(Law and Philosophy Library 38) Wojciech Sadurski (Auth.)-Freedom of Speech and Its Limits-Springer Net

FREEDOM OF SPEECH AND ITS LIMITS

Law and Philosophy Library VOLUME 38

Managing Editors

FRANCISCO J. LAPORTA, Department ofLaw, Autonomous University of Madrid, Spain ALEKSANDER PECZENIK, Department ofLaw. University of Lund. Sweden FREDERICK SCHAUER, John F. Kennedy School of Government. Harvard University. Cambridge. Mass .• U.S.A. Former Managing Editors AULIS AARNIO, MICHAEL D. BAYLESt, CONRAD D. JOHNSONt, ALAN MABE Editorial Advisory Board

AULIS AARNIO, Research Institute for Social Sciences. University ofTampere. Finland ZENON BANKOWSKY. Centre for Criminology and the Social and Philosophical Study of Law. University ofEdinburgh PAOLO COMANDUCCI, University of Genua. Italy ERNESTO GARZ6N VALDES, Institut flir Politikwissenschaft. Johannes Gutenberg Universitiit Mainz JOHN KLEINIG. Department ofLaw. Police Science and Criminal Justice Administration. John Jay College of Criminal Justice. City University ofNew York NEIL MacCORMICK, Centre for Criminology and the Social and Philosophical Study of Law. Faculty of Law. University ofEdinburgh WQJCIECH SADURSKI, Faculty ofLaw. University of Sydney ROBERT S. SUMMERS, School ofLaw. Cornell University CARL WELLMAN. Department of Philosophy. Washington University

WOJCIECH SADURSKI The University of Sydney. Australia

FREEDOM OF SPEECH AND ITS LIMITS

KLUWER ACADEMIC PUBLISHERS DORDRECHT/BOSTON/LONDON

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN-13: 978-1-4020-0281-6 e-ISBN-13: 978-94-010-9342-2 DOl: 10.1 007/978-94-010-9342-2

Published by Kluwer Academic Publishers, P.O. Box 17,3300 AA Dordrecht, The Netherlands. Sold and distributed in North, Central and South America by Kluwer Academic Publishers, 101 Philip Drive, Norwell, MA 02061, U.S.A. In all other countries, sold and distributed by Kluwer Academic Publishers, P.O. Box 322, 3300 AH Dordrecht, The Netherlands.

Printed on acid-free paper

All Rights Reserved © 1999 Kluwer Academic Publishers Softcover reprint of the hardcover 1st edition 1999 No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission from the copyright owner

TABLE OF CONTENTS

ACKNOWLEDGEMENTS PREFACE CHAPTER 1: JUSTIFICAnONS OF FREEDOM OF SPEECH 1 Search fQr Truth 2 Individ~al Autonomy 3 Democracy and Self-Government 4 Tolerance CHAPTER 2: SPEECH AND HARM 1 Levels of Scrutiny 2 Low and High Value Speech 3 Speech-Plus and Symbolic Action 4 Speech and Harm: A Case Study of R.A. V.

vii· 1 7 8 16 20 31 37 37 41 43 58

CHAPTER 3: SPEECH AND EQUALITY 1 Equal Opportunity and Public Speech 2 Silencing 3 Asymmetry of Fighting Words

73 73 98 III

CHAPTER 4: DISCRIMINATION AND ILLOCUTIONARY ACTS 1 Illocutions and Perlocutions 2 Authority in Discriminatory Illocutions

119 119 123

CHAPTER 5: VIEWPOINT NEUTRALITY AND ITS RATIONALES 1 Two Types of Neutrality 2 What is "A Viewpoint"? A Case Study of Rosenberger 3 Reasonableness and Viewpoint Regulations: A Case Study of Lamb's Chapel 4 Rationales for Viewpoint Neutrality and Subject-Matter Neutrality 5 The ContentlViewpoint Distinction and the Level of Generality 6 Indirect Viewpoint Discrimination 7 Paternalism and Intolerance

135 135 140 148 156 162 167 173

vi CHAPTER 6: RACIAL VILIFICATION AND FREEDOM OF SPEECH 1 The Context of the Hate-Speech Controversy 2 The Contours of "Racial Vilification" 3 The Harms of Hate Speech 4 Liberalism and Prohibitions of Hate Speech

179 179 187 195 217

INDEX

225

ACKNOWLEDGEMENTS This book owes its existence to a project which has produced several articles and papers presented to various academic audiences over the past few years. None of these papers are included here in their original form and, in the process of writing and revising them, I have incurred intellectual debts to a number of colleagues who have read and commented upon earlier versions of some parts of this book: to Kathryn Abrams, James Allan, Tom Campbell, Susan Dwyer, Alexandra George, Martin Krygier, Philip Pettit, Robert Post and Steven Shiffrin. I also wish to thank the audiences at the Australian National University, Catholic University of America, Cornell Law School, University of Hong Kong, McGill University, the Netherlands Institute of Advanced Studies, Saint Louis University, University of Tilburg, and University of Warsaw for giving me opportunities to present some of the ideas which later found their way into this book. A more tangible support was provided to me by the Australian Research Council and the Law Foundation of New South Wales in the form of research grants for which I am very grateful. I also thank Madeleine Cullen for her excellent work in producing the manuscript. I am very grateful to two institutions with which I have been associated while preparing this book: the Department of Jurisprudence at the University of Sydney Faculty of Law, of which I am a member, and Cornell Law School where I was a Visiting Professor in Spring Term 1995 and Fall Term 1996. Most of all, I wish to acknowledge the encouragement of my wife Anna to whom this book is dedicated.

PREFACE In authoritarian states, the discourse on freedom of speech, conducted by those opposed to non-democratic governments, focuses on the core aspects of this freedom: on a right to criticize the government, a right to advocate theories arid ideologies contrary to government-imposed orthodoxy, a right to demand institutional reforms, changes in politics, resignation of the incompetent and the corrupt from positions of authority. The claims for freedom of speech focus on those exercises of freedom that are most fundamental and most beneficial to citizens - and which are denied to them by the government. But in a by-andlarge democratic polity, where these fundamental benefits of freedom of speech are generally enjoyed by the citizens, the public and scholarly discourse on freedom of speech hovers about the peripheries of that freedom; the focus is on its outer boundaries rather than at the central territory of freedom of speech. Those borderline cases, in which people who are otherwise genuinely committed to the core aspects of freedom of speech may sincerely disagree, include pornography, racist hate speech and religious bigoted expressions, defamation of politicians and of private persons, contempt of court, incitement to violence, disclosure of military or commercial secrets, advertising of merchandise such as alcohol or cigarettes or of services and entertainment such as gambling and prostitution. Even the most ardent, card-carrying civil libertarians are not committed to an unconditional defence of a right to these types of public expressions, no matter what. But an attempt to give a principled, rather than an ad-hoc, answer about where to draw the line must fail unless it is based upon a rationale for having freedom of speech in the first place, and upon an understanding of the structure of the principle of freedom of speech. This rationale and this structure inform our understanding of the very central aspects of freedom of speech. In this way, a discussion about the peripheries is dependent on an understanding of the central values of freedom of speech. And, vice versa, thinking about those peripheries can illuminate our understanding of the very core of that freedom. Of course, the distinction between those aspects of freedom of speech that are central and those that are peripheral is itself contingent and often questionbegging. It is contingent because as we move from the core to the peripheries, the marginal benefit of a given type of expression decreases and the marginal harm increases, and yet the judgment of harm is contingent upon a broader moral theory about which people within a society may disagree. Be that as it may, there is no doubt that much speech is harmful, and that harms resulting from the exercise of freedom of speech may often prevail over its benefits just consider a catalogue of the types of speech mentioned in the first paragraph of this Preface.

2

PREFACE

That is why even very committed liberals must accept that the principle of freedom of speech, to use the words of a leading American constitutional scholar, "need not sanctify the deliberate infliction of pain simply because the vehicle used is verbal or symbolic rather than physical" and, therefore, that sometimes law "may create remedies for the damage done with words so long as these remedies display sufficient sensitivity to freedom of expression as well". 1 The purpose of this book is to place the arguments about legal restrictions of speech in a framework that facilitates legal reasoning about rights, and thus displays "sufficient sensitivity to freedom of expression". The content and the limits of the principle of freedom of speech must be informed by a rationale for adopting the principle in the first place. I begin the book with a discussion, in Chapter 1, of the main types of justification for having this principle. There is no doubt that the pursuit of truth, individual autonomy, democratic self-government and tolerance towards those with whom we disagree, are important values, and that there is at least some kind of relationship between a regime of freedom of speech and any of these important values. But this is hardly the point. The point is rather to determine which of these rationales justifies elevating the principle of freedom of speech to such a level as to exempt it from the ordinary operation of a harm calculus which instructs that those actions which produce more harm than good can be legitimately prohibited by law. As Frederick Schauer noted, in arguably the most important contemporary book on this subject in English-language literature, when the principle of freedom of speech is adopted, "a limitation of speech requires a stronger justification, or establishes a higher threshold, for limitations of speech than for limitations of other forms of conduct".2 In a more recent article, Schauer restated this point by formulating "the central question of free speech theory" as: "what are the grounds, if any, for entrenching a principle whose effect is to invalidate what would otherwise be a permissible form of governmental action?,,3 This suggests that the burden on the defenders of this principle is to produce not merely a rationale which connects a regime of freedom of speech with an attractive value (or with a set of values), but to show why this rationale justifies protecting free expression even if, at times, it may bring about more bad than good.

\.

2. 3.

Laurence H. Tribe, American Constitutional Law (Mineola, N.Y.: The Foundation Press, 1988, 2nd ed.), p. 856. Frederick Schauer, Free Speech: A Philosophical Enquiry (Cambridge: Cambridge University Press, (982), p. 8. Frederick Schauer, "Free Speech in a World of Private Power", in Freedom of Communication ed. by Tom D. Campbell & Wojciech Sadurski (Aldershot: Dartmouth, (994), pp. 1-2.

PREFACE

3

Chapter 2 applies this insight about the special weight of the principle of freedom of speech in competition with an ordinary harm calculus, to look at possible ways of defending restrictions on freedom of speech. The most promising strategies available to the defenders of such restrictions would be to show that some particular categories of speech should not enjoy such a high level of protection that they would be insulated from a calculus of harm caused by the expression in question. In other words, the strategies aim to show that restrictions of some speech warrant more lenient scrutiny than those restrictions which justifiably strike us as antithetical to central value,S of freedom of speech. The strategies for such an argument, discussed in Chapter 2, include the arguments that some categories of speech have a "lower value" than others, or that an expression that is not really "speech" but "conduct" should be subjected to those restrictions which fall properly on any other types of conduct in society. My focus then turns from the speech which is restricted to the nature of the restriction itself, as a criterion for lowering the scrutiny, and in the last part of Chapter 2, I look at the argument that restrictions which are neutral from the point of view of the content of restricted speech can be more acceptable (and thus subjected to a more lenient critical scrutiny) than viewpoint-based restrictions. This is because the harm of speech selected through its viewpoint may usually be mitigated by counter-speech, reflecting a counter-viewpoint. Therefore, the analysis of a speech-hann connection supports an intuitively strong aversion towards viewpoint-based regulations of speech. Chapters 3 and 4 deal with the relationship between speech and equality, and more specific!\lly with whether unrestricted freedom of speech can undennine equality. There are two main fonns of this argument, and they are examined in turn in these two chapters. First, in Chapter 3, I consider whether some categories of speech undermine other people's equality of expressive opportunity by the message they convey, and whether, accordingly, restrictions on these types of speech can be seen as a method of restoring the equality of opportunity, perhaps analogous to "affmnative action". In the process of considering this argument, I examine the character of "opportunity" in the area of speech, and in particular whether the rights to those plausibly defined expressive opportunities may impose some duties upon the listeners. I give a qualified positive answer to this last question, but I remain sceptical about the analogy between restrictions on speech and "affinnative action". There is one exception to this scepticism; namely a narrow category of so-called "fighting words", which may and should be proscribed in an "asymmetrical" way, and this asymmetry can be seen to resemble the principle of affinnative action. Chapter 4 considers an argument about the speech-equality relationship which is different to the argument examined in Chapter 3: rather than seeing the discriminatory character of some types of speech (notably, racist hate speech

4

PREFACE

and pornography) in the message they convey to their listeners, these types of speech are depicted as constituting acts of discrimination in themselves. For the wrongful discrimination to occur, there is no necessity for the speaker to convince the listener about anything. With the use of a theory of "illocutionary acts", I offer a critique of this analysis of the inequality produced by speech acts, and conclude that the analogy between hate speech and pornography on the one hand and the illocutionary acts of speech which unquestionably discriminate against some people on the other hand, regardless of the persuasion effect upon the listeners, is less than perfect. Earlier in the book, in the last part of Chapter 2,. I allude to a distinction between viewpoint-based, content-based, and content-neutral restrictions on speech, but my argument there is restricted to one specific theme related to the overall theme of that chapter; namely, to the connection between this distinction and harm analysis. Chapter 5 pursues this distinction in a more comprehensive way; the main question raised in this chapter is whether this distinction, and the related gradation of the levels of scrutiny of restrictions, illuminates something important about the moral characteristics of the principle of freedom of speech. My answer is positive, but I suggest that the difference between the levels of suspicion raised by these three types of restrictions only makes sense if we identify the types of moral defects typically associated with each of these varieties of speech restrictions. In particular, the distinction between those restrictions which are based on a viewpoint and those which are merely based on the subject-matter of a restricted speech, js significant in so far as it is founded on rationales for disliking these restrictions. I then argue that restrictions based on a viewpoint are typically motivated by intolerance to a given view, while restrictions based on a subject-matter are paternalistically motivated, and that there is a difference in the moral significance of these two types of motives for restrictions. Thoughout the book, I am concerned more about the structure of the principle of freedom of speech than about specific categories of speech and how they fare in the borderline territory of the principle: it is not my aim to declare whether this or that category of speech should be still protected by the principle of freedom of speech or whether they fall on the proscribable side of the borderline. But in the last chapter I depart from this approach, and consider in detail the arguments related to prohibitions of racist hate speech. This is because the question of racial vilification raises particularly difficult dilemmas for liberal defenders of freedom of speech, and it is with regard to this category of speech that, as one writer said, liberalism's ''theoretical vocabulary is least

PREFACE

5

compelling ... and least responsive to real world conditions".4 To examine this vexed issue, I bring together various insights from the earlier chapters, and suggest a way in which various arguments for and against the ban can be assessed. The conclusion, that such bans are usually unjustified under a plausible understanding of freedom of speech, is tentative and qualified; the purpose of the chapter is to explore the structure of the argument about harms of racist vilification, rather than to announce the author's verdicts about how the law should deal with this case of harmful speech. As this brief outline suggests, there is much about "the limits to freedom of speech" that is left beyond the ambit of this book. A number of issues are mentioned only passingly, or unexplored altogether: while legal regulations of pornography and hate speech receive a degree of attention, themes such as regulation of broadcast media, cable television, the Internet and workplace verbal harassment do not. If the aim of the book was to produce a comprehensive picture of how a principle of robust freedom of speech applies across a broad range of currently controversial matters, such an omission would be unforgivable. But no such ambition guided me in writing this short book; rather, I focus on a narrow range of illustrations aimed to serve the purpose of inquiring into the doctrinal complexities of the principle of freedom of speech. The selection of topics in a book of this size is, by necessity, arbitrary, and the reader should be warned about it in advance. Another caveat is perhaps also required: the aspiration here is not to provide yet another "general theory" of freedom of speech which would attempt to organize all the currently controversial topics around a single regulating maxim which can be easily stated (and as easily rebutted). Should some readers see it as a weakness, it may be mitigated by a sincere admission here, right at the outset. Finally, the book is not intended to be an interpretation of one particular legal system but rather a generalization at a theoretical level of some examples of the protection of freedom of speech taken from different systems and jurisdictions. As it turns out, by far the greatest number of illustrations of specific rules and decisions are taken from the United States. This should not be surprising. The body of judicial and scholarly doctrine generated by the First Amendment to the Constitution of the United States is by far the most influential and elaborate development of the principle of freedom of speech, and it provides a fruitful point of reference both as a positive inspiration and as a target of criticism. As the Canadian Supreme Court said in an important decision on freedom of speech: "In the United States, a collection of fundamental rights has been constitutionally protected for over 200 years. The

4.

Toni M. Massaro, "Equality and Freedom of Expression: The Hate Speech Dilemma", William and Mary Law Review 32 (1991): 211-265, p. 230.

6

PREFACE

resulting practical and theoretical experience is immense, and should not be overlooked by Canadian courts". s This may explain why so many examples of specific rules and decisions are taken from the United States. But these illustrations are meant to be just that: illustrations which serve the purpose of a broader theoretical argument rather than explorations into the actual set of authoritatively valid rules regarding freedom of speech in the United States. At times, my argument involves a rather detailed analysis of some specific court decisions. This is because, in my view, a detailed exploration of judicial reasoning may often help us detect the complex significance of a general principle in action, and see the conundrums and dilemmas rarely evident on the surface of a principle or of a doctrine. Again, the ultimate purpose is not to "criticize" or ''praise'' this or that judge for a "correct" opinion but rather to use their arguments on a par with the arguments put forward in scholarly discourse. Readers who are put off by such an exegesis are advised to omit Chapters 2.4,5.2 and 5.3, and go immediately to the concluding paragraphs of these sections of the book.

s.

R. v. Keegstra (1991) 61 C.C.C. 3d 1,32 (Dickson C.l.C.)

CHAPTER 1 JUSTIFICATIONS OF FREEDOM OF SPEECH Suppose someone puts forward a principle that speech which depicts or even advocates an odious social relationship (such as one that involves domination, subordination, or violence) should be suppressed on the basis of the same reasons which argue for suppression of the relationship itself. Surely no remotely robust system of free speech can survive under this principle. It is important to understand why this principle is false: the reasons that the government has to enforce freedom of speech are not equivalent to the set of reasons that people have for speaking. Stanley Fish recently suggested that freedom of speech would only be an important value if speech were inconsequential, that is, if the reason people had for speaking was speaking for its own sake: "[F]reedom of expression would only be a primary value if it didn't matter what was said, didn't matter in the sense that no one gave a damn but just liked to hear talk". 1 Since, Fish argues, people speak not only to proffer propositions but mainly to achieve specific purposes, and those purposes may be defeated by some forms of speech, freedom of speech is not valuable in itself: free expression could only be a primary value if what you are valuing is the right to make noise; but if you are engaged in some purposive activity in the course of which speech happens to be produced, sooner or later you will come to a point when you decide that some forms of speech do not further but endanger that purpose. 2

Such a conclusion would only follow if the rationale for freedom of speech was a mere aggregate of rationales for particular acts of speech, or, in other words, if there was a continuity between the legislative motive (for protecting citizens' rights to speak) and individual motives (for exercising these rights). But this is not the case. First, and most importantly, governmental motives underlying decisions about limits on the right of free expression are properly determined not only by a value assessment of the purposes for individual expression (in other words, of the value of the state of affairs produced by an aggregate of individual ideals), but also by what Ronald Dworkin calls "constitutive" reasons for freedom of speech: respect for individual autonomy, responsibility,

1.

2.

Stanley Fish, There's No Such Thing as Free Speech and It's a Good Thing, Too (New York: Oxford University Press, 1994), p. 106. Id., p. 107.

CHAPTER 1

8

etc.3 Second, individual motives for advocating certain states of affairs are not always properly described as the purpose for bringing about that state of affairs. There may be various legitimate strategic reasons for advocacy: attempts to test public opinion, to strengthen an under-represented point of view, a "devil' s advocate" type of argument, etc. The alternative, suggested. by Fish, that either speech is inconsequential or that freedom of speech lacks foundation, is therefore false. If the reasons for freedom of speech are distinct from the reasons for specific acts 'of speech, then the reasons for the suppression of certain states of affairs are distinct from the reasons for the suppression of views which advocate those states of affairs. But what are the reasons for freedom of speech? Disagreements about those reasons will affect disagreements about which categories of speech deserve strong protection. The purpose of this chapter is to review the main justifications for a strong protection of freedom of expression, and to examine their implications for the degree of scrutiny that should be accorded to laws that restrict speech.

1

Search for Truth

The traditional Millian justification, developed in the second chapter of "On Liberty" was subsequently transformed in American First Amendment jurisprudence into the "marketplace of ideas" slogan, and is best known through the famous words by Justice Holmes: "the best test of truth is the power of the thought to get itself accepted in the competition of the market".4 But, taken in isolation from other rationales, the purpose of seeking the truth supports a distressingly narrow scope for free expression. Most of us would probably find a regime of freedom of speech governed solely by the search-fortruth purpose excessively restrictive and, in any event, much more restrictive than the actual scope of freedom of speech in contemporary liberal-democratic states. Bollinger's observation that "the chance that the Nazi messages may tum out to be 'true' is hardly a persuasive basis on which to defend such

3.

4.

Ronald Dworkin, Freedom's Law: The Moral Reading of the American Constitution (Oxford: Oxford University Press, 1996), pp. 200-201. Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting).

JUSTIFICATIONS OF FREEDOM OF SPEECH

9

speech"5 applies to a broad range of messages which we would, nevertheless, wish to protect against legal prosecution. The fact that the "search for truth" rationale would justify a high degree of censorship indicates that there is a clash between the dominance of this rationale and widely shared intuitions about freedom of speech. As Judge Easterbrook wrote: "[T]he Constitution does not make the dominance of truth a necessary condition of freedom of speech".6 Nor, one might add, is it the inevitable outcome of freedom of speech, though we might well hope that more often than not it will be the case. To give some substance to this hope, we should distinguish among different types of speech, depending on their relevance to "the search for truth" and regardless of how indirect and remote this relationship may be. A crude distinction between statements of fact and expressions of opinion may be simplistic and liable to lead to a wide range of borderline cases. Yet it indicates the type of argument which questions the relevance of the "search for truth" rationale as applied to a general system of freedom of speech, and not with regard to specific areas such as speech in academic contexts. Notwithstanding the borderline cases, there are clear examples at both extremes of the factopinion spectrum which suggest the principle that underlies the distinction. Statements such as: "Cigarettes XXX contain 0.5 % tar", or "Joe Smith stole a pair of Nike running shoes in a department store on 1 February 1997" can hardly claim special protection, under the "search for truth" rationale, if demonstrably false. Indeed, we would expect the government to protect us against the unrestricted publication of lies of this type. A decision to prosecute or restrict false commercial advertising may well be based on a simple insight that individual citizens, each in isolation, have insufficient resources to test the veracity of an advertisement which makes representations about the volume of tar in cigarettes. Accordingly, to solve this collective-action problem citizens decide to endow a state authority with the power of testing its truthfulness on their behalf, and restrict the publication of false ads. Accompanied by the low social value of having unrestrained access to deceptive or misleading ads, this results in a decision about the rightness of the decision about a restraint (prior, or otherwise). This reasoning still places itself within the logic of the marketplace theory, except that market failure justifies state intervention in this case. And "market failure" similarly occurs in situations where defamatory

5.

6.

Lee Bollinger, The Tolerant SOciety: Freedom of Speech and Extremist Speech in America (London: Oxford University Press, 1986), p. 54; see, similarly, D.F.B. Tucker, Law, Liberalism and Free Speech (Totowa, N.J.: Rowman & Allanheld, 1985), pp. 15-18. American Booksellers Association v. Hudnut, 771 F.2d 323, 330 (7th Cir. 1985).

10

CHAPTERl

statements are based on false propositions and which concern the private person: the power of the market to make the truth prevail is extremely doubtful, since we (as interested individuals) lack sufficient incentives and resources to thoroughly investigate the truthfulness of the grounds of attacks upon private reputations. But now consider a different pair of statements: "There is nothing wrong with smoking, even if it shortens one's life; it is better to live happily, if shortly" and ''The author of this book has some very simplistic views about the law of freedom of speech". These are the sort of opinions that are hardly amenable to the "truth" analysis. They both come close to the "pure opinion" end of the spectrum. In such situations, surely we would not like the government to even begin considering the conduciveness of these statements to truth, just as we would not want our law to decide for us the "truthfulness" of statements such as "socialism is better than capitalism", or vice versa. As a useful tool for distinguishing between "facts" and "opinions", and a way of overcoming the "there are no crude facts out there" objection, one perhaps might apply Robert Post's distinction. Statements about facts, Post says, are those statements with regard to which we can expect a consensus to emerge about their truth or falsity; the criteria of truth are not dependent upon any particular community standards. In contrast, "judgments" make claims which are founded upon standards of a particular community, and which are not necessarily shared by the public at large.' Post further explains the First Amendment privilege for opinions (but not for statements of fact) by a concern for "the preservation of the neutrality of public discourse from the domination of community mores".8 But here is a paradox: legal protection for true statements of facts (but not for false ones) may be consistent with the marketplace of ideas theory because false statements of fact "interfere with the truth-seeking function of the marketplace of ideas".9 But it is hardly needed. Surely if the "factness" of a statement is defmed by its conduciveness to consensus, then why bother to support it with legal protection? However this argument is too quick a rebuttal: what Post has in mind is not an actual consensus about specific propositions, but rather a consensus about the standards for judging the truthfulness of these propositions. In the case of statements of fact, the standards are such that "any person, regardless of her

7.

8. 9.

R. C. Post, "'The Constitutional Concept of Public Discourse: Outrageous Opinion, Democratic Deliberation, and Hustler Magazine v. Falwelr, Harvard Law Review 103 (1990): 601-86, pp. 657-61. Id., p. 660. Hustler Magazine v. Falwell, 485 U.S. 46, 52 (1988).

JUSTIFICATIONS OF FREEDOM OF SPEECH

11

specific community, should in theory ultimately come to accept [a particular claim] if confronted with the relevant evidence".IO Contrary to Mill's argument, legal protection for true statements does not seem to detract from the dominance of truth. Whether legal protection for statements of fact should extend to false statements is a matter which probably cannot be determined in abstracto on the grounds of the theory of a "search for truth" alone. Consider, for example, the journey that the Supreme Court of the United States travelled, from asserting, with the authority of a quotation from Mill, that "[e]ven a false statement may be deemed to make a valuable contribution to public debate" 11 to an assertion that "[f]alse statements of fact are particularly valueless; they interfere with the truth-seeking function of the marketplace of ideas".12 The latter statement, in Falwell, is then somewhat qualified by recognition that "even though falsehoods have little value", nevertheless some tolerance for them is necessary in order to provide "breathing space", because strict liability for false factual assertions would create a "chilling effect" and result in the deterrence of true statements. This is the same argument which accompanied a celebration of the value of false statements, in New York Times v. Sullivan: limiting legal protection to only true statements would create a risk of "self censorship". However, in Sullivan a Millian rationale was also cited for the protection of falsity; the rationale dropped in Falwell. But here is a conundrum: if (relative) legal protection for false statements is based only on the "chilling effect" argument, then how do we know that the inevitable presence of some protected false statements in public discourse is less damaging to the prevalence of truth than the inevitable absence of some true statements as a result of the "chilling effect"? The reason why we must consider this question is as follows: the chilling effects argument is based on a particular choice of error risk. If a legal system only protects true statements of fact, then it risks underprotecting true statements (because some true statements will not find their way through to public discourse if they are difficult to prove in courts, and consequently the speakers will exercise "self censorship"). The error that the system risks is the error of underprotection for truth. But in a system of relative protection for false statements (supported, as it is, by the "breathing space" argument), the

10. Post, p. 661. 11. New York Times v. Sullivan, 376 U.S. 255,279 n. 19 (1964). 12. Hustler Magazine v. Falwell, 485 U.S. 46, 52 (1988).

12

CHAPTER 1

risk is of providing over-protection for statements of fact, as a result of which some false statements will receive protection. The decision to choose the former legal regime over the latter must therefore rest upon a calculation that the error of underprotection for truth is more damaging to the dominance of truth than a regime of overprotection, reaching into the protection of falsity. The error, one way or other, is inevitable, but why is it necessarily better for the pursuit of truth to err on the side of protection of falsity in the design of legal liability? An argument to support such a system would probably have to rely upon the sociological observation that the presence of falsity has a comparatively less pervasive effect on people's minds, in reaching their own views about a given matter, than the less-than-full presence of the truth. But it suffices to state the proposition in such general terms to realize how bizarre it is. A public presence of falsity can have enormous distorting effects upon the formation of public views about facts, and there is no reason to believe that this distorting effect is, as a general rule, less damaging to the truth itself. It could be less damaging if one accepted Mill's argument that the presence of false propositions helps to maintain the true views in a better shape, thanks to their collisions with errors. But not many people are prepared to have faith in the salutary function of such a collision, and one fears that if the false statements have sufficient amplifying volume, in propaganda and mass media, they can often win the battle. In any event, the Supreme Court of the United States, for one, having dropped the Millian factor from its recognition of the value of falsehoods, enjoys no benefit from falling back upon this rationale. The only other plausible explanation for such a choice of risk of error is by appeal to politics, not to the truth. When statements of fact concern the performance of public figures, one may argue that it is less politically harmful to risk underprotecting politicians against defamatory falsehoods than it is to risk overprotection, which would also deter true statements about politicians. Such a judgment is eminently reasonable: democratic polity needs free, undeterred public criticism and perceptive, investigative media. If an error in designing the legal regime, one way or the other, is bound to be made (and it is inevitable) then democracy is better off with some falsehoods about its politicians than with some truth deterred. But note that this conclusion has nothing to do with the "search for truth" rationale: it is a purely political judgment. And it is only applicable to some falsehoods: those expressed in the context of speech about public figures in connection with their role in governance. It does not provide for a general rule about the choice of error in the design of a legal regime, across the board.

JUSTIFICATIONS OF FREEDOM OF SPEECH

13

Now is the dominance of truth related to judgments which are statements of opinion (even in Post's sense, that is, statements which rely upon standards of a specific community)? It seems that in these cases, the "search for truth" argument is simply not applicable: it neither supports nor opposes governmental interference with speech. Or it may do both these things. The "pure opinion" argument is, therefore, double-edged. It might justify stronger protection for opinions because they do not easily lend themselves to falsification, and so cannot be shown to be. directly detrimental to the search for truth. But then, they can be denied special protection precisely because they do not contribute to the goal of truth winning out in a competition of ideas. They simply do not lend themselves to a truth analysis. Hence, depending on your perspective, Justice Powell's statement that "[u]nder the First Amendment there is no such thing as a false idea",13 might as well be read as saying that "there is no such thing as a true idea". The operative words are, of course, "under the First Amendment", and they indicate that the choice confronting us is political and legal, not epistemological. There are two possible ways out of this dilemma. One is to recite the presumption for non-restraint: it is restraint, not liberty, which calls for justification. In the present context, however, it becomes clear that the presumption itself must have some justification. Therefore there is another response: such a justification, and consequently a decision about the implications for the "search for truth" theory, must be found in the rationales for freedom of speech other than the "search for truth" theory alone. According to some writers, the search for truth theory is ultimately based on the Millian argument about uncertainty, and on the virtue of skepticism. 14 But surely there are cases when uncertainty about the truth of a particular proposition does not follow from a respectable philosophical attitude of doubting. As Frederick Schauer observes, "[ slome people may have so little confidence in their own factual judgments that they always sit down gingerly for fear that the chair they see is only an apparition, but we hardly need tailor first amendment doctrine around them".IS This sort of skepticism is not the right basis for non-intrusion. Conversely, the decision about interference may be silly for all sorts of reasons, but conceit about its own infallibility is the least

13. Gertz v. Robert Welch. Inc., 418 U.S. 323, 339 (1974). 14. See Jeffrey E. Thomas, "Statements of Fact, Statements of Opinion, and the First Amendment", California Law Review 74 (1986): 1001-1056, p. 1022. 15. Frederick Schauer, "Categories and the First Amendment: A Play in Three Acts", Vanderbilt Law Review 34 (1981): 265-307, p. 287; see also K. Greenawalt, "Free Speech Justifications", Columbia Law Review 89 (1989): 119-56, p. 132.

14

CHAPTER 1

likely of them. This is the point in a celebrated rebuttal of Mill's "infallibility" argument by James Fitzjames Stephen. Mill, in one of his most famous propositions on freedom of speech, declared: "[I]f any opinion is compelled to be silenced, that opinion may, for aught we can certainly know, be true. To deny this is to assume our own infallibility".16 The non sequitur of this proposition was pinpointed by Stephen: [T]here are innumerable propositions on which a man may have a rational assurance that he is right whether others are or are not at liberty to contradict him, and that although he does not claim infallibility.... There are plenty of reasons for not forbidding people to deny the existence of London Bridge and the river Thames, but the fear that the proof of those propositions would be weakened or that the person making the law would claim infallibility is not among the number. 17

The "infallibility" stick is too crude a weapon with which to attack the proponents of restraints on speech. If logically extended, it would undermine not only the legitimacy of restrictions on freedom of speech, but also any restrictions on any human freedom, simpliciter. As Tom Campbell says, if we take the assumption of fallibility seriously, then we are not going to be in a position to interfere in the conduct of any harm-causing person because we may be wrong to allege that the conduct causes harm. While this possibility may make us hesitant to enforce conduct, it does not exclude such intervention where the evidence of potential harm is strong. IS

Just as not all suppressions of speech are based on perceptions of falsity of the ideas expressed, neither is respect for speech always based on respect for truth. Truth or falsity, even if discernible beyond a reasonable doubt, may sometimes be irrelevant to a decision about restraint or toleration of speech. Indeed, in some circumstances the argument for restraint may be strengthened by the truth of statements. The truthfulness of public depictions of a private person's intimate life may be more damaging to that person's privacy than similar depictions that are generally understood to be a deliberate lie. Publishing a

16. John Stuart Mill, "On Liberty", in On Liberty and Other Writings, ed. by Stefan Collini (Cambridge: Cambridge University Press, 1989), p. 53. 17. James Fitzjames Stephen, Liberty, Equality, Fraternity, ed. by Stuart D. Warner (Indianapolis: Liberty Fund, 1993), pp. 26-27. 18. Tom Campbell, "Rationales for Freedom of Communication", in Freedom of Communication ed. by Tom Campbell and Wojciech Sadurski (Aldershot: Dartmouth, 1994), pp. 26-27.

JUSTIFICATIONS OF FREEDOM OF SPEECH

15

story about a public person's disreputable, though long abandoned, past lifestyle, is all the more despicable if retold with the utmost attention paid to factual accuracy. So much for skepticism about facts as a possible ground for freedom of speech. What about the skepticism about opinions and judgments? Clearly, this argument does not belong to the "search for truth" (understood as truth about verifiable facts) category of justifications but is rather a form of argument for toleration based on our uncertainty about our own views in light of the diversity of views and opinions in our, and in other, societies. This attitude is sometimes mistakenly taken to be a case of "moral relativism"; however, it is not a matter of moral relativism, but rather of skepticism. The fact that moral relativism does not provide a ground for toleration seems clear. Recall Bernard Williams on "vulgar relativism", that is, a combination of the proposition "that 'right' means ... 'right for a given society'" with the proposition "that (therefore) it is wrong for people in one society to condemn, interfere with, etc., the values of another society". 19 The inconsistency lies in the fact that the latter proposition uses a nonrelative sense of "right" disallowed by the first proposition: a "logically unhappy attachment of a nonrelative morality of toleration or noninterference to a view of morality as relative".20 . Such a contradiction does not arise when the grounds for tolerance are presented as a matter of moral skepticism, and yet, there is no clear inference from skepticism to toleration; indeed, skepticism may well be reconciled with the suppression of and intolerance for other people's views and lifestyles. A skeptical attitude, expressed more often as a question than as an assertion ("Who are we to know?"), is unhelpful as a basis for freedom of speech. This is not to deny that, as a matter of moral psychology, many people may find good rationalizations for their tolerant attitudes in a skeptical stance. Lee Bollinger may be right that "[a] person who feels healthy self-doubt toward her beliefs is a person who is not just prone to let more speech go uncensored, but a person who is going to behave differently, to insist on less and to compromise more, in the give-and-take of the political process.,,21 But it is important to note that even if it is true that tolerant attitudes are often supported by appeals to skepticism about truth (moral or otherwise), the reverse is not the case. That is,

19. Bernard Williams, "An Inconsistent Fonn of Relativism", in Relativism, Cognitive and Moral ed. by J.W. Meiland and M. Krausz (Notre Dame: University of Notre Dame Press, 1982), p. 171. 20. Id., p. 172. 21. Lee Bollinger, "The Tolerant Society: A Response to Critics", CCJiumbia Law Review 90 (1990): 979-1001, p. 985.

16

CHAPTERl

it is not the case that the adoption of a position of skepticism leads to tolerant attitudes. There is no such connection; rather, depending on other assumptions, skepticism may either lead to tolerance or to a suppression of unorthodox views. This latter conclusion was advocated, for example, by Hobbes who argued for the need to enforce uniformity in religious matters not from· the position of religious truth, but by using arguments about public order and peace.22 Indeed, as a conceptual proposition, there is no necessary connection between agnosticism about the truth and toleration for other views. Precisely because I cannot judge between your view and mine, I may say that while I cannot infer the superiority of my view, neither can I make a conclusion about the superiority of yours. Hence, I may impose my views on bases other than their truth. If I am uncertain about the possibility of proving the validity of my own judgments (an alleged skeptical basis for freedom of speech), then it must be because I am skeptical about the possibility of proving the validity of all judgments: otherwise it would be irrational for me not to accept those judgments which I feel more confident about. Hence, I cannot recognize the possibility of another person's capacity to prove the validity of her judgments, which are different from mine. To argue for my duty of tolerance for her judgments from such uncertainty would be to allow, in an arbitrary fashion, her judgments to prevail over mine, although both are equally unprovable. 23 There is, therefore, no connection between skepticism and freedom of speech, and the gap must be bridged by an appeal to some other grounds, for instance, that these are her judgments and that they concern her life more than mine. But these grounds have nothing to do with moral skepticism.

2

Individual Autonomy

The emphasis on the importance of self-expression and self-fulfillment of individuals through their autonomous action is one of the defining features of liberal theory. It has been variably expressed through different liberal maxims: the equal moral agency of individuals, the neutrality of the state towards particular conceptions of the good, and the priority of the right over the good.

22. See Alan Ryan, "A More Tolerant Hobbes?", in Justifying Toleration ed. by Susan Mendus (Cambridge: Cambridge University Press, 1988), ch. 2. On the impossibility of showing that skepticism leads to toleration see also Joseph Raz, "Liberalism, Skepticism, and Democracy", [owaLaw Review 74 (1989): 761-86, pp. 762-66. 23. This parallels an argument by Geoffi'ey Harrison, "Relativism and Tolerance", in Relativism. Cognitive and Moral, p. 235.

JUSTIFICATIONS OF FREEDOM OF SPEECH

17

But regardless of the idiom used, it boils down to the fundamental principle that autonomy of action is, in general, a more important good in liberal theory than the worth of the substantive consequences of one's action, precisely because it is a condition of human self-expression and self-fulfillment. In other words, the value of autonomy, as a rule, prevails over the disvalue which specific consequences of an autonomous action may have. There is little doubt that human communicative activities are crucial to our capacity for self-expression and self-fulfillment. They are crucial for selfexpression because speech is the most direct way of communicating to the rest of society who we really are. The messages we convey to other people about our identity form part of a process that includes feedback from other people, which in turn modifies our own self-perception. Communication is in the center of this complex process. And it is crucial for self-fulfillment because the exercise of our capacities is only made possible through self-defmition, and the determination of who we really are is impossible without open communication with our fellow human beings. It also allows our self-fulfillment in a trivial sense: there are many of us who rightly or wrongly perceive our best skills and capacities in areas connected with communicating our ideas to other people. To some, this link between the regime of protection of speech and the aim of self-fulfillment, will sound hollow. As a general theory of freedom of expression, one may perhaps say that the aims of self-expression and self fulfillment are patently inadequate. We may wish to protect the right of journalists to defame public figures (within limits) but we will not advance our cause by appealing to a journalist's need for self-expression. Generally, it seems that in most cases of political speech, where we face dilemmas about limits of protection for speech, appeals to the self-expression of speakers are off the mark. Ironically, they even seem to trivialize the importance of the speech in question. Speech as a vehicle of self-fulfillment has no other immediate aims than to be uttered and heard, while we believe that often the "point" of speech lies in its capacity to bring about the results desired by the speaker (or avert the consequences feared by the speaker). Speech as a means of self-expression actualizes its value, as it were, in the act of speech itself, and no wonder the United States Supreme Court once, while approvingly referring to "[t]he freedom to speak one's mind" as an aspect of (inter alia) individual liberty, suggested that it was "thus a good unto itself,.24 Yet, intuitively, it seems that in a great number of cases we defend a robust system of protection for freedom of speech not because we consider it to be "a good unto itself' or

24. Bose Corp. v. Consumers Union of United States, 466 U.S. 485, 503 (1984).

18

CHAPTERl

valuable insofar as it is an aspect of individual liberty, but rather because it is instrumental to values which may well be necessary to protect individual liberty, but which, strictly speaking, are not a part of the definition of individual liberty. Arguably, this is not a very persuasive objection: an appeal to intuitions does not give a great deal of mileage to a critique of the moral theory of freedom of speech. But the real problem about self-expression theory is not that it relies on intuitions which may not be very widely shared. The real problem is that, as a rationale, it is incapable of supplying the reasons for subjecting speech to a more lenient system of legal control than many other aspects of individual behaviour which may also be essential to one's self-expression and self-realization. If the "self-expression" rationale for freedom of speech is derived from a general underlying principle of individual autonomy (and what else can it be derived from?), then it must also share its limitations. If the reason for having freedom of expression is that it allows self-fulfillment, then self-fulfillment also determines the limits of expression. And these limits are the same as in the cas~ of any "self-expressing" or "self-fulfilling" conduct. If one believes that the "harm to others" principle best grasps a view about the limits of state coercive power towards the individual, then "harm to others", as a practical standard of protecting individual autonomy, must be applied to freedom of speech. This is because speech is an important, but not the only, means of self-expression and self-realization; indeed, compared to other aspects of individual behaviour, there is nothing particularly outstanding about speech, as a vehicle of expressing an agent's identity, desires, preferences etc, to others. Punching another person's nose or driving 120 kIn per hour through a school district during peak hours may be the best way for you to express your real self, or to fulfill your true capacities, but we will not allow you to exercise your autonomy in this particular way. We will be justified in preventing you from doing so even if your self-respect will be damaged as a result of our interference with your exercise of capacities central to your self-defmition. By the same token, insults directed towards members of racial minorities, publication of demeaning, sexist pictures, or the propagation of doctrines which we know may cause people to doubt the value of democracy, may be necessary for one's sense of unrestrained self-expression and self-fulfillment. But if the reason to protect freedom of speech is to enable the self-expression of speakers, then those instances of "bad" speech must be subjected to the same limits that we accept with regard to any conduct, communicative or otherwise. Hence an appeal to the values of self-expression and self-fulfillment is unhelpful to those who would apply a higher standard of scrutiny to laws that restrict freedom of speech. It does not follow that these values do not constitute an important part

JUSTIFICATIONS OF FREEDOM OF SPEECH

19

of the rationale for free speech. But it does follow that those who would insulate speech from the ordinary operation of the harm principle, or of the utilitarian calculus, or of any other standard that attempts to support the moral limits of state action vis-a-vis an individual, must find their rationale elsewhere. It may reside in the concern for the hearer's autonomy, and not that of the speaker, as the argument considered so far suggests. Thomas Scanlon appealed to "the sense of autonomy" (albeit, as he admitted, in an "extremely weak" sense) when defending the "Millian principle" with respect to freedom of expression. The principle, in Scanlon's version, was meant to place certain harms outside the scope of justification for legal coercion. Those harms which, under the Millian-Scanlonian principle, do not trigger legal coercion, are "harms to certain individuals which consist in their coming to have false beliefs as a result of those acts of expression", as well as harms resulting from the acts of expression ''where the connection between the acts of expression and the subsequent harmful acts consist merely in the fact that the expression led the agents to believe ... these acts to be worth performing". 25 I will have more to say about the substance of Scanlon's proposal later in Chapter 6.3.2, when discussing the question of advocacy to racial hatred. At present it is important to note that Scanlon's proposal that we do not count certain expression-related harms as harms which should trigger legal coercion is based on a respect for the hearers of those expressions. "The harm of coming to have false beliefs is not one that an autonomous man could allow the state to protect him against through restrictions on expression", claims Scanlon,26 and a similar claim is made for outlawing the advocacy of harmful conduct. But, even putting to one side Scanlon's contentious appeal to hypothetical contract, an appeal to the autonomy of the hearer resonates with a common intuition about what it means to be a free individual in a society. It is, among other things, not to be "protected" against the views which may be "bad for you", because they are false or they will lead you to commit violent or otherwise illegal actions. As stated, and as Scanlon himself has subsequently recognized, the principle is evidently overdrawn.2' We may well agree that the state should protect us against both some false beliefs and advocacy to crime. As to the first

25. Thomas Scanlon, "A Theory of Freedom of Expression", Philosophy & Public Affairs I (1972): 204-26, p. 213. 26. Id., p. 217. 27. T. M. Scanlon, "Freedom of Expression and Categories of Expression", UniverSity of Pittsburgh Law Review 40 (1979): 519-50, p. 532.

20

CHAPTER 1

hypothetical consent, harm consisting of a false belief that cigarettes do not cause cancer may be the sort of harm which we may want the government to protect us against. It is because, as suggested earlier, some false beliefs are enormously difficult to verify; the verification, while costly, may be reachable and reliable; the power of censorship may be seen as a rational solution to this particular collective-action problem. Or consider the vexed problem of Holocaust denial laws, such as those in Germany where it is illegal to publicly deny the fact of the Holocaust. I am not convinced about the wisdom of those laws, but one may imagine the following line of argument: the fact of the Holocaust is one that we can be reliably certain actually happened; the expense and effort involved in independently verifying this fact are excessively high for a single individual; the dangers of permitting widespread doubts about whether the Holocaust took place are very significant. Hence, we may, as a community, agree to be protected against views propounding those "false beliefs". A more persuasive argument can probably be made with regard to the link between advocacy and harms resulting from advocated action: we may want to protect ourselves against the harm of some of us being persuaded to commit harmful acts by others. But, in any event, this is not to say that the main thrust of Scanlon's principle is counterintuitive. I, for one, fmd it rather convincing. But the argument does not appeal to the value of self-expression or selfrealization, because these values are actualized on the side of the speaker. The values concerning the hearer are those of unconstrained access to the views and information held and conveyed by others. Those values call for a grounding in a broader theory about what it is that makes them so important as to convince us to depart from our usual standard of justification for legal coercion. Such a theory has to be supplied, and an appeal to the values of self-government, or an appeal to our reluctance to endow the government with censorship powers, seem to be good candidates for such theory. Therefore, it is to these conceptions that we will now turn.

3

Democracy and Self-Government

Discussions concerning this rationale for freedom of speech usually refer to the theory of Alexander Meiklejohn and his followers. 28 The argument is straightforward: democracy requires that citizens be free to receive all information which may affect their choices in the process of collective

28. See, in particular, Alexander Meiklejohn, Free Speech and Its Relation to Self-Government (port Washington: Kennikat Press, 1948).

mSTIFICATIONS OF FREEDOM OF SPEECH

21

decision-making and, in particular, in the voting process. After all, the legitimacy of a democratic state is based on the free decisions taken by its citizens regarding all collective action. Consequently, all speech that is related to this collective self-determination by free people must enjoy absolute (or near-absolute) protection. According to a leading contemporary proponent of the theory, "[s]peech that concerns governmental processes is entitled to the highest level of protection; speech that has little or nothing to do with public affairs may be accorded less protection".29 This applies not only to issues related to governmental processes, but more generally to issues of public concern, which in Meiklejohn's view extend to areas such as philosophy, literature and the arts. For, as Meiklejohn himself believed, "there are many forms of thought and expression within the range of human communications from which the voter derives the knowledge, intelligence, sensitivity to human values: the capacity for sane and objective judgment which, so far as possible, a ballot should express". 30 He went on to list four such areas which "must suffer no abridgment of their freedom": education, philosophy and the sciences, literature and the arts, and public discussions of public issues. But note that it is an individual qua "the voter", and her judgment through "a ballot", which are the focus of Meiklejohn's concerns. Meiklejohn was unequivocal about this specific, voting-related function of the free speech principle: responding to Kalven's critique, he confirmed his belief "that the people do need novels and dramas and paintings and poems, 'because they will be called upon to vote,,,.31 This point about the link between freedom of speech and the goal of self government has been taken up by a number of commentators. They observed that a "democratic self-government" theory is either unduly restrictive (by providing a rationale for protection only to political speech sensu stricto)32 or virtually meaningless (because it broadens the meaning of the "political" so as

29. Cass R. Sunstein, "Pornography and the First Amendment", Duke Law Journal (1986): 589627, p. 603, footnotes omitted. For other examples of political-process theory of First Amendment values, see Robert H. Bork, "Neutral Principles and Some First Amendment Problems", Indiana Law Journal 47 (1971): 1-35 and Lilian R. BeVier, "The First Amendment and Political Speech: An Inquiry Into the Substance and Limits of Principle", Stanford Law Review 30 (1978): 299-358. 30. Alexander Meiklejohn, "The First Amendment Is an Absolute", Supreme Court Review (1961): 245-266, p. 256. 31. (d, p. 263, quoting Harry Kalven, "The Metaphysics of the Law of Obscenity", Supreme Court Review (1960): 1-45, p. 16. 32. See David A.J. Richards, "Free Speech and Obscenity Law: Toward a Moral Theory of the First Amendment", University of Pennsylvania Law Review 123 (1974): 45-91, pp. 68-69; see also Tucker, pp. 26-27.

CHAPTER 1 to include in it every act of expression which deserves protection, regardless of its place in the political process).33 Indeed, it seems that there is some incongruency between the rationale and the scope of the protection: either the rationale is too narrow (thus leaving much of the worthy speech unprotected) or there is a lack of fit between the rationale and the breadth of protection. The first hom of the dilemma captures the sense of unease that many of us have about elevating politics to the apex of human activity; such an elevation is said to smack of elitism, and of neglect for other realms of human endeavor. Balkin complains that "[i]f we constitutionally value art as the servant or victim of politics, we risk skewing our estimation of it. It becomes art not for art's sake, but forpolicy's".34 The second hom of the dilemma (that is, lack of fit between the rationale and the breadth of protection) can only be remedied by watering down the sense of ''the political" to a point at which virtually any worthy activity is connected (albeit indirectly) with the fonnation of one's mind, character and preferences, which will in tum find its remote articulation in the voting decision. In such a case, there is a danger that we are arguing backwards: from the assessment of a domain of expression as worthy of protection to a characterization of the domain as "political", rather than the other way round, as the theory postulates. Others have observed that self-government is not necessarily linked to the principle of strong protection of freedom of speech; indeed, one may perhaps argue for restricting free speech on the basis of self-government. As Lee Bollinger observes: if the people themselves, acting after full and open discussion, decide in accordance with democratic procedures that some speech will no longer be tolerated, then it is not 'the government' that is depriving 'us', the citizens, of our freedom to choose but we as citizens deciding what the rules of conduct within the community will be.3 s

Such a self-governing restriction on future exercises of self-government (which may include attempts at speech-acts now declared intolerable) may seem paradoxical, but it is no more paradoxical than any act of self-constraint,

33. See R. George Wright, "A Rationale from J. S. Mill for the Free Speech Clause", Supreme Court Review (1985): 149-178, p. 152; Laurence H. Tribe, American Constitutional Law (Mineola, N.Y.: The Foundation Press, 1988, 2nd ed.), p. 786; Steven H. Shiffrin, The First Amendment, Democracy, and Romance (Princeton: Princeton University Press, 1990). 34. J.M. Balkin, "Populism and Progressivism as Constitutional Categories (Book Review)", Yale Law Journal 104 (1995): 1935-90, p. 1962, footnote omitted. 35. Bollinger, The Tolerant Society, p. 50, emphases in original.

JUSTIFICATIONS OF FREEDOM OF SPEECH

23

individual or collective. The paradox of pre-commitment may be a theoretically difficult phenomenon to conceptualize but, without entering into this vexed problem here, at least this much is certain: just as it would be improper to say that an individual can never exercise her liberty to constrain her future actions (or increase the costs of her future actions), so it would be an intolerable. restriction on self-government to say that a self-governing community can never, deliberately and purposefully, restrain the scope of its future actions (or heighten the costs of its future actions). Indeed, to postulate that no such collective self-restraint is ever compatible with true self-government would strike at the very idea of constitutionalism. And it is not clear why, if the very principle of self-restraint is accepted as compatible with self-government, selfrestraint consisting in limiting the scope of acceptable speech must be immune from this characterization. As Schauer observes: Any distinct restraint on majority power, such as a principle of freedom of speech, is by its nature anti-democratic, anti-majoritarian. If this were not the case, then the majority would be no more restrained in dealing with speech than in dealing with any other form of conduct, and free speech would be little more than a platitude.36

In the end, we may (and we often do) think that restraints on speech are special, and therefore call for a particularly stringent scrutiny, but a mere invocation of the principle of self-government is insufficient to explain why. Against this last conclusion, one can perhaps claim that a way to reconcile the principle of free speech with the idea of self-government is by seeing this particular "collective self-restraint" as a particular solution to a collectiveaction problem, that is, a case of Prisoner's Dilemma in which everyone will be better-off as a result of the coordinated restraint of certain categories of speech. If we collectively agree that some acts of speech are detrimental to all of us (such as telling strategic secrets to our foreign enemies), but few of us are prepared to suffer self-restraint unless assured that others will do so, then a way of achieving an optimal social outcome is to guarantee, with the force of legal coercion, that each will pay a heavy price for not complying with the collectively agreed-upon result. And if we all consider this outcome as superior, as the argument goes, then we have no grounds for complaining about the violation of our freedom to speak.

36. Frederick Schauer, Free Speech: A Philosophical Enquiry (Cambridge: Cambridge University Press, 1982), pp. 40-41, footnote omitted.

24

CHAPTER 1

This standard explanation for self-restraint is convincing as a rationale for collective self-restraint but it is inadequate to reconcile the principle of free speech with the notion of self-government. To do so, the collective-action explanation would have to demonstrate that a strategy of foreclosing certain options (such as prohibiting certain acts of speech) is not damaging to the freedom of speech of individual actors in a society. But to say that it is beneficial to them, in the sense of attaining a public good which would be unavailable in the absence of mutual abstinence from certain acts of speech, is not equivalent to saying that it is consistent with the free speech principle. Here is the point of disanalogy between individual self-restraint and collective selfrestraint. Individual self-restraint may be shown to be both optimal to an individual and consistent with her liberty because individual freedom must also include freedom to foreclose certain future options available to oneself. But "collective self-restraint" is different in that the identity of those who do the restraining is not identical to those who are subject to the restraint, and so "collective self-restraint" is never truly self-restraint. There may always be someone who disagrees with the ranking of values in which the particular solution to the Prisoner's Dilemma is grounded, and who believes that the price to be paid for the collective result is not justified by the outcome. By imposing an optimal solution, as ascertained by a majority, a "[g]ovemment may indeed overcome a collective-action problem, but it may overcome the wrong one for many of us". 37 For those whose preferences do not coincide with majority preferences (which can be implemented only by mutually assured compliance), the restraint is not an expression of their autonomous choice. Even if, under the principle of majority rule, they have to obey it, it becomes a violation of their freedom of speech, rather than an articulation thereof. As an instance of "democratic paradox" where (alleged) restrictions on democracy are democratically enacted, consider the question of restrictions on political advertising in election campaigns. Such restrictions are commonplace in a number of countries, otherwise considered democratic and free (including Germany, Spain and France). Is this reputation of democratic freedom unfounded, and is the practice of banning political ads at certain periods an infringement of liberty? This was the view of the Australian High Court which, in 1992, struck down a law which attempted to prohibit the publication of political ads during an election period, and to allocate, at the same time, "free

37. Russell Hardin, "Public Choice versus Democracy", in Majorities and Minorities: Nomos XXXII ed. by J. W. Chapman and A. Wertheimer (New York: New York University Press, 1990), p. 194.

JUSTIFICATIONS OF FREEDOM OF SPEECH

2S

time" for eligible political broadcasts and for policy launches by recognized political parties.38 In its judgment, the Australian court argued that, in a representative democracy, public participation in political discussion is a central element of the political process. Bans on political ads, Chief Justice Mason thought, impair the "freedoms previously enjoyed by citizens to discuss public and political affairs".39 Since freedom of communication is an indispensable element of representative government, restrictions on the former undermine the latter. As Mason claimed, "[t]he raison d'etre of freedom of communication in relation to public affairs and political discussion is to enhance the political process ... thus making representative government efficacious".40 But if the raison d'etre of freedom of communication is to respect and enhance democratic political process, and if the arguably democratic process (untainted by the restrictions under consideration) produced a particular law, wouldn't respect for the democratic process dictate that we respect its outcome as well (reserving, of course, a right to try to repeal the law by those same democratic means that brought it about)? If concern for democracy is the basis of the rationale for freedom of expression, shouldn't it extend to the output of democracy? The question is all the more valid since the view that bans on political advertisements adversely affect the principle of freedom of political discussion is controversial in itself. Some of its critics are as adamant as the Australian Court's majority in celebrating the importance of freedom of communication, but they differ in their judgment about what specific legal regime concerning paid political ads is more conducive to an overall freedom of political communication. Tom Campbell, for example, claims that the Court's conclusion about the negative effect of prohibiting political ads is only rendered possible because "the Court simply discounts the corrupting influence of money in the election process, a corruption which does affect the equal competition of ideas which could be regarded as what freedom of communication is meant to be all about".41 So the line is drawn not between those who put freedom of speech at the apex of their political values and those who would restrict it for the sake of other values, but between two conflicting

38. Australian Capital Television v. Commonwealth, (1992) 66 AUR 695. 39. Id. at 699. 40. Id. at 706. 41. Tom D. Campbell, "Democracy, Human Rights, and Positive Law", Sydney Law Review 16 (1994): 195-213, p. 202.

26

CHAPTER 1

interpretations of what is required in order to maximize freedom of communication overall. The merits of the controversy involve a number of complex and contextspecific judgments about the specific operation of such laws. Both the Court and its critics may have good prima facie cases to argue for a causal connection between the specific regulation in question and the consequences which they depict as damaging to a system of freedom of speech (as the Court claims), or as salutary for such a system (as the critics claim). No doubt Mason has a point when he suggests that the combination of a ban on paid ads with the protection of news and current affairs programs could result in a situation where some politicians would enjoy invitations by "the powerful interests which control and conduct the electronic media", while others would have no opportunity to reply. And no doubt he has a point when he complains that the "replacement regime", that is, a system of free time for recognized political parties, is "weighted in favour of the established political parties represented in the legislature immediately before the election" and "discriminates against new and independent candidates".42 But then, critics of the decision also raise points which are prima facie reasonably related to a concern for genuine freedom of expression. As Andrew Fraser complains, "the court never considered whether, by protecting the communicative freedom of the wealthy few, it might diminish the civic freedoms available to the many" - and those "civic freedoms" include "a civic right to freedom of communication".43 In this complaint, he is accompanied by Tom Campbell who notes that the Court perceived freedom of communication in a purely negative sense, as "the absence of legal prohibitions on any form of communication", and did not contemplate that freedom of communication may have a positive meaning, "in the sense of correlating with the duty of government (or other) to take positive steps to further communication by, for instance, the provision of free time on television".44 It is impossible to adjudicate between these conflicting claims in abstracto, and it is unnecessary to do so for our present purposes. What concerns us are not the merits of this particular controversy but rather that it provides a good illustration of a democratic paradox: can democracy enact (what some would consider) abridgments of democratic process? And a converse aspect of this question is: are we justified in using non-democratic means (such as judicial

42. Australian Capital Television at 706. 43. Andrew Fraser, "False Hopes: Implied Rights and Popular Sovereignty in the Australian Constitution", Sydney Law Review 16 (1994): 213-28, p. 225. 44. Campbell, "Democracy, Human Rights", p. 208.

mSTIFICATIONS OF FREEDOM OF SPEECH

27

review, which is by its very nature non-majoritarian) to restore the democracy into a system? Can a non-democratic body constrain a democratic one for the sake of freedom of communication grounded in concern for the democratic process? The suggestion implied by this question would be damaging to the theory only if we adopted an unrestrained majoritarian view as the underlying conception of a democratic polity. On this view, democratically adopted limits upon free political speech would have to be respected and observed without dissent. And yet, the system of constitutional democracy, as it is generally known and celebrated, does not presuppose an unrestrained majority rule. Quite apart from all other restraints upon majority rule, which I shall not discuss here, at least this much seems obvious: the principle of democratic majority rule presupposes the free formation and expression of societal preferences and desires, and so it cannot condone restrictions on the channels of political communication. John Hart Ely defmes the central purposes of the free-speech clause of the First Amendment as "to help make our governmental processes work, to ensure the open and informed discussion of political issues, and to check our government when it gets out ofbounds".45 This is the theory which underlies the famous fourth footnote of the Carolene Products46 decision: laws which tend to restrict the operation of democracy cannot be subject to an ordinary, democratic, majoritarian, deferential-to-Iegislator review, but must be subject to a stricter extra-legislative review. In the absence of this (or a similar) theory, even if we adopted the democracy-based argument about freedom of speech, we still would not have solved the closer-scrutiny problem. That is, we still would not have solved the problem of identifying the grounds for affording extra weight to the free speech principle, compared to a general standard for state interference with individual liberty. It is, of course, true that free expression and communication are necessary for the healthy functioning of democracy, self-government and so on. But that is not the point of our inquiry; the point is to find out why this 45. John Hart Ely, Democracy and Distrust (Cambridge, Mass.: Harvard University Press, 1980), pp.93-94. 46. United States v. Carolene Products, 304 U.S. 144, 152 n. 4. The relevant passage reads: "It is unnecessary to consider now whether legislation which restricts those political processes which can ordinarily be expected to bring about repeal of undesirable legislation, is to be subjected to more exacting judicial scrutiny ... than are most other types of legislation". For a discussion, see Bruce A. Ackerman, "Beyond Carolene Products", Harvard Law Review 98 (1985): 713-46; J.M. Balkin, "The Footnote", Northwestern University Law Review 83 (1989): 275-320; Ely, passim; Wojciech Sadurski, "Judicial Protection of Minorities: The Lessons of Footnote Four", Anglo-American Law Review 17 (1988): 163-82.

28

CHAPTER 1

particular rationale would warrant the insulation of free speech from the ordinary operation of the harm calculus. In order to consider this issue, it is not enough to accept an uncontroversial thesis about the relevance of free speech to the democratic process; we must determine what it is about the democratic process that justifies subjecting free speech to a closer than usual scrutiny. Consider what "strict scrutiny" is about: it is really about a suspicion that politicians and legislators will overstate the degree of harm produced by a given activity (in this case: by exercises of free expression) and/or that they will undervalue the harm inflicted by the restriction. By subjecting a given restriction to a higher than usual standard of scrutiny, we express our mistrust in the political judgment which affects the harm calculus. We must, therefore, have a theory about why such mistrust is more justified in one area than in the other. Why, for instance, is the law in a democratic community more likely to distort the harm calculus in the field of public concerns than in the area, say, of private defamation? Roughly speaking, there are two plausible types of answers to this question and these yield two types of theories of strict scrutiny of speech related to public issues. One theory refers to legislative motives, and the other refers to the risk of legislative miscalculation. The motive-related theory stems from the observation that lawmakers tend to be self-interested and, hence, have distorted perceptions of what constitutes harm to be averted by legislative action. They will, understandably enough, be likely to view strong criticism of their own actions with dislike or suspicion, and they will perceive it as harmful to the public good. In reasoning about the legal restrictions that they propose, they will tend to neglect the incidences of overinclusion of a proposed legal restriction. Overinclusion of the law regulating freedom of speech means that a statute "comprehends a substantial range of applications to activity protected by the First Amendment, in addition to the unprotected activities it legitimately prohibits".47 At the same time, legislators will tend to be excessively receptive to arguments of underinclusion, the arguments which point to a need to expand a given restriction upon allegedly similar cases. The argument is that legislators win tend to overvalue harm that affects their electoral chances, and will under-value possible harm produced by their own political activities. In order to compensate for this distortion motivated by self-interest, it is necessary to enhance the importance of free speech directed at public issues, and so to insulate it from the ordinary operation of the harm calculus.

47. Turchick v. United States, 561 F.2d 719, 721 (8th Cir. 1977), footnote omitted.

JUSTIFICATIONS OF FREEDOM OF SPEECH

29

It is important to note that the motive-based argument does not rely upon a

simple notion that parliamentarians are less likely to articulate constitutional principles (including the principle of freedom of speech) in their decisionmaking than constitutional judges are. In conventional arguments for the power of judicial review of legislative acts, it is often assumed that, in the division of labor between legislators and judges, the former tend to consider a number of factors, including their own self interest, the passions of the day, and the pressure of interest groups. Judges, on the other hand, attempt to articulate the true meaning of law, and nothing else. This is part of judicial mythology all over the world, and its widespread popularity is, in itself, a partial (if paradoxical) rebuttal of the judicial claim of disinterestedness: the power of such a bizarre myth surely suggests that much judicial rhetoric is self serving. Both judges and legislators, in the ideal division of power, attempt to articulate the meaning of constitutional rules as best they can; and both judges and legislators, in the real world, are prone to self interest, myopia, and external pressures. There is no general reason to believe that, for example, in the clash of views about political ads, the parliamentarians who enacted a law banning political advertising were less concerned about the particular articulation of the principle of freedom of speech than were the judges who struck down the law. The constitution (including the constitutional principle of freedom of speech) binds all authorities, and it is odd to suggest that legislators, as a rule, are disingenuous in their fidelity to the constitution while judges, as a rule, are not. If we thought that the majority is inherently unable to respect and honour the right to freedom of speech, and that is why we need a countermajoritarian body to ensure the legislative respect for constitutional rights, then we would be incapable of understanding how constitution-making (including the constitutional proclamation of freedom of speech) is possible at all. After all, it is the majority which ultimately makes decisions about the constitution - a qualified majority, and a majority acting in a special way, but a majority nonetheless. And if we never trusted the majority to be able to consider, in good faith, the legitimate interests of the minority, then we could never have a genuine bill of rights in the first place. But if there are some circumstances in which we can trust the majority to determine our rights (partly because we have no other choice), then it opens the way to trusting the majority in other circumstances as well, provided that these circumstances significantly resemble the circumstances which supported the trust in the first place (that is, the circumstances of constitution-making). To be sure, there are important differences between constitution-making and ordinary lawmaking, but the differences are only of degree. To draw a sharp contrast between the majority deliberating on the constitution and the majority

30

CHAPTER 1

deliberating on other statutes, including those which would restrict the constitutional rules, would be in essence to rely on the fiction that the same group can act, in different circumstances, on the basis of totally different motives. While it may sometimes apply to an individual agent, it is much less plausible with regard to the community. So the motive-based theory must make a narrower claim: it must demonstrate that, due to a specific institutional design, legislators' typical motivations will incline them to overestimate the harm attacked by the regulation, and to underestimate the costs of the proposed regulation, including the incidence of over- and under-inclusion. The theory must also show why the institutional design of the judiciary makes these distortions less likely to occur in the minds of judges. Such a judgment will hinge on a great number of variables, and on their relevance to an institutional ability not to be motivated by considerations of self-interest, myopia and prejudice. These variables include, among other things, matters such as procedures of selection and the recruitment of members of a given body, the conditions of job security of the decision makers, the flexibility in determining one's agenda, the access to information and empirical studies on matters affected by a decision, requirements for giving reasons for one's decisions and defending them against critics, patterns of responsibility for unpopular decisions, et cetera. But it is question-begging to declare a priori that the judiciary is better qualified to determine the best interpretation of general, textually indeterminate provisions of constitutional bills of rights, such as the right to' freedom of speech. In tum, the risk-oriented theory borrows from Carolene Products the insight that legislative decisions, which may distort the future operation of democratic lawmaking, must be remedied through extra-legislative processes, and without the usual democracy-based deference to legislative judgment. Legislative restrictions upon freedom of speech, even if properly expressing current societal preferences, contain a high risk of distortion of preference-formation and preference-expression in the future. In particular, they will make it more difficult for future lawmakers to remove these restrictions in accordance with changed public preferences because the restrictions will continually distort the overall picture of preferences. The process of democratic lawmaking must, therefore, be subjected to greater vigilance when the outcome of this process threatens to alter the process itself. Furthermore, restrictions upon free speech not only tend to distort the process itself but also to distort the input of the process: lawmakers obtain a distorted picture about the actual distribution of various preferences and desires in a society. Hence, in a variety of ways, restrictions on public-concern speech pose a higher risk for the overall democratic process than many other restrictions do.

JUSTIFICATIONS OF FREEDOM OF SPEECH

31

This is a crude summary of a theory which might justify stricter scrutiny of restrictions of speech in areas of public concern. One clear advantage of this theory is that it does not make any express appeal to individual rights to free expression: it does not say, for example, that democratic decision-making must be subject to limitations related to individual rights. Such an appeal would render the theory circular because the scope of rights to free expression is precisely the issue at hand. The upshot is that restrictions of speech trigger a higher standard of scrutiny of the speech-harm relationship if the speech at issue is related to public and, in particular, political concerns. I do not wish to offer a definition of what makes a given subject-matter "removed from politics,048 and what makes it closer to political and public concerns. I assume, however, that such distinctions can be made. They have been made, using formulas such as speech that "influences social relations and politics on a grand scale", or is part of a "debate about public affairs", or makes a "contribution to the public debate", or concerns ideas which may "bring about ... political and social changes".49 It is, of course, an altogether different question to consider who is best capable of ascertaining whether a particular expression belongs to that category of speech and, in particular, it is far from clear whether courts are competent to judge what information "is relevant to self-government". 50 But this issue of institutional competence is independent of the issue of the conceptual characterization of a given expression as belonging to the category of speech pertaining to public issues.

4

Tolerance

The most original feature of an influential book by Lee Bollinger The Tolerant Society is that it shifts the focus of arguments about freedom of speech away from the speakers and towards the audience. In Bollinger's view, it is not so much for the sake of protecting speakers but rather in order to promote the right attitudes of tolerance among the audience that speech-acts call for higher protection. "[T]he tolerance principle ... is intended and designed to perform a self-reformation function for the general community and not . . . to offer a shield of protection either for the majority against the government or for

48. American Booksellers Association v. Hudnut, 771 F.2d 323, 331 (7th Cir. 1985). 49. See, respectively, id. p. 331; Hustler Magazine v. Falwell, 485 U.s. 46, 53 (1988); New York Times v. Sullivan, 376 U.S. 254,279 n. 19 (\964); Roth v. United States, 354 U.S. 476,484 (1957). 50. Rosenbloom v. Metromedia, Inc., 403 U.S. 29,79 (1971) (MarshalJ, J., dissenting).

31

CHAPTERl

minorities against unfair treatment at the hands of the majority".51 Regardless of the lack of any social value in a given speech-act, promoting the attitude of tolerance towards the speaker helps to create "a general intellectual character',52 and enhances ''the capacity for general tolerance".53 Consequently, justified cases of non-protected speech, such as "fighting words", libel, or obscenity, reflect those situations where self-restraint towards speech acts cannot be reasonably expected.54 This last issue is a good starting point for critical reflection about Bollinger's theory. As it is, it fails to convince me that the underlying purpose of promoting tolerance can provide justification both for permitting "extremist speech" (which is the central focus of his book) and for refusing protection to certain categories of speech that are outside First Amendment protection. There is a distinct air of arbitrary rationalization in Bollinger's argument: taking the Supreme Court's current doctrine as a moral yardstick, he suggests that the protection of much extremist speech can be justified by an appeal to the virtues of tolerance, while a refusal to grant higher protection indicates that it would be too much to expect people to tolerate some types of speech. One feels that this sort of argument might be used to justify any scope for protected speech, narrow or wide. Perhaps this is somewhat unfair to Bollinger, since he attempts, at times, to explain why some instances of speech, such as extremist political speech, lend themselves to improper intolerant responses more than other categories, such as obscenity or libel. But at this stage, his explanations become anecdotal and of very limited applicability. For instance, this is exemplified when he asserts that a significant amount of outcry against anti-Semitic marches and speeches in the United States is motivated by an attempt by the Jews to appease their own sense of guilt.55 More generally, Bollinger seems to merge two ideas: a plausible (but not very original) one and a radical (but implausible) one. A radical, but hardly plausible, idea is that promoting right attitudes among the audience, regardless of the protection of speech and/or of speaker, warrants a need for the strong protection of some categories of speech, even if they are harmful under an ordinary calculus of bum. However, this sounds implausible. Under this theory, the law prevents A from interfering with B's speech in order to protect something important about A. This sounds like a moralized

51. Bollinger, The Tolerant Society, p. 134. 52. Id., p. 124. 53. Id., p. 182. 54. Id., pp. 181-86. 55. Id., pp. 129-130,274-75 n. 17.

JUSTIFICATIONS OF FREEDOM OF SPEECH

33

paternalism: restricting individual liberty for the agent's own good, where his or her good is defined in terms of promoting right virtues and commendable character. I do not think that such a reason for governmental action passes the muster of a liberal theory both because it is paternalistic and moralistic. It is, first, paternalistic, because the good which it is alleged to protect is the good of. those who are constrained by the policy. And it is moralistic because the nature of the "harm" it is alleged to addressis "moral harm", which is not a genuine harm to one's interests but rather a "harm" in not having the virtues which are thought to be morally worth having.56 And, as Joel Feinberg argues, "if legal coercion designed to prevent self-inflicted moral harm is an indefensible violation of personal sovereignty, then all the more so still is legal coercion to promote the actor's own moral goocf,.57 In the specific setting of a neo-Nazi march in Skokie, which is the central example used by Bollinger, the results are almost perverse. As Pierre Schlag puts it, "can you imagine telling Jewish survivors of the concentration camps that the major (the only?) reason the Nazis must be allowed to march is to enable the Jews and the rest of the community to learn the virtues oftolerance?".58 But there is a more plausible (even if not very original) idea in Bollinger's theory, and it is to his credit that he points our attention in this direction. The plausible idea is that even if much offensive or extremist speech is without any social value, restrictions are often demanded for the wrong reasons, and those wrong reasons taint those regulations of speech. Bollinger claims that, in cases of public responses to extremist speech, "if one looks not at the speech but at the motivation behind the restrictions, one may properly conclude that the restrictions were imposed for bad reasons". 59 Hence, arguments about the level of scrutiny should not rest upon the nature of speech (the restriction of which is to be scrutinized), but rather upon the motives for a governmental restriction. The underlying theory is that the legitimacy of governmental prohibitions has to do with the nature of the motives for these prohibitions, rather than with the nature of the prohibited conduct itself. It has been argued, for instance, that a liberal theory of limited governmental powers, properly understood, means that

56. See Joel Feinberg, Harm to Others: The Moral Limits of the Criminal Law, vol. I (New York: Oxford University Press, 1984), pp. 65-70. 57. Joel Feinberg, Harmless Wrongdoing: The Moral Limits of the Criminal Law, vol. 4 (New York: Oxford University Press, 1988), p. 299, emphases in original. 58. See Pierre Schlag, "Freedom of Speech as Therapy (Book Review)", UCLA Law Review 34 (1986): 265-83, p. 279, footnote omitted. 59. Boninger, The Tolerant Society, p. 126.

34

CHAPTER 1

the government may never act for certain reasons, which are deemed improper or unconstitutional. Such an argument has been made both in a moral theory and in the theory of constitutional law. In moral theory, a good example is provided by C.L. Ten's interpretation of Mill's "harm to others" principle. In Ten's reading, Mill's ban on state interference with self-regarding acts means that "[t]here are certain reasons for intervention in the conduct of individuals which must always be ruled out as irrelevant" so that "[e]ven when intervention iS'justified in a particular case it is on the basis of certain reasons rather than others".60 More directly on our subject-matter, Scanlon suggested that "an important component of our intuitions about freedom of expression has to do not with the illegitimacy of certain restrictions but with the illegitimacy of certain justifications for restrictions".61 In constitutional law, it has been noted with approval that the US Supreme Court often considers an "illicit reason" in the legislature's deliberation over a law as "impermissibly skewing" the calculus of costs and benefits, and thus, tainting the restriction. 62 As Geoffrey Stone says, "if an improper motivation played a substantial role in the government's decision to restrict expression, the restriction must be invalidated even if alternative, proper justifications are available".63 The general implication is that in order to judge the propriety of a particular prohibition it is neither necessary nor sufficient to look at the external outcome of the prohibition, that is, at the class of actions which are prohibited as compared to the class of actions which are pennitted. Rather, we must inquire into the nature of the motives which are behind a given restriction; consequently, prohibition of one and the same action may be legitimate or illegitimate, depending on the type of motives at work. The most influential version of the motive-related strategy refers to a distinction between viewpoint-based, as opposed to viewpoint-neutral, regulations of freedom of expression. The connection between motive-based scrutiny and the hostility toward regulations which are based on the viewpoint of a speaker is clear: "[0]ne reason for the hostility toward viewpoint discrimination is that the government may have a motive for excising a particular point of view that is unrelated to the public welfare; the government

60. C.L. Ten, Mill on Liberty (Oxford: Clarendon Press, 1980) p. 40.

61. Scanlon, "A Theory of Freedom of Expression", p. 209. 62. Tribe, pp. 821-22. 63. Geoffrey R. Stone, "Content Regulation and the First Amendment", William and Mary Law Review 15 (1983): 189-252, p. 230, footnote omitted.

JUSTIFICATIONS OF FREEDOM OF SPEECH

35

may excise a point of view simply because it disagrees with it.,,64 As Sunstein concludes: ''The special hostility directed at viewpoint restrictions ... is thus designed to 'flush out' impermissible bases for government action".65 The general principle might therefore be formulated as follows: restrictions that express a viewpoint on a given issue that is preferred by the government (or conversely, that is disapproved by the government) should be subjected to very strict scrutiny. By implication, restrictions that are viewpoint-neutral, and which do not discriminate among different viewpoints, may be subject to lenient scrutiny because there is little ground for suspicion that they result from improper governmental motives. But is a distinction between viewpoint-based and viewpoint-neutral regulations (especially, those viewpoint-neutral regulations which are not neutral with regard to a subject-matter of speech) clear and noncontroversial? And, even if it is, is the link between viewpoint discrimination and the likelihood of impermissible motives for restrictions as manifest as the theory of viewpoint neutrality suggests? It is this theory, and its underlying rationales, that we will discuss in detail in Chapter 5.

64. Sunstein, p. 610. 65. Id., p. 611, footnote omitted.

CHAPTER 2

SPEECH AND HARM

1

Levels of Scrutiny

In what became perhaps the most oft-quoted sentence in the First Amendment jurisprudence of the United States, Justice Stevens observed that "few of us would march our sons and daughters off to war to preserve the citizen's right to see 'Specified Sexual Activities' exhibited in the theaters of our choice".1 Much less celebrated is a reply by Archibald Cox: Few of us would march our sons or daughters off to war to preserve the citizen's right to see pictures of American Nazis marching in uniform in Skokie, Illinois, or to hear advocacy of Stalinist Communism, or to read advertisements stating the price of prescription drugs. The test is both unreasoning and insufficient. 2

Cox is right, of course, but he may, in turn, be accused of taking a rhetorical figure for a strict legal argument. The point conveyed by Stevens' proposition is both substantial and intuitively plausible: freedom of speech, though of great value, is not absolute, and some domains of freedom of speech are more valuable than others. To insist that all speech, regardless of its nature, should presumptively be equally protected, would be both patently absurd and inimical to the freedom of really valuable speech. Confronted with a demand for a uniform standard of protection of all speech, law would inevitably respond by lowering the degree of protection across the board. A plea for a uniform standard of protection sounds counterintuitive: there are words that hurt, and that produce harm to other people and to entire communities. The damage produced by words may be very high; for example, public statements that express racial hate or contempt for an entire group of people hurt their victims more than many other unpleasant words. It is one thing to say that offensiveness itself is an insufficient reason to punish the speaker, and that the regime ofliberty demands that we put up with expressions of unwelcome ideas, unpopular thoughts and offensive views. As the United States Supreme Court said, "the fact that society may find speech offensive is not a sufficient reason for suppressing it. Indeed, if it is the speaker's opinion

I. 2.

Young v. American Mini Theatres, Inc., 427 U.S. 50,70 (1976). Archibald Cox, "The Supreme Court, 1979 Term - Foreword: Freedom of Expression in Burger Court", Harvard Law Review 94 (1980): 1-73, p. 29.

38

CHAPTER 2

that gives offense, that consequence is a reason for according it constitutional protection".3 But it does not follow that gratuitous harm, of whatever gravity, must always be tolerated merely because it results from words rather than from actions. If there is something undoubtedly wrong under most moral systems, it is deliberate infliction of pain for pain's sake, and the fact that it is done through the medium of words should not automatically immunize it from legal interference. However, one clear implication of talking about a ''right'' to free speech is that it insulates speech from the ordinary operation of the harm principle. By the "ordinary operation" of the harm principle I mean the use of a calculus of benefits (in terms of harm avoided) and costs (in terms of speech restrained). Under the calculus, the government would be obliged to prohibit a speech act whenever it would be reasonable to expect that any resulting harm outweighed the harm of restraining people from speaking. The regime of ''rights'', it is generally thought, insulates a given sphere of activity from the use of such a calculus; the notion of ''rights'' would be redundant if their exercise were conditional upon a case-by-case confirmation that the benefits resulting from an individual action outweighed its costs. This does not mean that a right to act in a particular way makes any interference with such action inappropriate, but only that the threshold of justification must be substantially raised. Net benefits resulting from restriction will not always justify the restriction if the notion of "rights" is to retain its significance and not merely become a substitute calculus of costs and benefits of governmental action. Whilst this much may be obvious, the question is how to translate it into a language suited to legal argument. The framework offered here is in terms (familiar in American constitutionalism) of a level of scrutiny of a given regulation. The idea is to imagine yourself as a judge whose task is to review a challenged regulation from the point of view of its conformity with a constitutional right (in this case, the right of free expression). A preliminary question that one must answer concerns the level of scrutiny: once we have agreed that rights are not absolute, we accept that their exercise can sometimes be restrained by an appeal to a higher social good (in particular, harms avoided by such a restraint). But how great must a good be in order to warrant a restriction of the right, and how closely must the restriction be related to a likely achievement of this good? At one end of the spectrum of possible answers (lenient scrutiny) we may decide that any net surplus of benefits, however minimal, will justify the restriction of a harmful action, and that even the slightest likelihood of a positive relationship between the restriction and the benefit will convince us about the legitimacy of the restriction. At the opposite end of the spectrum (strict scrutiny) we may require that a social good of very 3.

FCC v. Pacifica Foundation, 438 U.S. 726, 745 (1978).

SPEECH AND HARM

39

great importance must be demonstrated in order to warrant a restriction, and that, for a restriction to be legitimate, a very close link between the restriction and the achievement of this good must be shown. In this latter case, any reasonable uncertainty of achieving the purported aim of the restriction, or the incidence of over- and under-inclusion (so that not all the harmful expressions will be captured by the restrictions, and some non-harmful expressions will be) will defeat the challenged restriction. For an illustration of the language of "strict scrutiny" consider these words of Justice Holmes from his famous dissent which became part of the canon of liberal free-speech jurisprudence. All speech acts, Holmes argued, should be protected unless they "so imminently threaten immediate interference with the lawful and pressing purposes of the law that an immediate check is required to save the country".4 The italicized words indicate both the immediacy and the gravity of the danger that the restriction aims to arrest. More recently, the Supreme Court of the United States announced that, in order to pass constitutional muster, a governmental restriction must be a "a precisely drawn means of serving a compelling state interest,,;5 or that the regulators must choose "the least restrictive means of achieving some compelling state interest".6 A similar idea was expressed by the Canadian Supreme Court when it established that a governmental measure which infringes a constitutional right is only justifiable when it impairs the right in question "as little as possible", in addition to being rationally connected to an "objective of pressing and substantial concern in a free and democratic society".7 Hence, with regard to the legal suppression of hate speech, it must be shown that "the means embodied in hate propaganda legislation should be carefully tailored so as to minimise impairment of the freedom of expression".8 A decision concerning the choice of a point on the scale at which to locate our test, between the strict-scrutiny and the lenient-scrutiny ends of the spectrum, will affect the results of the test. Naturally, the closer we approach the strict-scrutiny end, the less likely the regulation is to survive. In contrast, to choose lenient scrutiny displays a much higher deference to the lawmakers, and increases the likelihood that the restriction will survive. But we must have a rational method of determining the proper level of scrutiny, and (as will be

4. 5. 6. 7. 8.

Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes J, dissenting), emphases added. Consolidated Edison Co. v. Public Servo Comm'n., 447 U.S. 530, 540 (1980). Thomas V. Review Bd., 450 U.S. 707, 718 (1981). The formula quoted in the text concerned the Free Exercise of Religion Clause. R. V. Oakes, [1986] 1 S.C.R. 103, 138-39. R. V. Keegstra, (1991) 61 C.C.C. 3d 1,55.

40

CHAPTER 2

explained later) such a method should, itself, be independent of our views about the gravity ofharm of a given category of speech. While a framework of various levels of scrutiny has been elaborated in the context of a judicial review of laws under a constitutional bill of rights, the underlying idea is of universal application. As long as we presuppose that there is a right to free speech in a given society (whatever the sources of this right may be, constitutional or otherwise), we can analyse the proposed restrictions on that right in the language of different levels of scrutiny. It really does not matter who does the subjecting of a given restriction to critical scrutiny: a constitutional judge, a legislator, a citizen, or an academic. The scheme of different levels of scrutiny is not dependent upon any institutional pattern of judicial review; it is merely a framework to facilitate understanding of what is going on when speech-restrictions are evaluated from the point of view of their capacity to avert harm. At the end of the day, what will matter in our judgment about the rightness of laws which restrict speech, is our view about the nature and degree of harm produced by the speech targeted by these laws. However, it does make a difference whether the speech-harm relationship will be subjected to a lenient, or to a strict, scrutiny. The same degree of harm may defeat the regulation under one test and leave it unaffected under another. The reasons for the use of a given test are related to the nature of the speech in question. Surely there is a difference between a speech act that consists of communicating to another the secret combination of a bank safe, and a speech act that calls for the abolition of the present government. The difference does not lie in the degree of harm: a safe box may contain nothing, and the government under criticism may be as good as they come. If both were subjected to the same level of scrutiny, the disclosure of the bank safe's combination would, in these circumstances, attract a lesser legal sanction than persuasive criticism of the government - a strongly counterintuitive result. As Frederick Schauer says: ''To lie under oath violates the law even if the lie turns out to cause no harm whatsoever in the particular case, a result unexplainable if courtroom testimony under oath were an activity covered by the first amendment".9 If for "the first amendment" we read "the principle of freedom of speech", Schauer's proposition identifies a general problem: why does the 'principle accord differential protection to various categories of speech even before we know the actual gravity of harm produced by those different categories? In other words, can there be any principled distinction which can be drawn among speech acts before the examination of the link between speech and harm, and as a result of which we can lower the usual, presumptive strict scrutiny of restrictions upon speech?

9.

Frederick Schauer, "The Aim and the Target in Free Speech Methodology", Northwestern Law Review 83 (1989): 562-68, p. 563.

SPEECH AND HARM 2

41

Low and High Value Speech

It might seem that the most obvious argument for lowering the level of scrutiny

appeals to the social value of the speech in question. Racial vilification, pornography, offensive epithets and derogatory remarks about respectable persons have a lower social value than the kind of speech that a society has in mind when it entrenches rights to free expression. The fact that not all speech is of equal importance from the point of view of First Amendment values has long been an established doctrine of the United States Supreme Court. A list of "low value" speech categories goes back to the landmark decision Chaplinsky v. New Hampshire lO and, notwithstanding some recent dilutions, II it includes obscenity, libel, express incitement, commercial speech, child pornography and "fighting words". This inferior value of "certain well-defined and narrowly limited classes of speech" was, according to the Chap/insky Court, due to the fact that they "are no essential part of any exposition of ideas, and are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.,,12 As commentators noted, the "low value" speech is not, eo ipso, totally unprotected but, rather, when speech is perceived as having a low value only, the restrictions upon such speech are subject to a balancing of costs and benefits which would be improper in the case of restrictions upon speech of higher value. 13 Some writers take a principled, high ground in rejecting the gradation of speech, by law, into higher and lower values; they consider such a distinction incompatible with the libertarian approach to freedom of speech. Archibald Cox defended a principle "that a court should not attempt to differentiate or allow the state to differentiate the value of particular messages protected by the first amendment".14 Undoubtedly, the idea that courts should not differentiate among the values of different messages can claim support from powerful libertarian arguments: the risks of entrusting an authoritative body with the

10. 315 U.S. 568 (1942). 11. See Laurence H. Tribe, American Constitutional Law (Mineola, N.Y.: The Foundation Press, 1988, 2nd ed.), pp. 850-56. 12. 315 U.S. at 571-72. 13. See Geoffrey R. Stone, "Content Regulation and the First Amendment", William and Mary Law Review 15 (1983): 189-252, pp. 194-195. 14. Cox, p. 28. See also, similarly, Larry Alexander, "Low Value Speech", Northwestern University Law Review 83 (1989): 547-54; Kenneth L. Karst, "Equality as a Central Principle in the First Amendment", University o/Chicago Law Review 43 (1975): 20-68, p. 28; Tribe, pp.940-44.

42

CHAPTER 2

power of assigning different values to different categories of speech, even before they consider the harm of specific speech, are manifest. But a demand for a uniform standard is unrealistic: even in the United States "it is clear that some speech may be regulated not merely because it is harmful but also because the government need not, for that category of expression, meet the ordinary, highly protective standards for regulating speech". 15 More importantly, uniformity of levels of scrutiny could be, ironically, damaging to the general system of freedom of speech. Its adoption may lead to a general lowering of the standard of scrutiny of restrictions of expression across the board. As an example, consider a Harvard Law Review comment about the Supreme Court's 1990 decision on child pornography in Osborne v. Ohio.t 6 Since the majority employed an "intermediate scrutiny", but failed to explicitly classify the speech in question as "low value", the failure may, according to the authors of the Note, "erode the protection of 'core' speech as wen".17 As the Note suggests: "Because Osborne did not openly endorse the 'low-value' approach, lower courts might conclude that child pornography is core ~h and that Osborne is an example of a relaxed strict scrutiny analysis".18 As a result, "[c]ourts could view Osborne as generally relaxing the strict scrutiny requirements that the government have a compelling interest in order to regulate protected speech and that such regulations are the least restrictive means to advance that interest" .19 What kind of argument must be made by a proponent of lenient scrutiny of a restriction on a given category of speech? To make a claim about the low value of a particular utterance may mean either of two things. First, it may mean that the harm produced by such speech is significantly greater than in the case of many other categories of speech subjected, as they are, to a high level of constitutional protection. But remember, the aim of a strategy to lower the level of scrutiny of laws which restrict some types of speech is to justify a standard, deferential test for the measurement of harm produced by a given category of speech. Hence, the reasons for lowering the level of the test must themselves be harm-independent. Otherwise, our appeal to a lower standard of scrutiny would collapse into the harm principle. But if this is all that is meant, the appeal to a lower standard of scrutiny is rendered redundant. We might just as well have directly proceeded to the speech-harm relationship argument especially since, if the harm perceived is really great, a given restriction will probably pass the muster of even a very stringent level of scrutiny. This is not a patently implausible argument; indeed, some writers suggest that we should "treat the

15. Cass R. Sunstein, Democracy and the Problem ofFree Speech (New York: Free Press, 1993), p.124. 16. 495 U.S. 103 (1990). 17. Note, "The Supreme Court, Leading Cases", Harvard Law Review 104 (1990): 129-358, p. 245. 18. Id., footnote omitted. 19. Id., emphasis added.

SPEECH AND HARM

43

realm of messages as an undifferentiated whole and distinguish protected from unprotected speech on the basis of the nature of the harm that is feared and the linkage between the message and the harm".20 Such direct scrutiny of the speech-harm connection renders it unnecessary to differentiate between the relative values of various categories of speech, and the stringency of the test remains uniform for all speech. But the risk is, as observed above, that the uniform test will be less protective than the "strict" one, and much of the speech which we would like to exempt from the routine harm calculus will become endangered. Alternatively, a "low-value" argument may mean that a given category of speech is so remote from the main rationale for freedom of expression that there are no moral or political reasons to accord it the usual freedom-ofexpression protection. As Cass Sunstein says, "any well-functioning system of free expression must ultimately distinguish between different kinds of speech by reference to their centrality to the First Amendment guarantee".21 This test for "low" or "high" value easily escapes the charge of redundancy addressed to the former interpretation of "low value". It also sounds like a serious moral argument: there are some reasons for having freedom of expression, and if these reasons do not apply to the category of speech in question, then there is no justification for treating it on a par with core categories of expression. In order to assess the weight of this argument, one must reflect upon the reasons for having freedom of expression in the first place, and only then test their application to the category of speech in question. It is only once one knows what values are promoted by this freedom that one is able to judge which categories of speech are central and which are marginal, penumbral or remote (hence, of lower value). The previous chapter offered a review of various possible reasons for adopting the principle of freedom of speech; it was suggested that those reasons support special protection for speech on public issues. Such speech can be considered to have a high value from the point of view of the purposes of freedom of expression.

3

Speech-Plus and Symbolic Action

One strategy for supporting a lowered level of scrutiny of speech acts is to detect a high "conduct ingredient" in a given act. Since it is only speech and not the conduct which warrants a high degree of protection under the principle of freedom of expression, "speech-plus" will be entitled to a standard level of 20. Alexander, p. 554. 21. Sunstein, p. 126, footnote omitted, emphasis added.

44

CHAPTER 2

protected behaviour, that is, subject to the nonnal operation of the hann principle. As Thomas Emerson suggested, ''the focus of inquiry must be directed toward ascertaining what is expression, and therefore to be given the protection of expression, and what is action, and thus subject to regulation as such".22 And, in the words of the US Supreme Court, to the extent to which a conduct "involve[s] more than just communication ideas", it should not be "immune from all state regulation".23 But the speech/conduct distinction is ridden with problems, and ultimately may prove to be unhelpful for the purposes of detennining a level of scrutiny of speech restrictions. As I will argue, the distinction is more plausible when viewed as part of the discourse about the reasons given for particular restrictions of speech rather than about the nature of the speech (or conduct) in question. As such, the speech/conduct distinction is too crude to identify those communicative acts which deserve special protection. First, any speech is accompanied by conduct of one sort or another, which may be burdensome to others: for example, talk may be noisy, or the distribution ofleaflets may cause litter. 24 Speech is never "pure"; it always has a "plus" (conduct which makes the speech possible), although the fonns of the ''plus'' are varied. This immediately suggests that the characterization of an act as non-speech may be a result of, not a prerequisite to, our willingness to regulate it. As an illustration, consider a dissent by Justice Douglas to the Supreme Court of the United States decision from which the immediately preceding quotation was taken. In Teamsters, the Court decided that a state could regulate picketing in a manner unaffected by the First Amendment rules because picketing was more than just a communication of ideas. Douglas protested: "where, as here, there is no rioting, no mass picketing, no violence, no disorder, no fisticuffs, no coercion - indeed nothing but speech",25 then full First Amendment protection should be provided. But, strictly speaking, it is untrue that there was "nothing but speech"; there was conduct (picketing). According to Douglas, the conduct did not meet the criteria of hann because there was "no violence, no disorder, no fisticuffs" et cetera, but the fact that the conduct was not hannful (according to Douglas' criteria ofhann) did not render it any less conduct-like, and certainly did not transfonn it into being "nothing but speech". Hannful conduct and pure

22. Thomas I. Emerson, ''Toward a General Theory of the First Amendment", Yale Law Journal 72 (1963): 877-956, p. 917. See also Thomas I. Emerson, The System of Freedom of Expression (New York: Random House, (970), p. 17. 23. Teamsters Local 695 v. Vogt, 354 U.S. 284, 289 (1957). The decision concerned trade union picketing. See also, similarly, Cox v. louiSiana, 379 U.S. 559 (1965) where a statute prohibiting picketing near a courthouse was upheld as a valid regulation of conduct as distinguished from pure speech. 24. See Tribe, p. 827. 25. 354 U.S. at 296 (Douglas, J., dissenting), emphasis added.

SPEECH AND HARM

4S

speech do not exhaust all the possibilities: harmless conduct is a third option. So the real reason for characterizing the regulation in question as targeting the speech was that the target of the regulation (conduct, speech, or whatever) was acceptable to Douglas; and it was not the case (contrary to Douglas' rhetoric) that the regulation was unacceptable to Douglas because it targeted pure speech. Second, there are many forms of conduct which aim primarily, or exclusively, to be communicative. Wearing armbands or uniforms, saluting or refusing to salute a flag, or burning a flag are all kinds of "conduct" but their main aim is symbolic and communicative. It is to convey a message that can be expressed in words, but which (according to a speaker) will be more effective when the vehicle of symbolic action is used. To realize that the conduct is primarily communicative, one can ask oneself whether "[a]n intent to convey a particularized message was present, and [whether] the likelihood was great that the message would be understood by those who viewed it".26 To protect speech but not conduct that is equivalent to speech in the motives of an agent and/or in the meanings it actually conveys to the audience, seems inconsistent, and the Supreme Court of the United States has recognized that conduct may be "sufficiently imbued with elements of communication to fall within the scope of the First and Fourteenth Amendments".27 How far one should go in characterizing various forms of conduct as equivalent to speech involves a subtle judgment of degree, and there is an unfortunate, significant risk of hoax in the argument for protection of expressive conduct. If nude dancing on the stage can be characterized as behaviour protected by the principle of freedom of expression, then why not social dancing or nude bathing? And why not roller skating? The examples are not invented. The United States Court of Appeals for the Seventh Circuit declared nude dancing performed as entertainment to be protected under First Amendment. 28 The Court found that the theme communicated by the dancers to the audience was one of eroticism and sensuality, and refused to draw the line between the communication of ideas and the communication of emotions, or between "high" art and "low" entertainment: "[n]ude barroom dancing, though lacking in artistic value, and expressing ideas and emotions different from those of more mainstream dances, communicates them, to some degree, nonetheless".29 But social dancing and nude bathing did not make it to the

26. 27. 28. 29.

Spence v. Washington, 418 U.s. 405, 410·11 (1974). Id. at 409. Millerv. City of South Bend, 904F.2d 1081 (7thCir.1990). Id. at 1087.

46

CHAPTER 2

higher circles of protected "expressions" in the eyes of two other American appellate courts. The Court of Appeals for the 11th Circuit found that nudity is protected as speech only when combined with some expression which itself is constitutionally protected, and it did not accept the argument that nude sunbathing was expressive in the sense of communicating the philosophy ''that the human body is wholesome and that nudity is not indecent".30 Similarly, the Court of Appeals for the 8th Circuit determined that social, recreational dancing is not expression in the First Amendment sense because it does not convey any kind of message ''unless it be the message that the plaintiffs do not believe that dancing is wrong".31 After all, a message of this kind is implied in any voluntary behaviour by any rational agent, and so the principle of freedom of expression would lose any meaning if such a message were allowed to count as rendering conduct "communicative". It is also for this reason that the Supreme Court of California had "difficulty finding [the element of communication of some ideas or beliefs] in the context of a roller skating rlnk",32 contrary to the petitioners' claim that the operation of a rink was entitled to the First Amendment protection. In the examples just mentioned, one can see the distinction between the conduct which is performed before an audience (nude dancing in a nightclub) and those activities which are undertaken for the sole (or primary) enjoyment of the participants rather than for the audience, or at least not predominantly for the audience. But, admittedly, the line will often be difficult to draw. This is partly because, in a trivial sense, every human action is "communicative" or "expressive" in the sense that it "communicates" to other people something about the agent or "expresses" her preferences, values, character, etc. But not every action is intended to express something about the agent to others, and not every action is seen as aimed at communicating something about the agent to others; as Frederick Schauer rightly observes, "[a]ny coherent formulation of a Free Speech Principle requires communicative intent as well as a perceived message".33 If I go to social dances, as the petitioners' children did in Jarman v. Williams, then my aim is to enjoy myself rather than to communicate my emotions to other people. This is typically so, but it may also happen that my main purpose is to impress my dancing partner and other people, and at that point the distinction between communicative conduct and non-communicative conduct begins to blur. But this only shows that the distinction is often difficult

30. South Florida Free Beaches. Inc. v. City ofMiami, 734 F.ld 608, 609-610 (11th Cir. 1984). 31. Jarman v. Williams, 753 F.ld 76, 78 (8th Cir. 1985) 32. Sunset Amusement Co. v. Board of Police Comrs. of City of Los Angeles, 7 Cal.3d 64, 74, 496 P.ld 840, 845-846 33. Frederick Schauer, Free Speech: A Philosophical Enquiry (Cambridge: Cambridge University Press, 1982). p. 98.

SPEECH AND HARM

47

to make in specific circumstances and that there will be many borderline cases, not that the distinction is without a difference. The argument about the communicative nature of given conduct should be distinguished from a "speech plus" argument. The former, it seems, carries more weight for an increased level of protection, and is therefore less capable of supporting a lower level of scrutiny of restrictions. The difference is this: in the "speech plus" argument it is the "plus" which is problematic from the point of view of a clash with other legitimate values or interests. The "plus" in itself is not equivalent to communication but is a carrier or vehicle of communication. It is therefore a category of conduct which, when not accompanying the speech, can be restrained by an appeal to the routine harm calculus. What adds an extra protection to it is that it is a carrier of speech which enjoys a higher degree of protection than any other action under the routine harm calculus. But this extra protection does not come without a price. By hypothesis, the conduct ingredient endangers some other legitimate interests: otherwise we would not have a problem of claims for restriction. Hence, the proper level of protection should result from balancing the benefits of protecting a "plus" (that is, conduct) as an essential carrier of a speech act, and, on the other hand, the costs that the conduct produces by colliding with legitimate social or individual interests. The case of symbolic conduct is different. Here the whole point of a given conduct is not that it accompanies and carries speech but that it is speech, or rather, it is its equivalent. Symbolic conduct does not produce a conflict with other legitimate social and individual interests, except for those interests which are allegedly violated by the message. When a flag is burned, it is not the fact that someone's piece of fabric is destroyed without the owner's permission, or the fact that a fire hazard is created, or that the burnt fabric left on the street reduces the cleanliness of the city, that provoke the calls for prohibition. It is the fact that some people are deeply offended and upset by the symbolic meaning of an act, which is the true reason for a prohibition, or at least, for the calls for a lower scrutiny of a prohibition. To be sure, other troubling consequences may occur as well (destruction of property, fire hazard, litter on the streets), but they are clearly trivial compared to what constitutes, in the eyes of the critics of legal impunity of flag burning, genuinely troubling aspects of those acts. And, no doubt, one may think of other examples of symbolic conduct which, at times and in various places, have been outlawed (even though the corresponding speech, which expressed an identical meaning, remained legal) and when the occurrence of those secondary, harm-producing effects is minimal or nil: think of wearing certain uniforms or armbands, or saluting certain signs. So it seems that a genuine ground for calls for

48

CHAPTER 2

restrictions of symbolic conduct is that they convey offensive, dangerous, distressing meanings. This is exactly the same ground for calls for restrictions of speech which may convey the same meanings, and which in the case of speech is, in itself, ineffective in lowering the standard of scrutiny of restrictions. But perhaps the distinction between "speech plus" and symbolic conduct should not be pushed too far. If the troubling aspect of a ''plus'' is that, on the one hand, it is a carrier of speech but, on the other hand, it can collide with other legitimate individual and social interests, then a natural question to ask is: to what extent is a ''plus'' an essential carrier of the speech, so that without the carrier, the speech would not be uttered and heard? The more essential the carrier is, and, in consequence, the less separable the ''plus'' is from the speech itself, the stronger is the argument that it should enjoy the same level of protection as the speech it conveys. Posing the problem this way lends itself to a complex balancing of the marginal increase in the "effectiveness" of speech and the marginal increase of non-communicative costs of the enhanced "plus". A political demonstration is a "plus" attached to the substance of the protest the demonstrators want to convey. The substance may be conveyed without a ''plus'' (or with a plus which has a lower collision effect with other interests), such as by publishing a paid advertisement in a newspaper. Let us assume for the sake of argument that, even putting aside the costs of the ad, the effectiveness of the ad is much lower than that of the demonstration. In such a case, to insist upon a different "plus" would greatly detract from the effectiveness of the message. And in this sense, the plus is not easily separable from the speech itself. On the other hand, the organizers of the demonstration may insist that not just any demonstration, but a demonstration through the busy city centre, during peak hours, lasting a very long time et cetera is necessary to convey the message truly efficiently. This shows that there comes a point where the marginal increase of the efficiency of the speech-plus is minimal, and the marginal costs imposed by the "plus" on others are very high. At this point, it is not destructive of the speech itself to refuse the right to a "plus". But a similar refusal at a lower end of the scale ("why don't you express your message in a way which is less disturbing to public order"?) may seem to deny the protection for a "plus" which is inseparable from the message. It would seem that a similar calculus can be made with regard to symbolic speech: "can't you express the message without burning the flag?" In fact, the difference between both these situations is obvious. While speech-plus imposes substantial costs upon the public, and they are produced by a collision of the "plus" with the legitimate interests of other people, in the case of symbolic speech the only non-trivial costs are in terms of harm resulting from the communication of the message in a symbolic, powerful way. The message and the conduct are not separable here, in the way they are separable in the "higher"

SPEECH AND HARM

49

echelons of speech-plus (consider an example of a demonstration in the city center which really disturbs traffic), where the marginal improvement to the effectiveness of speech is minimal, and the marginal cost upon the others is high. In the situation of symbolic speech, the suggestion: "Can't you express your feelings in a civilized way, without burning the flag?" misses the point: the form is inseparable from the message. This is precisely the answer Justice Harlan gave to the same question in. Cohen, a decision of the US Supreme Court concerning the conviction of a man for wearing a jacket in a courthouse with the words "Fuck the draft". Expressions, Harlan said, serve a dual communicative function: not only "to convey . . . ideas capable of relatively precise, detached explication" but also "otherwise inexpressible emotions"; "that emotive function . . . may often be the more important element of the overall message sought to be communicated" and it is often not the case "that we can forbid particular words without also running a substantial risk of suppressing ideas in the process".34It is irrelevant for our purposes that Harlan spoke about "words" while the dissenters in Cohen characterized Cohen's behaviour as "mainly conduct and little speech".35 What matters is that often the form of the message is to such a degree part of the message itself that you cannot separate one from the other. Proponents of prosecuting uncivil expressions believe that they do not suppress the message but they ignore the connection between the message (especially the emotive aspect of it) and the form of expression. R. George Wright, in his criticism of Cohen conceded that "[w]hile not all of us can rise to the level of a Daniel Webster or a Thoreau", it should be constitutionally permissible "to require an even slightly closer approach to the level of a Daniel Webster than Cohen manifests if we are to address the public" and, hence, "to require of all speakers a certain minimum, easily attained level of respect for the feelings of many among the audience".36 The hope that such requirement is neutral with regard to the content of the message is, however, illusory, and betrays a view that the only protected aspect of an expression is an intellectually accessible thought, and not the emotion of the speaker. So when, later in his article, Wright says that "[w]hat we actually learned from Cohen's speech was mostly about Cohen",37 he is correct in one sense, but the conclusion is in fact opposite to the one he intended. Wright's view is

34. Cohen v. California,403 U.S. 15,26 (1971). 35. Id. at 27 (Blackmun, J., dissenting). 36. R. George Wright, "A Rationale from J. S. Mill for the Free Speech Clause", Supreme Court Review (1985): 149-178, p. 173. 37. Id. p. 173.

so

CHAPTER 2

dismissive of Cohen's expression and sceptical of the legal protection accorded to it, while the unintended truth of Wright's argument is that we would miss something important about Cohen's emotions which he wanted to communicate to us if he were "required" to translate his expletive into "I strongly !iisapprove of the draft". Whether we would be worse off for this loss is another matter but this is the sort of judgment that a society, which lays down a principle of freedom of speech, is reluctant to entrust to the authorities. Incidentally, it may seem ironic that Wright borrowed support from John Stuart Mill, of all people, for his plea for legal requirements of civility, when it was Mill himself who was very caustic about precisely such demands in his time. Responding to those who demanded, in Mill's words, "that the free expression of all opinions should be permitted, on condition that the manner be temperate, and do not pass the bounds of fair discussion", Mill noted "the impossibility of fixing where the supposed bounds are to be placed".38 This is not merely because the judgment is one of degree and people may disagree about where to draw the line; the objection is much more fundamental. The judgment about the incivility of an expression is not independent of our disagreement with the substance of the expression. It is not the case that we can separate our judgment about the "manner" of speech (whether it is ''temperate'', and, to use Wright's words, whether it approaches ''the level of a Daniel Webster") from our agreement and disagreement with the speech; we are rarely offended by the "manner" of speech which agrees with our viewpoint. This is a point Mill understood very well when he cautioned against the ''temperateness'' requirement: [I]f the test be offence to those whose opinions are attacked, 1 think experience testifies that this offence is given whenever the attack is telling and powerful, and that every opponent who pushes them hard, and whom they find difficult to answer, appears to them, if he shows any strong feeling on the subject, an intemperate opponent. 39

There is one other related point worth taking from Mill: the pleas for civility hit harder against the dissenters than the supporters of mainstream views. Our disagreement with the expression informs our judgment of incivility, and dissenting views are most likely to be the targets of public disagreement. Edward Shils, who drew a strong connection between the virtue of civility and a commitment to liberal democracy because "[c]ivility treats others as, at least,

38. John Stuart Mill, "On Liberty", in On Liberty and Other Writings, ed. by Stefan Collini (Cambridge: Cambridge University Press, 1989), p. 54. 39. Id.

SPEECH AND HARM

51

equal in dignity, never as inferior in dignity",4o argued that, among other things, "[s]oftly spoken, respectful speech is more pleasing to listen to than hard, contemptuous speech".41 This may well be true, but the psychological sense of pleasure in a hearer is a relatively unimportant goal of public speech, and the "respect" may not be earned. This is a "staff common room" model of public speech, in which conflict avoidance is facilitated by a perception of the relative equality of power of the various speakers. But society as a whole neither shares the goal of conflict avoidance as a paramount value, nor does it recognize the equal power of participants in public discourse, ~d so it cannot be analogized to a college staff common room where "softly spoken, respectful speech" is favoured and incivility penalized. John Stuart Mill again: With regard to what is commonly meant by intemperate discussion, namely invective, sarcasm, personality, and the like, the denunciation of these weapons would deserve more sympathy if it were ever proposed to interdict them equally to both sides; but it is only desired to restrain the employment of them against the prevailing opinion: against the unprevailing they may not only be used without general disapproval, but will be likely to obtain for him who uses them the praise of honest zeal and righteous indignation. 42

Mill goes on to argue on utilitarian grounds that unpopular views should be more protected against the civility requirement than mainstream ones. We need not pursue this argument here; what matters is that the civility requirement is not content neutral, and tends to put unpopUlar, non-mainstream views at a disadvantage. An analogous point can be made with regard to expressive, non-verbal conduct. One can describe an expressive message conveyed, say, by flag burning, in a verbal way ("by burning the flag Mr Johnson expressed his deep abhorrence to the United States' system of government, its foreign policy, etc."), but it is precisely that: a description, or a paraphrase of the message rather than a restatement of the same message in a different fashion. It suffices to try to render the meaning of flag burning, or cross burning, or saluting the flag, or other similar symbolic actions in a verbal form, to see how much is lost in translation.

40. Edward Shils, "The Virtue of Civil Society", Government and Opposition 8 (1981): 3-20, p.

12. 41. Id., p. 13 42. Mill, p. 54.

52

CHAPTER 2

What really matters, in deciding the level of scrutiny, is' not so much whether the object of the restriction can be characterized as "speech" or as "conduct" but, rather, what the purposes of the restriction are. In laying down a principle of freedom of speech, a society decides that speech is special, not in the sense that verbal communication is distinct from other forms of conduct but rather that the communication deserves or calls for a higher level of protection than many other domains of human conduct. This means that restrictions upon communication should be viewed with greater suspicion, and what matters in our characterization of a governmental restriction as implicating the principle of freedom of speech, is whether the purposes and/or effects of the restrictions are related to suppression of communication. In cases of "speech-plus", what matters is whether a restriction of the ''plus'' significantly hinders the communicative message so that the message cannot be conveyed without a "plus" and we may have a reasonable suspicion that the purpose for regulating the "plus" was to target the "speech", that is, the communicative message. In cases of symbolic or communicative conduct, what matters is whether the purposes of the restriction have nothing to do with the communication itself and address non-communicative aspects of the action, or whether the purposes attach to the communication itself. In that case, the purpose of the restriction is indistinguishable from the purposes of restrictions upon purely verbal speech. Contrary to appearances, such a determination need not be hopelessly difficult: often the purposes of a restriction will be suggested by the nature of the expressive action in question. At one end of spectrum, some actions are such that restrictions upon them can be easily supported by reasons unrelated to the suppression of the message. Consider Sunstein's examples of "[p]rotection of the President from assassination, or of the Lincoln Memorial from graffiti".43 Even if the main purpose of the would-be assassin was to express his hostility towards the policies of the president, and if the graffiti writers' aims were to express their views about the ideological legacy of Lincoln, the restrictions upon such activities can undoubtedly be "supported by reference to powerful, legitimate reasons, such as protecting life and promoting clean public monuments".44 These reasons would still stand if it turned out that the message of another would-be assassin, or the message of another group of graffiti writers, was completely different from those mentioned above: restricting their action would still be supported by those same reasons because those reasons are independent of the content of the communication itself. But at the other end of the spectrum lie those regulations of symbolic actions which are unlikely to be triggered by any reason other than a desire to suppress the communication: consider flag burning or the public destruction of a draft card. It is hard to think

43. Sunstein, p. 181. 44. Id.

SPEECH AND HARM

53

of a reason for the suppression of these actions which is independent of the suppression of the message conveyed by the action. Or is it? Suppose a proponent of a legal prohibition of flag burning says: "My purpose is not to suppress expressions of hate towards our country but to prevent breaches of the peace: it is likely that an act of public flag burning will provoke street fights, riots, demonstrations etc." And suppose a proponent of prosecuting those guilty of the public destruction of draft cards says: "My purpose is not to prosecute the anti-draft message. My purpose is to protect the smooth operation of the military draft which is a policy laid down by a democratically elected legislature". Protection of public order on the streets and of efficient conduct of military recruitment may seem to be as "powerful, legitimate reasons" for governmental action as "protecting life and promoting clean public monuments". The examples of non-communicative arguments for prosecution of flag burning and the destruction of draft cards are, of course, not invented but are taken from real cases. In Texas v Johnson, a decision in which the Supreme Court declared that conviction for flag desecration was inconsistent with the First Amendment, Chief Justice Rehnquist argued, in a dissent, that "the public burning of the American flag by Johnson was no essential part of any exposition of ideas, and at the same time it had a tendency to incite a breach of the peace".45 And earlier, in a draft card burning decision, the Supreme Court found that conviction for such action was not in conflict with the First Amendment because the legislator furthered "an important or substantial governmental interest ... unrelated to the suppression of free expression", namely, to further "the smooth and proper functioning of the system that Congress has established to raise armies".46 But there is an important difference between the examples of the assassination of the President and writing graffiti, on the one hand, and burning the flag or destroying an army registration certificate, on the other. In the former pair of examples, the harm attacked by a prohibition is truly independent of the message intended by the actors, and the prohibition is not supported by an attempt to curtail whatever communicative effects the act may have. But in the second pair of examples, the harm cited by proponents of the prohibition is inseparable from the communicative impact of the act. Breaches of the peace resulting from public desecration of a flag follow only as a result of the impact that the message of the burning has upon the audience. It is not that the breach of the peace follows the fact that a piece of fabric is burnt at the 45. Texas v. Johnson, 491 U.S. 397,430 (1989) (Rehnquist, C.J., dissenting). 46. United States v. O'Brien, 391 U.S. 367, 377, 381 (1968).

54

CHAPTER 2

steps of a public building: the breach results from the fact that many people will be outraged, shocked and deeply offended by the destruction of a symbol which they hold in great respect. These effects of outrage, shock and offence cannot take place unless the act is received and interpreted in a particular way by the audience. The restriction is aimed at the direct effect of the communication of a message, and it is supported by an interest in avoiding. the message producing this effect. As Justice Brennan noted, [ilt would be odd indeed to conclude both that 'if it is the speaker's opinion that gives offense, that consequence is a reason for according it constitutional protection,' ... and that the government may ban the expression of certain disagreeable ideas on the unsupported presumption that their very disagreeableness will provoke violence.47

The point is not so much that the "presumption" is "unsupported" but, rather, that even if it was fully supported, it would not add any weight to the justification of a restriction because the law has to protect the unpopular speaker against the hostile audience, not the other way round. Similarly, the effects feared by the legislators and the Court in O'Brien, will only occur if David O'Brien's action has sufficient effect upon others who will then also refuse to cooperate with the military registration authorities. The action of O'Brien produces harm (as perceived by the legislator) only if it becomes an example followed by others; it is not intrinsically harmful. If O'Brien destroyed his card in the privacy of his home, unbeknown to anyone (except himself and, perhaps, military authorities), rather than on the steps of South Boston Courthouse in front of a large crowd, his action would not undermine "a system of registration for individuals liable for training and service".48 So Chief Justice Warren was not quite believable when he claimed that O'Brien was convicted for the "noncommunicative impact of his conduct [namely, frustrating the governmental interest in the smooth and effective functioning of the Selective Service System], and for nothing else".49 It seems that the test the 0 'Brien Court established for symbolic action such as draft card burning, suffers from a major ambiguity about what it means for the regulation to target the communicative aspect of expressive conduct. The well known four-tier test is that when "speech" and "nonspeech" elements are combined in the same course of conduct, regulation is justified if it is within the constitutional power of the Government; if it furthers an important or substantial governmental interest; if the governmental

47. 491 U.S. at 409, reference omitted, quoting FCC v. Pacifica Foundation, 438 U.S. 726, 745 (1978). 48. 391 U.S. at 377. 49. Id. at 382.

SPEECH AND HARM

55

interest is unrelated to the suppression offree expression; and if the incidental restriction on alleged First Amendment freedoms is no greater than is essential to the furtherance of that interest. 50

The key element of the test is the third, italicized, requirement. However, it suffers from a major ambiguity, and Warren's view that O'Brien was convicted for the noncommunicative impact of his conduct is a testimony to this ambiguity. Read literally, the requirement is so thin and lenient that it will hardly invalidate any restriction. It is rare to find a piece of legislation the purpose of which is "related to the suppression of free expression"; usually the purpose is related to some perceived good or the avoidance of some perceived harm and, in order to further the interest, the legislators find it necessary to restrict free expression. Clearly Warren applied this interpretation of "unrelatedness to the suppression of free expression" when he declared that O'Brien's communication was not targeted by the law. But the requirement can be read differently. It can be read: "if the governmental interest is unrelated to communicative effects of the proscribed action".51 So understood, regulations of flag burning and of draft card destruction are defeated by this requirement, because the interest promoted by these restrictions is related to the communicative impact of the proscribed conduct, while prohibitions of the assassination of the President or the writing of graffiti on public monuments survive the test because the purposes promoted are unrelated to the messages conveyed by the proscribed actions. So much for "speech plus" and for communicative (or symbolic, or expressive) conduct. These are cases when conduct (which in an ordinary parlance would not be characterized as "speech") can and should be treated by law as speech because their legal protection is supported by the same rationale as the protection of the speech. But the opposite is also true: not all instances of "pure speech" are necessarily considered "speech" for the purposes of the jurisprudence of free expression, because not all acts of speech are related to the rationales for freedom of speech. If freedom of speech does not encompass (to use Schauer's examples) "freedom to fix prices, breach contracts, make false warranties, place bets with bookies, threaten, extort and so on",52 this is not because these verbal acts belong to inferior categories of speech which do

50. Id. at 377, emphasis added.

5 I. For a classical exposition of this approach, see John Hart Ely, "Flag Desecration: A Case Study in the Roles of Categorization and Balancing in First Amendment Analysis", Harvard

52.

Law Review 88 (1975): 1482-1508. Frederick Schauer, "Categories and the First Amendment: A Play in Three Acts", Vanderbilt Law Review 34 (1981): 265-307, pp. 270-71.

56

CHAPTER 2

not obtain protection from the principle of :freedom of speech. Rather, it is because they do not even register as speech for the purposes of that principle.53 As Schauer wrote more recently: "there are, and properly so, vast domains that are speech in the ordinary language sense of that term but to which the doctrine and the discourse of the First Amendment are not admitted".54 It is not that they are relegated to categories at the lower end of the spectrum of speech and assigned, accordingly, a lower level of protection. Rather, there are many cases of verbal conduct which are rarely, if ever, even contemplated as possible candidates for assigning any level of protection under the freedom of speech principle. Incitement to violence (which is "speech", both in the legal and ordinary senses of the word) is prosecutable if it meets standards of proximity of speech and harm such as the "clear and present danger" test. However, the examples of verbal actions in Schauer's list are not subjected to such a test. Similarly, when an owner of the dog shouts "Kill!" he could not rely on the protection of freedom of speech principles: the reason is not that his words are harmful (which is, of course, also the case) but because they are not taken to constitute "speech" in the first place. This may seem puzzling because a different approach seems to be more straightforward: to characterize all verbal actions as "speech" so that whenever one opens mouth and emits sounds, or takes a pen and writes something, it is automatically "speech", but then to say that ''freedom of speech" covers only some cases, namely those that are related to the rationales for freedom of speech. But illocutions do not necessarily become "speech" in the legal sense. If they did, we would have difficulty understanding the lack of protection for many verbal acts which are no more harmful than many protected acts of speech. Consider Schauer's test: "If no first amendment and principle of free speech existed, would we say that a telephone bookie's operation constitutes a greater danger of harm than the march of the American Nazi Party in Skokie?,,55 One may think of the following answer: if no principle of :free speech existed, we would probably say that the march of the Nazis constitutes a greater danger harm. This does not prove that telephone bookie operations are not "speech" in a legal sense, but only that the march of the Nazi Party derives its protection from the principle of freedom of speech. This is because it resonates with some free speech rationales, such as equal political participation in a discourse on public matters. The existence of freedom of speech forces us to disregard some harms related to speech, the answer goes, as a result of which the calculus of harm between a Nazi demonstration and a bookie operation is

53. See id., p. 268. 54. Frederick Schauer, ''The Speech of Law and the Law of Speech", Arkansas Law Review 49 (1997): 687-701, p. 692, footnote omitted. 55. Schauer, "Categories and the First Amendment", p. 271.

SPEECH AND HARM

57

reversed, compared to the situation that existed in the absence of the principle. The bookie's operation still keeps its status as speech but its comparative value is lower. But this answer will not do: the principle of freedom of speech does not just affect the relative value of different kinds of speech but also defines some categories of verbal acts out of the realm of "speech". If it did not, we would be unable to explain why we do not subject extortions or breaches of contract (through words) to those same tests as some other very harmful expressions, such as incitement to violence or racial hatred. An analogy can be drawn between the technical meaning (that is, a meaning that doesn't coincide with the ordinary meaning) of "speech" in "freedom of speech", and the technical sense of "religion" in "freedom of religion". The analogy has limits which are worth noting at the outset: in freedom of religion the tendency is to extend the legal notion of religion beyond the everyday meaning because the risk of underinclusion is higher than the risk of overinclusion. If the definition of "religion" is overinclusive, then the risk is that some non-religions will obtain privileges related to freedom of religion; if the definition is underinclusive, then the risk is that some genuine religions will suffer discrimination because they will fail to obtain the same rights and privileges as obtained by other religions. Since the harm related to the latter risk is usually seen as higher than the harm related to the former error, the tendency is to err on the side of overinclusion, thereby adopting an expansive definition of "religion" for the purposes of the principle of free exercise of religion. The principle of non-establishment of religion may have the opposite effect; an issue which I have discussed at length elsewhere, and which need not be pursued here.56 Consequently, divergences between the legal and everyday meanings of "religion" are such that the former is usually more expansive than the latter. But in the case of freedom of speech, the divergences between the legal and everyday meanings go both ways: legal notions of speech are both overinclusive (speech-plus, symbolic conduct) and underinclusive (see Schauer's list) when compared with the everyday meanings. But there is also an important similarity with definitions of religion: just as the rationale for having freedom of religion in the first place informs the way in which "religion" is understood for constitutional purposes, rationales for having freedom of speech also affect the boundaries of the meaning of "speech" in the principle of freedom of speech. Since "[t]here are many acts of communication that simply have nothing to do

56. See Wojciech Sadurski, Moral Pluralism and Legal Neutrality (Dordrecht: Kluwer, 1990), pp. 171-80.

CHAPTER 2

58

with what the concept of free speech is all about",57 the concept of "speech" (and not only of ''free speech") is dependent upon the reasons for adopting the principle of freedom of expression. This is inevitable, but it also suggests that the very distinction between "speech" and "non-speech" is manipulable, and that if one wants to deny a degree of protection to a particular act, then one may simply try to define it as non-speech.

4

Speech and Harm: A Case Study of R.A. V.

The two strategies for determining the level of scrutiny which we have discussed so far, the lowlhigh value distinction and the pure speech/speech-plus (and communicative conduct) distinction, both focused on the nature of speech which is subject to putative regulation. But there is a third strategy which is both jurisprudentially and judicially very influential; its focus is not on the nature of speech but rather on the motives for a governmental restriction. The underlying idea is that the legitimacy of governmental prohibitions is related to the nature of the motives for these prohibitions, rather than to the nature of the prohibited conduct itself. The most important version of the motive-related strategy refers to a distinction between viewpoint-based and viewpoint-neutral regulations of freedom of expression. The general principle might be formulated as follows: restrictions that express a viewpoint that is preferred by the government (or conversely, that is disliked by the government) should be subjected to very strict scrutiny. By implication, restrictions that are viewpoint-neutral, and which do not discriminate among different viewpoints, may be subject to more lenient scrutiny because there is little ground for suspicion that they result from improper governmental motives. This theory will be discussed at greater length in Chapter 5. Here my interest is only in one aspect of the theory which is strictly related to the preceding discussion, namely, in the connection between viewpoint-neutrality analysis and harm-related scrutiny. My main purpose in the remainder of this chapter will be to see whether there is any link between the manner in which harm is produced by "viewpoints" and the justification of strict scrutiny of regulations based on such viewpoints. To address the question one must, first of all, distinguish between speech regulations based on viewpoints and regulations based on the entire subject matter. Again, this distinction (much more complex than an initial formulation might imply) will be subjected to a detailed discussion in Chapter 5, but for our present purposes it is important to keep in mind at least a very simplified version of this distinction. The distinction intuitively seems to make sense: it is one thing to restrict an act of expression because it reflects a

57. Schauer, Free Speech, pp. 101-2.

SPEECH AND HARM

59

particular viewpoint on a given matter, but it is something else to restrict public discussion of a particular issue, regardless of the point of view represented by a speaker. For many commentators and judges, the latter regulation seems to be less dangerous from the point of view of the values served by the principle of freedom of speech. That is why it has become a currently accepted doctrine of the United States Supreme Court that subject-matter restrictions should trigger an intermediate level of scrutiny located halfway between that of contentneutral restrictions (which warrant a very lenient test, the only purpose of which is to see whether there is any rational basis for a regulation aimed at promoting a legitimate governmental interest) and viewpoint-based ones (which trigger the strictest scrutiny, the purpose of which is to determine whether there is no other method of achieving this particularly important purpose). The "intermediate" scrutiny is usually represented as one which aims to check whether there is a substantial link between a restriction and an important governmental interest. One immediate comment about the distinction drawn in the preceding paragraph is that the link between the demands for subject-matter neutrality and the motive-based theory of legitimate state prohibitions is much more tenuous than that between the requirement of viewpoint neutrality and the motive-based theory. While it is relatively easy to attribute particular motives to a regulation that disproportionately affects one "viewpoint" more than another, it is much less clear whether restrictions over an entire subject matter may be as easily traced to a specific type of motivation. It may be possible, but it seems less determinate and obvious. But is the assumption behind this whole argument plausible? Is there really a strong link between a viewpoint-specific regulation and the likelihood of improper governmental motivations? Is Geoffrey Stone correct to claim that "the probability that an improper motivation has tainted a decision to restrict expression is far greater when the restriction is directed at a particular idea, viewpoint, ... than when it is content-neutral,,?58 To answer this question one must try to flesh out the notion of "improper governmental motivations". It is not good enough to dismiss viewpoint-based restrictions on the grounds that they reveal improper motivations unless we have a theory of what these improper motivations really are in the context of freedom of expression. A standard answer is this: disfavour for a particular idea or, conversely, preference for another idea, are improper legislative motivations; they must not constitute the grounds for the restriction of a particular utterance. The law in a liberal state, it is often said, must not "declare one perspective right" and seek to

58. Stone, p. 230.

60

CHAPTER 2

create "an approved point of view",S9 just as, vice versa, it must not "infringe speech on grounds actuated by state distaste for what is said".60 But surely this is insufficient: the law prohibits child pornography on the basis that lawmakers disfavour the view that exploiting children for sexual purposes is proper. Proponents of a "viewpoint neutrality" theory may reply that the true motive for this particular governmental restriction is rather to prevent an identifiable harm to children. They may say: we stand by our view that "the government may not restrict expression simply because it disagrees with the speaker's views".61 But in the case of child pornography, the restriction is not imposed simply because of disagreement; it just so happens that pursuing the aim of prevention of harm coincides with the effect of restricting the views which are conducive to the harm. The primary aim, the argument goes, is to avoid harm, and "[t1hat concern might dwarf, or even be unaccompanied by, any bias toward the point of view expressed".62 But this reply does not sound very convincing. It just sounds pedantic to distinguish between a "point of view" that a particular form of conduct is harmful and a judgment about the harm produced by that very conduct. Isn't a viewpoint-discourse just another way of bringing the issue of harm into the reasoning about restrictions of freedom of expression? If so, conclusions about the appropriate level of scrutiny of the speech-harm connection are redundant because they collapse into a determination of the gravity of harm. In order to fulfil their role in the theory of freedom of speech, strategies for determining the level of scrutiny must be harm-insensitive. This point will be directly addressed at the end of this chapter but, to reach that point, we first need to look more carefully at the relationship between, on the one hand, a distinction between regulations based on viewpoints and those based on subject matter, and, on the other hand, a harm analysis. At the outset it is important to note that subject-matter regulations are often based on the distinctiveness and severity of harm of a particular category of speech. For example, the Supreme Court of the United States declared that a state could regulate false, deceptive or misleading advertising by lawyers without also regulating other professions. 63 Although in Bates the Court invalidated the rule which barred truthful advertising by lawyers, Justice Blackmun, who delivered the opinion of the Court, suggested (with regard to false advertising) that

59. The words in quotation marks are from American Booksellers Association v. Hudnut, 771 F.2d 323, 325, 332 (7th Cir. 1985), a decision of the US Court of Appeals for the Seventh Circuit which invalidated an Indianapolis anti-pornography ordinance. 60. David AJ. Richards, Toleration and the Constitution (New York: Oxford University Press, 1986), p. 193. 61. Stone, p. 227, emphasis added, footnote omitted. 62. Steven H. Shiffrin, The First Amendment, Democracy, and Romance (Princeton: Princeton University Press, 1990), p. 18, footnote omitted. 63. Bates v. State Bar ofArizona, 433 U.S. 350, 383 (1977).

SPEECH AND HARM

61

"because the public lacks sophistication concerning legal services, misstatements that might be overlooked or deemed unimportant in other advertising may be found quite inappropriate in legal advertising".64 The Supreme Court, likewise, had upheld a state law which banned the advertising of casino gambling while the advertising of other forms of gambling remained lega1.65 In Posadas the then Justice Rehnquist deferred to the state legislature's opinion that "for Puerto Ricans the risks associated with casino gambling were significantly greater than those associated with the more traditional kinds of gambling in Puerto Rico",66 and that residents of that state would be "induced by widespread advertising to engage in such potentially harmful conduct". 67 In those cases, there was no reasonable suspicion (in the eyes of the Court) that a selective ban was dictated by a dislike of a particular opinion, viewpoint or belief. The regulations seemed to have been firmly based on an evaluation of harm. However, in the case of viewpoint-based regulations the relevance of the harm-prevention motive is much less obvious, and a suspicion that the restriction is motivated by a dislike of the message (regardless of further harms) seems more justified. R.A. V. v. City of St. Paul is a good illustration of the distinction between viewpoint-based and the subject-matter regulations, and of the relevance of the distinction to harm analysis. For our purposes, it will be helpful to examine this decision at length because the opinion for the Court (written by Justice Antonin Scalia) illustrates the connections between these two types of regulations and the harm analysis very nicely. In this decision the Supreme Court struck 'down a city ordinance under which a teenager was prosecuted for burning a cross inside the fenced yard of a black family. The City of St. Paul's ordinance made it a misdemeanour for anyone to place on public or private property a symbol, object, appellation, characterization or graffiti, including, but not limited to, a burning cross or Nazi swastika, which one knows or has reasonable grounds to know arouses anger, alarm or resentment in others on the basis of race, color, creed, religion or gender. 68

64. Id. at 383, footnote omitted. 65. Posadas de Puerto Rico Assoc. 11. Tourism Co., 478 U.S. 328 (1986). 66. Id. at 343. 67. Id. at 344, emphasis added. 68. R.A.V. 11. CityofSt, Paul, 505 U.S. 377, 380 (1992).

62

CHAPTER 2

As Steven Shiffrin observes, "[o]f course, St. Paul was hostile to the messages it sought to criminalize".69 But it was not the hostility to racist messages per se which provided the best justification for the ban. The ban might have been equally well or even better defended on the basis of the special severity of injuries caused by fighting words related to race, colour, creed, religion or gender. Scalia respond~ to the special-hann argument in the following way:' This is word-play. What makes the anger, fear, sense of dishonor, etc. produced by violation of this ordinance distinct from the anger, fear, sense of dishonor, etc. produced by other fighting words is nothing other than the fact that it is caused by a distinctive idea. conveyed by a distinctive message.70

However, the "distinctiveness" of the idea banned by st. Paul collapses into a distinctive harm. It is a plausible judgment that, considering the specific circumstances of time and place, people are more severely injured by verbal assaults related to their race than by those related to, for example, their political affiliation. In this way, there is an analogy between St. Paul's selective ban on some fighting words and the regulation of deceptive advertising by lawyers (as approved in Bates) or of casino advertising (as upheld in Posadas). St. Paul's ban has been characterized by Scalia as viewpoint-based but a subject-matter characterization seems more accurate: on the face of it, no particular view has been selected by the legislature as the subject of legal restriction. Rather, the fighting words "theme" has been selected on the basis of its greater likely hann. The very characterization of "fighting words" is neutral between different (and mutually conflicting) opinions which may be employed in the process of uttering those "fighting words". It is the likely effect upon a hearer which makes these words more akin to an assault (albeit a verbal assault) than to the communication of an idea. This is clear from the judicial construction of "fighting words" as personally abusive epithets addressed in a face-to-face manner to a specific individual and uttered under such circumstances that the words are likely to directly cause an immediate violent response by an average recipient. 7 • Scalia's attempt to show that only one side would be suppressed as a result of the ordinance is not convincing. One can well imagine a pair of mutually conflicting fighting words equally prohibited by the ordinance, such as verbal assaults by white supremacists and by black racists. If, in practice, the former would be more likely to be punished under the ordinance than the latter, this would not be a result of greater legislative hostility towards 69. Steven H. Shiffrin, "Racist Speech, Outsider Jurisprudence, and the Meaning of America", Cornell Law Review 80 (1995): 101-66, p. 119. 70. 505 U.S. at 392-93. 71. See Stephen W. Gard, "Fighting Words as Free Speech", Washington University Law Quarterly 58 (1980): 531-81, p. 580; for more on fighting words. see Chapter 3.3.

SPEECH AND HARM

63

anti-black racism than towards anti-white racism built into the ordinance. Rather, it would be a result of a reasonable judgment made in the process of enforcing the ordinance that, in their specific social context, the former would cause more injury than the latter. The regulation which provided for the selective treatment of fighting words related to "race, color, creed, religion or gender", such as the City of St. Paul's ordinance, seems, therefore, to match more adequately a characterization of "subject matter regulation" than a description of "viewpoint-based regulation". No clear viewpoint is captured by this selectivity. Scalia disagrees, and he responds with two bizarre hypotheticals which supposedly illustrate the viewpoint-discriminatory effect of the ordinance: '[Flighting words' that do not themselves invoke race, color, creed, religion, or gender - aspersions upon a person's mother, for example - would seemingly be usable ad libitum in the placards of those arguing in favor of racial, color etc. tolerance and equality, but could not be used by that speaker's opponents. One could hold up a sign saying, for example, that all 'anti-Catholic bigots' are misbegotten; but not that all 'papists' are, for that would insult and provoke violence 'on the basis of religion.' st. Paul has no such authority to license one side of a debate to fight freestyle, while requiring the other to follow Marquis of Queensbury Rules. 72

But these alleged consequences of St. Paul's ordinance do not follow from the prohibition under challenge, and so the hypotheticals fail. These hypotheticals would be effective if they could show that the ordinance handicaps one, but not another, viewpoint on the same subject. But this is not the case in Scalia's examples. In the former one, a "fighting words" utterance based on a ground other than the prohibited ones and aimed at supporting the argument in favour of racial tolerance, is not being "responded to" by an utterance based on race, colour, etc. arguing in favour of racial intolerance but, rather, by other fighting words based on a non-prohibited ground used to support an anti-tolerance argument. Such utterances would not be captured by the ordinance. The ordinance prohibits fighting words which arouse "anger, alarms or resentment in others on the basis of race, color, creed, religion or gender". Whether the fighting words are used by someone who wants to argue for or against tolerance (racial or otherwise) is irrelevant here. What is relevant for the applicability of the ordinance is whether the hearer's anger, alarm or resentment is caused by expressions related to his or her race, colour etc. It is

72. 505 U.S. at 391-92.

64

CHAPTER 2

therefore conceivable that someone who makes a pro-tolerance argument by attacking the race of an opponent of tolerance in a way which meets the criteria of fighting words, will be prosecuted under the ordinance. And, on the other hand, it is conceivable that the use of fighting words by an opponent of tolerance will be non-prosecutable under the ordinance because the speaker does not refer to the hearer's race, gender etc., but makes, for example, "aspersions upon a person's mother" .. This shows that, contrary to Scalia's suggestion, no particular viewpoint in the controversy concerning racial tolerance is being adversely affected by the ordinance. So the two verbal assaults which he compares are not equivalent to each other. The same applies to the second hypothetical. According to Scalia, a sign attacking all "anti-Catholic bigots" would have to be tolerated while a sign attacking the "papists" would be prosecuted under the ordinance, because only the latter would be based on "religion" which is one of the ordinance's prohibited subjects. It takes some imagination to think that someone could be stirred to anger and alarm by a placard proclaiming: "All anti-Catholic bigots are misbegotten". But even leaving the unreality of the hypothetical aside, the verbal assaults upon "anti-Catholic bigots" and "papists" are not two opposed viewpoints in one controversy. An assault upon "anti-Catholic bigots" reflects a viewpoint in a controversy about religious tolerance and bigotry.73 But an assault upon "papists" involves a viewpoint in a religious controversy within organized Christianity. The opposing viewpoint to the first placard would be one attacking "all anti-Catholic tolerationists" (that is, those who object to what they see as Catholic bigotry and intolerance). Thus, there is no equivalence between the two placards imagined by Scalia. If they were equivalent, that is, if the court found that an assault upon "anti-Catholic bigots" was a disguised verbal assault upon non-Catholics because of their religion, then the placard would include prohibited fighting words and would be liable to prosecution under the ordinance. Either way, the validity of Scalia's hypothetical would be undermined. Assume for the moment; however, that a selective regulation of fighting words, such as the regulation in St. Paul's ordinance, is viewpoint-based rather than subject-matter based. Why should it matter? This question brings us closer to our starting point in this Section: the relationship between the level of scrutiny of speech restrictions and the viewpoint/subject-matter distinction. One would expect that characterization of a regulation as viewpoint based would be important to Scalia in supporting his conclusion about the invalidity of the regulation. And yet, what is puzzling is that the characterization of the ordinance comes after Scalia announced that St. Paul's ordinance is subject-

73. This point was made by Justice Stevens in R.A. V., 505 U.S. at 435 (Stevens, 1., concurring in the judgment).

SPEECH AND HARM

65

matter based, and that there is a strong First Amendment presumption against such subject-matter distinctions: ''The First Amendment does not permit St. Paul to impose special prohibitions on those speakers who express views on disfavored subjects".14 This sounds like a non-differentiated treatment of all content distinctions, whether based on viewpoints or subject matters. It provoked the response by Justice Stevens, in his dissent in R.A. v., that "[nJot all content-based regulations are alike . . . [and] that some content-based restrictions raise more constitutional questions than others,,75 and, extracurially, that the R.A. V. decision failed "to follow through on its insight that content discrimination is not the same as viewpoint discrimination".76 This insight is ignored in the crucial parts of Scalia's opinion where he establishes the main legal principle for which R.A. V. stands now as the authority. This principle is that content distinctions are not permitted even with regard to otherwise unprotected categories of speech such as "fighting words". The rule of content neutrality has now been extended to categories of proscribable speech (which include, under the First Amendment current doctrine, fighting words, as well as obscenity and defamation). In Scalia's view, supported by four other Supreme Court judges, these categories are not "entirely invisible to the Constitution,,77 and "content discrimination unrelated to their distinctively proscribable content" is not permitted. 78 Partly as a result of the undifferentiated treatment of content neutrality (that is, without drawing a line between viewpoint neutrality and subject-matter neutrality) in the central part of his opinion, there is a strange myopia in R.A. V. towards the relations between subject-matter distinctions and the harm caused by speech. Scalia's exception to the prohibition against content discrimination: "When the basis for the content discrimination consists entirely of the very reason the entire class of speech at issue is proscribable, no significant danger of idea or viewpoint discrimination exists."79 An obvious response would be that this is precisely the case with the St. Paul's ordinance which proscribed an entire class of fighting words in order to prevent severe psychological injury to hearers. This very reason forms the basis of "content discrimination" manifested in the selective treatment of fighting words based on race, colour, religion and gender. The same ~rgument applies to another possible reason for 74. Id. at 391, references omitted, emphasis added. 75. Id. at 429 (Stevens, J., concurring in the judgment). 76. John Paul Stevens, "The Freedom of Speech", Yale Law Journal 102 (1993): 1293-1313, p. 1309. 77. 505 U.S. at 383. 78. Id. at 384. 79. Id. at 388.

CHAPTER 2 proscribing "fighting words", namely, the avoidance of violent responses which are likely to be provoked by such words. On the basis of real-life experience, the legislator may well fmd that the fighting words which refer to race, colour, religion or gender inflict more severe injury, and tend to be more fightprovoking, than fighting words based on some other characteristics. It is difficult to impeach this diagnosis. As Kent Greenawalt observes: . "A legislature has chosen to punish the instances of speech within a category that cause the greatest harm, although it could also punish other instances within the category"; the selectivity is plausibly based on the ground that "such symbols of race hate [as proscribed by the ordinance] probably do have a greater propensity to cause violence".80 Similarly, Shiffrin points out that, in a prosecution under a nondiscriminatory fighting-words statute (that is, one which does not specify the grounds on which the fighting words are proscribed), "a jury could determine that racist speech was especially damaging" and so "[t]he puzzle is why the St. Paul City Council could not do the same".81 The rationale that Scalia produces for this exception to the general rule against content discrimination and which renders the entire class of speech proscribable is as follows: "Such a reason, having been adjudged neutral enough to support exclusion of the entire class of speech from First Amendment protection, is also neutral enough to form the basis of distinction within the class.,,82 This proposition implies that Scalia would attribute the virtue of neutrality to a distinction between proscribable and protected categories of speech if there is a proper basis for the distinction, such as where severe harm is expected to follow from a particular speech. The following example illustrates the operation of this principle: [T]he Federal Government can criminalize only those threats of violence that are directed against the President . . . since the reasons why threats of violence are outside the First Amendment (protecting individuals from the fear of violence, from the disruption that fear engenders, and from the possibility that the threatened violence will occur) have special force when applied to the person of the President. 83

The difference between this example and St. Paul's ordinance is questionable. One can, plausibly, paraphrase Scalia's defence of the special penalty for threats against the President in the following way: "The government can criminalize only those fighting words that are based on race, color, creed, religion or gender since the reasons why fighting words are outside the First 80. 81. 82. 83.

Kent Greenawalt, Fighting Words (Princeton: Princeton University Press, 1995), pp. 57-58. Shiffrin, "Racist Speech", p. 118. 505 U.S. at 388. Id. at 388, reference omitted.

SPEECH AND HARM

67

Amendment (that is, protecting individuals from anger, alarm or resentment, from the disruption that such emotions engender, and from the possibility that violence will occur) have special force when verbal assault uses the victim's race, color, creed, religion or gender". Indeed, this is suggested separately by Justices White and Blackmun in R.A. V.84 It is important to emphasize that the purpose of my argument so far has not been to suggest that subject-matter distinctions are more likely to identify a real harm arising from an utterance than viewpoint-based distinctions. The relationship between speech and harm is independent of the taxonomy of regulations as content-neutral, subject-matter-based and viewpoint-based ones. 85 The reason why content-neutral regulations are more palatable than content-based regulations is not because the former tend to capture the harmful speech better than the latter - this would be an unprovable assertion. Rather, content-neutral regulations have less impact on the values served by the freedom of speech principle. The same applies to a distinction within the category of content-based regulations. Thus, restrictions based on an entire subject matter may not be more capable of reducing the harm of the speech than restrictions based on a viewpoint, but, as will be demonstrated in Chapter 5, the motivations and effects of the former type of restrictions are less problematic than is the case with the latter ones. To foreshadow the argument developed in Chapter 5: as far as motivations are concerned, it is less likely that a subject-matter regulation is triggered by the legislator's intolerance of a particular idea or opinion. As far as the effects are concerned, shrinking the agenda of a debate in a particular setting is less harmful to free debate within a society than official discrimination against a particular side in a given controversy. Harms caused by speech affected by both types of regulations may be equal - there is no apparent correlation between the nature of the regulation and the severity of the harm avoided. With regard to R.A. V., the characterization of a selective treatment of some categories of "fighting words" as subject-matter selectivity (contrary to Scalia's characterization) does not necessarily add anything to the justification of a selective prohibition to prevent harm caused by those fighting words which come within the ambit of the prohibition. Indeed, one may claim that an openly viewpoint-based regulation would even more effectively capture the really harmful speech-acts than one which merely

84. Id. at 407-09 (White, J., concurring in the judgment); at 415-16 (Blackmun, J., concurring in the judgment). 85. For a good discussion of various aspects of relations between viewpoint regulation and harmbased regulation, see Shiffrin, The First Amendment. Democracy. and Romance. pp. 17-23.

68

CHAPTER 2

penalizes those speech-acts defined selectively through their subject matter. The subject-matter definition was based on the judgment that assaultive expressions based on race, religion etc. are more harmful than others. The argument could be made that assaultive expressions which are based on claims about the racial inferiority of a traditionally under-privileged racial group are even more harmful than all other race-related fighting words. The fact that such a viewpoint-based regulation would apply to "fighting words" only, which are more like assaults than the communication of ideas and which therefore inflict a proscribable harm on their victims, largely removes the usual concerns which are raised by viewpoint regulations. But such viewpoint-based regulation of fighting words would certainly be much less politically palatable to the community than a subject-based regulation which is immune to the objection that the state is siding with one racial group against another, or that it affords different levels of protection to different groups. The "fighting words" example may be atypical from the perspective of the relationship between viewpoint regulation and harm because ''viewpoint'' and "content" count for very little in "fighting words". By their nature, the communicative ingredient (which is a vehicle of a viewpoint or a subject matter) of such expressions is very low, and what matters in characterizing an expression as "fighting words" is the context and form of the expression. These are personally abusive words which are directly targeted at a particular individual whom they describe, and the target of the assault is unable to avoid (or to avoid easily) the assaultive message. A somewhat different situation occurs when an expression has a high communicative ingredient (for example, in a speech given to a wide audience, where any particular hurtful or shocking words cannot be seen as a targeted assault upon an individual) but the proximity and severity of the harm produced by the expression may be a reason for a legitimate suppression. Let us therefore restate our initial question of this section: can one make a meaningful distinction between a viewpoint regulation, which should be avoided, and a harm-based regulation, which may be justified? The argument against drawing such a distinction might run as follows. Suppressing an expression based on a resulting alleged harm reflects a viewpoint that the expression is indeed harmful. Since the characterization of a given effect as "harmful" is morally laden and thus necessarily controversial, it reflects a viewpoint which may not be, and is usually not, shared by all. A harm discourse, when applied to expressions, must therefore collapse into a viewpoint discourse. And, vice versa, viewpoint regulations are usually supported by alleged harms. Laurence Tribe puts this point more strongly: "[alII viewpoint-based regulations are targeted at some supposed harm".86 Hence, the answer to the question posed as the starting point of this section 86. Tribe, p. 925.

SPEECH AND HARM

69

about the redundancy of viewpoint analysis in the conceptual framework of levels of scrutiny of restrictions, would be that a viewpoint discourse does indeed collapse into a harm discourse. But this argument proceeds too quickly. The role of the free speech principle is to insulate the sphere of expression from legal restrictions based on an ordinary harm calculus - that is, on the determination that the negative consequences of an act prevail (even if only marginally) over the positive ones. No robust regime of speech protection would survive if a finding of net harm were sufficient to justify a restriction. Harm resulting from an expression may be mitigated by counter-expressions, by the fact that people have the good sense and strong enough moral values not to adopt harmful views, and by the fact that harmful opinions will disqualify themselves from general acceptance when people realize the consequences of acting on them. To be sure, these are not fool-proof guarantees which would eliminate the harm of an expression, but they nevertheless mitigate the harm. However, these mitigating effects only apply to some types of speech-harm linkages, namely those where the harm results from the fact that a large number of people may be persuaded to accept the view expressed, and act on it. "Revealing military secrets to the enemy of your nation is the right thing to do"; "Our President is a crook"; "Having sex with minors is legitimate" - all these views may be harmful when expressed to others, but this harm only occurs as a result of these views being believed and possibly acted upon by the hearers. A number of intervening factors, such as counter-speech or prior beliefs by the hearers, can do much to mitigate the harm. Causation of harm through the mental intermediation of the hearers is well captured by the "viewpoint" discourse. "Viewpoints" produce harm by persuading the hearers to adopt one of a number of competing opinions about a given thing but the process of persuasion is vulnerable to the impact of counteracting factors. The fact that expressions of viewpoints are typically subject to a number of harmmitigating factors supports the idea that legislators should and may abstain from regulating viewpoints - the net harm of viewpoints expressed is usually low. But there are many other ways in which expressions produce harm. The modus operandi of "fighting words" is one such way - injury inflicted upon a hearer by assaulting him with an unexpected or difficult-to-escape verbal attack is typically not subject to mitigation before injury occurs. This is well captured by the original characterization of "fighting words" in Chaplinsky: "those which by their very utterance inflict injury or tend to incite an immediate

70

CHAPTER 2

breach of the peace". 87 Strictly speaking, it is incorrect to say that fighting words inflict injury "by their very utterance" (they must be heard to create a harmful effect), yet it is correct to say that they do not have to be heard, understood and accepted by third parties to victimize the hearers. Other ways of producing harm include issuing a threat which immediately causes fright; advocacy to violence where there is a reasonable likelihood that imminent and grave violence will follow; false commercial advertising where the hearers/readers have little means to verify the truthfulness of the claims; verbal disclosure of important and damaging private information about a person whose privacy is thus violated, and so on. The absence or ineffectiveness of usual mitigating factors can be due to different reasons in these various cases. For example, in some cases the harm occurs in the very moment and by the very fact of an expression, or it may be so imminent that there is no time or means to counteract it, or there may be no practicable way for the hearers to check the facts stated in the expression. This explains why, in a classical formulation, actionable libel concerns only statements of fact and not of opinion. But the common ground is that the harm does not occur through a process of persuading the hearers to adopt an expressed opinion. In such circumstances there are no mitigating factors which, in the case of an expression of a "viewpoint", may stand between the expression and the resulting harm. The argument for regulation is therefore stronger than in the case of viewpoint restrictions, and the case for more lenient scrutiny of regulation obtains additional support from this analysis It is "additional" to the standard argument that in the case of regulations other than those based on viewpoints, suspicion about improper legislative motives (that is, intolerance towards a point of view) is less justified than in the case of viewpoint restrictions. This standard argument will be examined in Chapter 5. My suggestion here is that an examination of the way in which various expressions produce harm and, in particular, of different opportunities for various factors to mitigate possibly harmful effects of an expression, also supports a call for a less demanding scrutiny of restrictions based on a subject matter rather than on a point of view. A ban upon child pornography88 is a good example of a regulation which demands a less-than-strict scrutiny because it can be seen as one where the usual mitigating factors cannot be relied on to avert the harm. A restriction on written material which describes or advocates sex with minors might be construed as a viewpoint regulation because a number of factors which mitigate the harm produced by the dissemination of the material are likely to intervene. For example, strong social reprobation against such views, or the readers' own

87. Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942). 88. See New York v. Ferber, 458 U.S. 747 (1981).

SPEECH AND HARM

71

pre-existing contempt towards paedophilia, may nullify the harmful effect of dissemination of such materials. But a ban upon the production and dissemination of films and photos depicting sex with minors is justified by the injury already inflicted upon the children used to produce these pictures. Neither the counter-arguments nor the revulsion provoked by the pictures can prevent or erase this injury. Even if it may be claimed that these materials express a viewpoint, what legislators are attacking here is not viewpoint-related harm, that is, not a harm produced by the fact that readers will be persuaded that sex with minors is a right thing to do, but rather the harm which has been already inflicted upon some children, and which is unrelated to the "perlocutionary"S9 power of the material. The ban upon the' possession, viewing, distribution etc. of such material may therefore be seen as an indirect way of banning the production of it, and preventing the harm (which is independent of any persuasion-related harm) inflicted upon its child models. This is not to say that no persuasion-related harm can ever occur, or that restrictions of viewpoints are always impermissible. Hostility to viewpoint regulations is not absolute, and no more absolute is hostility to legal actions which are undertaken to avoid the harm produced by persuading listeners to adopt the views expressed. The rationale for hostility towards viewpoint regulation is related to the distorting effects of governmental interference with public debate. But when this rationale does not apply, or poses an insignificant danger compared to the harm avoided by the interference, legal attack upon the persuasion-related harm may be justified. As an example, one may posit restrictions on tobacco advertising. This arguably involves viewpoint regulation and also a regulation aimed at the persuasive force of the viewpoint. Anti-smoking advertisements are legal, while the opposite viewpoint is banished. This may be seen by some as a regrettable deviation from the principle of freedom of speech.9o But one can argue that the restrictions on tobacco advertising, though viewpoint discriminatory, are nevertheless justified, even under a strict scrutiny of legal restrictions. This is because the gravity of harm resulting from the ads is high, the effectiveness of mitigating factors is limited (due to pro-smoking messages prevalent in the mass culture, older films, entrenched habits, etc.), levels of suspicion that the government acts for illegitimate reasons are low, and the distorting function of the ban is minimal. Rather than skewing the debate against the pro-smoking viewpoint,

89. On "perIocutions" see John R. Searle, Speech Acts (Cambridge: Cambridge University Press, 1969) pp. 25,45-47. 90. See, e.g., Martin H. Redish, ''Tobacco Advertising and the First Amendment", Iowa Law Review 81 (1996): 589-639.

71

CHAPTER 2

the ban may be seen as preventing the distortion resulting from the dominance of the tobacco industry's financial power over the more meagre means of the anti-smoking lobby. The combination of these characteristics suggests that the rationale for disliking the persuasion-related viewpoint in the anti-smoking advertising legislation is less justified than usual.

CHAPTER 3 SPEECH AND EQUALITY

1

Equal Opportunity and Public Speech

In an important article in 1975, Kenneth L. Karst claimed that "[t]he principle of equality, when understood to mean equal liberty, is not just a peripheral support for the freedom of expression, but rather part of the 'central meaning of the First Amendment'''.1 Since then, the call for recognition of the egalitarian ingredient in the principle of freedom of speech has become an important theme in the literature. In the United States, a number of writers have suggested that freedom of speech must be viewed not only through the prism of a "libertarian" First Amendment but also of an "egalitarian" Fourteenth Amendment. Most famously perhaps, Catharine MacKinnon has postulated an approach to the constitutional doctrine of free speech which "tak[es] equality seriously", on the basis of her diagnosis that the "law of equality and the law of freedom of speech are on a collision course" in the United States. As a positive model, she identified the Canadian approach to constitutional freedom of speech as exemplified by its restrictions upon pornography and racist speech. These restrictions, MacKinnon claimed, made "the hate propaganda prohibition and the obscenity law of Canada (properly interpreted) into equality laws".2 But there is a difference between Karst's claim and the more recent demands for an egalitarian approach to freedom of speech. Karst was primarily concerned with the application of the principle itself. When he discerned the traces of recognition, by the Supreme Court, of the equality principle in the First Amendment, his focus was on the rejection of content censorship by the government, and on various ways of equalizing access to means of public communication, in particular, access to the broadcasting media. He was not concerned about how the content of speech may produce inequality which the government is required, by the principle of freedom of speech, to eradicate through speech control. What has emerged, however, in much of the recent writing about the collision of freedom and equality, particularly with reference to such issues as pornography and racist speech, is that content restrictions on

1.

2.

Kenneth L. Karst, "Equality as a Central Principle in the First Amendment", University of Chicago Law Review 43 (1975): 20-68, p. 21 (citing New York Times v. Sullivan, 376 U.S. 254,273 (1964)). Catharine MacKinnon, Only Words (Cambridge, Mass.: Harvard University Press, 1993), p. 103.

74

CHAPTER 3

private speech are seen as effective ways of ensuring access to public discourse. As one proponent of this approach suggested: "In granting equal liberty, the first amendment may implicitly require dismantling even private structures of prejudice as vital to the fulfillment of its free speech guarantee".3 In order to examine these claims, it is important to be clear about what sort of equality is demanded in these contexts, and also the equality of what. As I will argue further in this section, the most plausible type of equality that one can demand in the realm of communicative acts is the equality of opportunity of reaching one's desired audience with one's communications, or, as I will put it, the equality of opportunity to speak and be heard. Regrettably, often questions about the nature of equality stipulated in the domain of speech are left unanswered, with a resultant lack of conceptual clarity. For example, at times MacKinnon describes her ideal as that of "equal access to speech",4 as if speech itself was a good, the access to which is subject to the rules of distributive justice, and at other times, she describes it as "equal speech",5 as if the attribute of equality belonged to the speech itself. In the same vein, she deplores "a substantial lack of recognition that some people get a lot more speech than others".6 Yet, taken literally, the ideal of "equal speech" is unintelligible unless we assume a radically relativistic position within which no judgmental distinctions about the value of different speech acts can ever be legitimate.7 At other times, however, she defines her egalitarian ideal as that of equality of opportunity.8 The importance of establishing clarity about these matters is suggested by the fact that often the proponents of an equality-based theory of freedom of speech oscillate between the notion of equality simpliciter and equality of opportunity, and these are not the same things. The just-quoted ambiguities in MacKinnon are one example. Another is provided by Mary Ellen Gale who moves between arguing that "everyone [should be] entitled to the same free speech resources as everyone else,,9 and demanding "not equal speech resources, but equal opportunities to develop them".10 This latter concept of "equal opportunity to develop speech resources" is puzzling. It may, for example, mean equal access to broadcasting and press media, to public fora, etc.

3. 4. 5. 6. 7. 8. 9.

10.

Mary EJlen Gale, "Reimagining the First Amendment: Racist Speech and Equal Liberty", St. John's Law Review 65 (199 I): 119-85, p. at 154. MacKinnon, p. 72. Id. Id., footnote omitted, emphasis added. See Frederick Schauer, Free Speech: A Philosophical Enquiry (Cambridge: Cambridge University Press, (982), p.62. MacKinnon, p. 99. Gale, p. 157. Id., p. 158.

SPEECH AND EQUALITY

75

Yet clearly something more is envisaged by Gale because she explains, for instance, that her principle of equal opportunity may be violated when the "government's. failure to restrict certain messages (such as racist ones) guarantees that other messages (such as the views of the victims of racism) will not have an equal opportunity to be formulated, articulated, or heard". II The obvious implication is that the very presence of certain messages, identified by their content, is destructive to the equality of opportunity with regard to speech. But note that if equality of opportunity is now being attributed, not so much to the speakers but to the messages themselves, so that all messages should have an equal opportunity to be "formulated", then surely much more is needed than just excising the opposite messages from the public discourse. Clearly, one should also aim to create positive conditions in which there is an equal likelihood that all possible messages will be "formulated, articulated, or heard". Such an ideal is patently implausible because an unequal opportunity for some messages to be articulated may have nothing to do with the government's failure to censor other messages, or with anything vaguely invidious in any way, but rather with the fact that no-one, or very few people, would wish to articulate them even in a totally level playing field of ideas. Do messages praising the lifestyle of an ancient Spartan warrior have as equal an opportunity to be "formulated and articulated", for instance, in contemporary Switzerland, as all the other messages regarding a recommended lifestyle? Recall Nozick's advice: "It is helpful to imagine cavemen sitting together to think up what, for all time, will be the best possible society . . .. Do none of the reasons that make you smile at this apply to us?"12 The likelihood of a particular viewpoint being publicly expressed is dependent upon there being a "constituency" of the viewpoint or idea; and speakers are dependent upon the listeners, for we are both speakers and listeners most of the time. The government or powerful media can (unfortunately) do much to stifle the views present in a given community but can do very little to generate views that appeal to no-one. As one critic of the "fairness doctrine" in American electronic media asked: "is it really necessary to the proper functioning of a democracy that the federal government assure platforms in every medium, in every community, for the ranting of bizarre conspiracy theorists, paranoid delusionists, flat-earthers, anarchists, and others witho~t any significant constituency?,,13 Regardless of the grounds for his distaste of the "fairness doctrine", which need not concern

11. Id., p. 160. 12. Robert Nozick, Anarchy, State, and Utopia (Oxford: Blackwell, 1974), pp. 313-14. 13. Adrian Cronauer, "The Fairness Doctrine: A Solution in Search of a Problem", Federal Communications Law Journal 47 (1994): 51-77, p. 74.

76

CHAPTER 3

us here, Cronauer's question suggests, at the very least, that the meagre public presence of some viewpoints need not be seen as a matter of unequal opportunities, calling for a remedy. Obviously the concept of "equality of opportunity" with regard to speech must mean something more modest than an equal likelihood that all ideas or viewpoints will be articulated or heard: it cannot mean that every single idea that we can think of will have an equal opportunity to be formulated and articulated, but only that certain constraints on their formation and articulation will be absent. If this restricted notion of equal opportunity is adopted, then the difficulty is how to draw the line between those constraints upon equality of opportunity that do count (and that the government has a duty to mitigate or remove) and those that do not, even though in fact they make some messages less likely to be formulated and articulated. The line, to many proponents of an equality-based theory of freedom of speech, is not between the access-based constraints and the content-based constraints; they make it clear that the very presence of certain contents in public discourse adversely affects the equal opportunity of proponents of opposite contents. Gale, for one, invites us "to liberate our concept of free speech to permit regulation based on content when necessary to ensure that all speakers have, not equal speech resources, but equal opportunities to develop them".14 So the clear implication from this is that the very presence of certain contents is hannful to the equal opportunity to "speech resources". In this case, the subject-matter of this "opportunity" must include at least some conditions which assure that all messages will have an equal opportunity to be formulated in the first place, and then be publicly articulated and heard by others. I say "some conditions", but not all, because as we have just suggested, equalizing all conditions which affect the likelihood of all messages to be formulated is patently absurd. How, then, can we draw a line between these two types of conditions in a principled way? Note that one can protest that the preceding reasoning falsely assumes that an "equal opportunity" must apply to all possible messages that we can think of, and that the reductio ad absurdum is unwarranted: after all, only some ideas are likely to be generated in a given society at a given time, and the point is to make sure that those ideas should have an equal opportunity to be heard, and should be protected against the destructive force of opposite messages. But such a protest is question-begging because, if the concern of equality-based theorists of freedom of speech is that the toleration of some contents will annul (or reduce) the opportunity of some messages to be formulated, then we must have some baseline idea about which messages would have been formulated in the absence of those deplorable conditions. However, such a baseline is impossible to construct other than by imagining a stock of all possible ideas that we can think

14. Gale, p. 158, emphasis added.

SPEECH AND EQUALITY

77

of, because any actual conditions which would justify the construction of such a baseline may be presented as affecting an equal opportunity for some messages to be formulated. To return to the example on which we based the earlier reduction ad absurdum, if someone claims that the possibility of messages praising the ancient Spartan lifestyle in contemporary Switzerland should be removed from the baseline because of the actual real-life conditions of a contemporary Swiss society, then a proponent of such messages, albeit an arguably lonely one, would say that this is precisely the problem: that those very conditions operate to adversely affect the equal opportunity of those messages to be formulated, articulated, and heard. This is not to deny that in modem societies, as we know them, some types of contents have a much greater impact upon the domination of certain viewpoints in a society than others, and that often these dominant viewpoints, if based on racist, sexist or class-based superiority, affect the hearers much more than other viewpoints. The point, rather, is to see how the presence of these viewpoints can be said to affect equal opportunity in the area of speech. This is the only question which I am concerned with in this chapter, precisely because of the popularity of the "equal opportunity" discourse with regard to the freedom of speech. But we cannot intelligibly tackle this question unless we have a clear answer to a prior question: an equal opportunity to what? Before I take up this question more directly, it is important to ascertain the characteristics of an equal-opportunity discourse in general. Is the conceptual framework of "equality of opportunity" applicable to freedom of expression? A good point at which to begin this inquiry is by reflecting upon the notion of "opportunity" and the function of equal-opportunity discourse. The very notion of "opportunity" implies a capacity to access socially valued goods through one's effortful action. "Opportunity" is an instrumental good in that it is a stepping stone to attaining another good which is the real object of human desires and pursuits. There is nothing inherently good about having "opportunities"; their goodness is parasitic on the good which the opportunities lead to. What an opportunity does is empower an individual to attain those other goods through her own action, and redress the impact of those factors which are outside the individual's control and which obstruct the attainment of those goods. To have an opportunity to X; means that the obstacles or prohibitions to the attainment of X are not insurmountable. Of course the criterion of "insurmountability" is a matter of degree: it may be that the obstacles and prohibitions are such that they are theoretically "surmountable" through the agent's action, but the costs incurred by the agent in order to achieve X are so great that, in the eyes of common sense, they reduce the "opportunity" to nil.

78

CHAPTER 3

But this only demonstrates an unswprising fact that "having an opportunity" is itself a matter of degree. To say that we have an opportunity to X suggests that the only thing which stands between us and an X is our own choice and effort, and that no factors which are fully beyond our control affect the attainment of X This suggests that the language of opportunities tacitly assumes at least two things: frrst, thai we can identify reliably the factors which we can affect through our own action, and second, that our choices and our capacities to undertake an effort are dependent on the factors which are ultimately under our control. Both these assumptions may be legitimately seen as question-begging. They raise fundamental problems of free will and determinism, and, as the second assumption suggests, there is a danger of an infmite regress inherent in the assumption that the willingness to undertake an effort is itself something which we can control through our own will. The regress is suggested by the question of whether we can control the decision to control our willingness to undertake an effort. But I need not get embroiled in this discussion here; all that matters is that the language of opportunities presupposes that these assumptions are true. Choice and effort are the only factors which connect an opportunity with a desired good, and if that choice and effort are outside the individual agent's control, an opportunity is illUSOry. But it is important to emphasize that choice and effort do stand between the opportunity and the desired good: if we can get from here to there without any effort on our part (or when an effort is trivial), then it is misleading to talk about an opportunity: talk about a "chance" or a "likelihood" is more apposite. As Tom Campbell observes: "[ilt is always ... relevant to ask concerning an opportunity what effort is involved in taking advantage of it and to doubt in cases where no real effort is involved whether it is appropriate to speak of an opportunity at all".15 For instance, it does not make sense to talk about an "equal opportunity" to win a lottery because the result, by the very nature of the practice, cannot be decided by human effort. However, talk of "equal opportunity" to be admitted to a university, or to be hired by a company, or to receive a contract from the government (the spheres where "equal opportunity" talk typically applies) does make sense. If all the conditions of attainment of a particular good were only dependent upon the will and effort of those who desire this good, we would have a situation of full opportunity, and talk of equality of opportunity would be redundant. But it is rarely, if ever, possible to eliminate the impact of factors which are outside an agent's control, upon the attainment of a desired good. Equal opportunity is then the second-best ideal: it demands that the impact of those factors which are beyond individual control be equalized. Sometimes, this

15. T.D. Campbell, "Equality of Opportunity", Proceeding of the A.ristotelian Society 75 (197475): 51-68, p. 56.

SPEECH AND EQUALITY

79

can be done directly, by equalizing the access to a resource the use of which is under a person's control. In other cases it can only be done indirectly, by lessening the weight (or the degree of the impact) of a factor on attainment of the desired good, if the equalization of access to the factor is impossible or very costly. But, whether the equalization proceeds in a direct or indirect manner, the ideal of "equal opportunity" requires equalizing those factors which determine the desired outcome and are not under one's control so that, once the opportunities are equalized, the outcome is to an equal extent a function of both the choices and effort of all the agents who desire the good in question. The question is now whether "speech" lends itself to such an analysis. Although MacKinnon, at times, refers to the ideal of "equal speech", the concept is so puzzling that it is hard to treat it other than as a metaphor for equal expressive opportunities (and, as we have just seen, "an opportunity to X' and "X' are not identical since they are, under ideal conditions, separated by factors that are under the agent's control). But what is an opportunity in the area of speech an opportunity to? Surely it would not make much sense to demand "an equal opportunity to speak". This is too thin an opportunity to be meaningful. On the other hand, it would not be plausible to claim "an equal opportunity to convince". The likelihood of convincing the audience is largely dependent upon the audience's prior views which belong to the category of factors outside of the speaker's control, and which are not the subject of "equalization" in the equal-opportunity discourse. This is due to various causes of disagreement between people which can be explained neither by a lesser rationality of some people than others, nor by the fact that those different views are merely rationalisations of people's narrow interests, which themselves are different. John Rawls labelled these reasons "the burdens of judgment".16 In Rawls's view, those causes include the fact that the evidence bearing on a case is often conflicting and complex; there may be disagreement about the weight of different kinds of considerations that are relevant; concepts are often vague, indeterminate and subject to hard cases; our experiences are different and these differences affect the way we assess evidence, etc. The resulting moral and political differences among people do not point to any aberration or pathology, nor to lesser persuasive powers of some speakers compared to others (although all these factors may naturally occur in particular cases). Rather, they have to do with notorious facts of social morality: we neither can, nor should we, attempt to remove these sources of moral disagreements. Even assuming that hearers are

16. John Rawls, Political Liberalism (New York: Columbia University Press, 1993), pp. 54·58.

80

CHAPTER 3

fully receptive to speakers' arguments, a number of decisive factors of the effectiveness of persuasion are, and should be, outside the powers of speakers. The most plausible interpretation of "equal opportunity" in the field of public expression is gleaned by equating it with the "equal opportunity to speak and to be heard", or, more briefly, the "equal opportunity to be heard". This is not an unusual formula, and the proposition that "government must afford all points of view an equal opportunity to be heard" can be found not only in academic writings but also in judicial doctrine. 17 An immediate suggestion is that it postulates not so much an equal opportunity to persuade one's audience, but simply to get one's message across to the audience. This is more than an opportunity to speak, and less than an opportunity to convince; rather, it is an opportunity to convey one's message to the audience which the speaker wants to reach. Such an opportunity sounds in accordance with the language of opportunities because what stands between the opportunity itself and the desired good (that is, actually reaching the audience with one's message) is predominantly a speaker's will and effort. Further, such an opportunity cOITesponds to an ideal which is attractive and easily recognisable. When I think about what opportunities I want to have regarding my public speech, I may visualize the following ideal without any difficulties: speaking to other people and knowing that they listen to me, take me seriously (if I intend to be serious), do not tum away from me once they recognize who I am, and determine their attitude towards me on the basis of their consideration of what I have said, rather than on the basis of preconceived conceptions shaped by my opponents, etc. This sort of opportunity is both important and desirable. I do not merely want an equal opportunity to speak in public: I may end up confined to a Hyde Parkstyle speakers' comer where the substance of what I have to say will not matter much. It is the effectiveness of a right which makes it morally attractive (which is not to say that having a right, even if ineffective, is not a desirable thing compared with a situation in which no such right is acknowledged and respected), and in the case of communicative rights, such effectiveness hinges upon reaching the audience. As a judge of the Supreme Court of California once said in a dissent (which upheld an anti-noise ordinance): "[W]ithout amplification so that it can be heard by others, freedom to use the human voice on the highways and thoroughfares is a freedom without value".18 And then, similarly, the same court in a decision taken en bane, made this remarkable and, by the standards of the American judiciary, highly unusual statement:

17. Police Department ofChicago v. Mosley, 408 U.S. 92, 96 (1972). 18. Haggerty v. Associated Farmers of California, Inc., 44 Cal. 2d 60, 72, 279 P.2d 734, 742 (l9SS) (Carter, J., dissenting).

SPEECH AND EQUALITY

81

To take the position that the right of free speech consists merely of the right to be free from censorship of the content rather than any protection of the means used, would, if carried to its logical conclusion, eliminate the right entirely. The right to speak one's views aloud, restricted by the ban that prevented anyone from listening, would frame a hollow right. Rather, freedom of speech entails communication; it contemplates effoctive communication. 19

On the other hand, we cannot reasonably demand an equal opportunity to the "effectiveness" of communication, in the sense of actually. converting other people to our views because, as noted above, so many factors that shape other people's views are not subject to equalization. However, we can plausibly demand an equal opportunity to get our points through to the audience which matters to us. This ideal translates itself into equal access to the media, printing houses, and public fora; into the "fairness doctrine"; into the right of reply, etc. These are all ways of equalizing an opportunity to be heard, and the United States Supreme Court is justifiably criticized by various commentators for neglecting these rights in its First Amendment interpretation.20 From this perspective, the words of Jerome Barron from thirty years ago that the American law "is indifferent to creating opportunities for expression,,21 seem to be still valid. In his influential paper, Barron lamented the absence of an affirmative responsibility by the quasi-monopoly media (such as newspapers in onenewspaper cities) to act as sounding boards for new ideas and old grievances. As one example of such an affirmative duty to expand expressive opportunities, one might consider the right of reply. The Supreme Court of the United States invalidated a statute which granted an individual, attacked by a newspaper, the right to have a response printed in that newspaper. The statute in question was described as "the compulsion exerted by the government on a newspaper to print that which it would not otherwise print".22 Strictly speaking, this is true, but this characterization ignores the crucial fact that, given the dominant position held by a paper in a particular market, the newspaper's decision not to publish a reply, may amount to a virtual denial of access for a particular view.

19. Wollam v. City of Palm Springs, 59 Cal. 2d 276, 284, 379 P.2d 481,486 (1963), emphasis added. 20. See, e.g., Eric Barendt, "Importing United States Free Speech Jurisprudence?", in Freedom of Communication ed. by Tom Campbell and Wojciech Sadurski (Aldershot: Dartmouth, 1994), pp.57-76. 21. Jerome A. Barron, "Access to the Press - A New First Amendment Right", Harvard Law Review 80 (1967): 1629-78, p. 1641. 22. Miami Herald Pub. Co. v. Tornillo, 418 U.S. 241,256 (1974).

82

CHAPTER 3

When stated in abstract and general tenns, as the Tornillo Court did in its opinion, a law "[c]ompelling editors or publishers to publish that which "'reason" tells them should not be published,,,23 indeed sounds damaging to freedom of speech. But when compared with the specific situation of a person criticized by a paper and then refused the right to reply, the "compulsion" in question seems much less hannful and the refusal to publish appears grossly unfair. The accompanying suggestion that the "[g]overnment-enforced right of access inescapably 'dampens the vigor and limits the variety of public debate",24 in this context sounds almost bizarre, since it is the refusal of a right to publish the opposite side of a given controversy which limits the diversity of viewpoints. Even if that suggestion is based on speculation that the imposition of a duty to publish replies would be a counter-incentive for publishers and editors to print controversial materials (because they "might well conclude that the safe course is to avoid controversy"),2S it is just that - a speculation. It seems at least equally plausible to speculate that replies (and other materials polemical with earlier materials) are attractive to a great number of readers and that, if anything, they will improve the circulation of the paper. Additionally, as a thoughtful American commentator observes: "The prospect that some regulated editors will forgo coverage of some political discussion because of reply requirements need not necessitate rejection of access regulation, for the benefits may still outweigh this cost".26 The American position on the right of reply contrasts with regulations in almost all European countries. For example, in Gennany, a newspaper or other periodical publication is obliged to publish a reply by any person who is the subject of an assertion of fact in the paper, subject to certain conditions. In some countries, such as France, the right of reply may be triggered by critical opinions as well as factual allegations. Most countries which do recognize a right of reply allow the offended party to seek a court order if the paper refuses to publish it.27 The right of reply is of course no substitute for a general right of access to the press, as Barron himself had noted,28 but it may serve as a useful device to equalize expressive opportunities. As another device, one might consider something equivalent to the "fairness doctrine" which has become defunct in the United States. It required a broadcast licensee to provide a balanced

23. Id., quoting Associated Press v. United States, 326 U.S. 1,20 n. 18 (1945).

24. Id. at 257, quoting New York Times Co. v. Sullivan, 376 U.S. 254, 279 (1964). 25. Id. at 257. 26. Lee C. Bollinger, Images of a Free Press (Chicago: The University of Chicago Press, 1991), p. 112. 27. See Sandra Coliver, "Comparative Analysis of Press Law in European and Other Democracies", in Press Law and Practice (London: An Article 19 Report, 1993), pp. 272-73. 28. Barron, p. 1660.

SPEECH AND EQUALITY

83

presentation of controversial ideas of public importance. It also required the licensees to provide a reasonable amount of time for the presentation of programs devoted to public issues, thus defeating the argument that the requirement of a ''balanced presentation" would induce licensees not to broadcast controversial programs. The doctrine went some way towards ensuring that no single viewpoint dominated the most pervasive broadcast media and so enhanced (even if insufficiently) opportunities for airing of traditionally underrepresented viewpoints. In upholding the doctrine, the Supreme Court held that "[i]t is the right of the viewers and listeners, not the right of the broadcasters, which is paramount. ... It is the right of the public to receive suitable access to social, political, esthetic, moral, and other ideas and experiences which is crucial here".29 Apart from content-related means of equalizing expressive opportunities, a legal system may develop various structural devices which are aimed at preventing the existence of monopolies in communication, and thus enhance the opportunities for those with lesser access to private media. The problem is that structural devices aimed at diversification and the pluralism of sources of information and communication are usually defended on the basis of the audience's interest in diversity rather than on the basis of the speakers' access to means of communication. For example, when the Supreme Court of the United States recently emphatically affirmed the goal of diversity in broadcasting, it stated: "Assuring that the public has access to a multiplicity of informational sources is a governmental purpose of the highest order, for it promotes values central to the First Amendment".3o But the separation between diversity as the audience's interest in plurality of communication and diversity as a matter of access for diverse communicators to media is somewhat artificial, and the two are closely interrelated. For example, in a 1996 decision, the U.S. Court of Appeals for the District of Columbia explained, while upholding "leased access" provisions (which compelled cable operators to set aside a certain number of their channels for commercial use by persons unaffiliated with the operator), that these provisions were necessary to ensure the widest possible diversity of information sources. It further explained that "diversity" in this context referred not to the substantive content of the program on a leased access channel, but to the entities (the "sources") responsible for making it available. 3! The interrelation between these two aspects of diversity was acknowledged by the Supreme Court when it endorsed the Federal Communications

29. Red Lion Broadcasting Co. v. F.C.C., 395 U.S. 367, 390 (1969). 30. Turner Broadcasting System. Inc. v. F.C.C., 512 U.S. 622, 662 (1994). 31. Time Warner Entertainment Co. v. F. C. c., 93 F.3d 957, 967 (1996).

CHAPTER 3

84

Commission's (FCC's) "theory that diversification of mass media ownership serves the public interest by promoting diversity of program and service viewpoints, as well as by preventing undue concentration of economic power".32 There is a clear link between the "diversification of ownership" and the "diversity of viewpoints" established by the FCC and by the Court: "the Commission acted rationally in finding that diversification of ownership would enhance the possibility of achieving greater diversity of viewpoints".33 The Commission's action was concerned with cross-media ownership, and the Court quoted with approval the FCC's judgment that "it is unrealistic to expect true diversity from a commonly owned station-newspaper combination. The divergency of their viewpoints cannot be expected to be the same as if they were antagonistically run".34 Restrictions on media concentration and, in particular, on cross-media ownership are well-established in most European legal systems. Courts in Italy, France and Germany have ruled not only that anti-trust regulations of commercial broadcasting are compatible with constitutional guarantees of freedom of expression, but even that these regulations are mandated by the requirements of plurality of opinion. The most famous of these rulings is perhaps a decision of the Conseil constitutionnel in France regarding the 1986 Broadcasting Bill.35 The Conseil decided that the principle of plurality of sources of information has constitutional status, and should be taken into account in designing anti-concentration rules. As a result, the Broadcasting Act of 30 September 1986 contains numerous guarantees against concentration and cross-media ownership. For example, no person may control more than 25% of the capital in a national broadcasting company, and no license for TV or radio can be issued if the holder would then find that he had interests in more than two kinds of media (radio, TV, cable and the press) with more than the prescribed coverage, etc. With the development of cable and satellite methods of transmission of programmes, the question of equalizing communicative opportunities has become all the more pressing' as the access to these means of communication may be restricted by the monopolistic position of carriers, and the monopoly may be due either to the enormous start-up costs or to the technical characteristics of the medium. One way of broadening access, and thus enhancing communicative opportunities, is by so-called "must carry" rules,

32. F.C.C. v. National Citizens Committee for Broadcasting, 436 U.S. 775, 780 (1978). 33. Id. at 796. 34. Id. at 797. 35. See Claude Leclercq, Libertis pub/iques (Paris: Editions Litec, 1991), p. 252; see also a

discussion of this decision in John Bell, French Constitutional Law (Oxford: Clarendon Press, 1992), p. 172-76; Eric Barendt, Broadcasting Law (Oxford: Clarendon Press, 1993), p.

127.

SPEECH AND EQUALITY

8S

which currently require cable systems in the United States to carry certain local commercial TV stations and noncommercial stations. The Supreme Court upheld the constitutionality of those rules on the basis that, among other things, they promote the widespread dissemination of information from a multiplicity of sources.36 In a more recent decision, the Court reaffirmed its support for "must carry" rules specifically referring to a "governmental purpose of the highest order in ensuring public access to a mUltiplicity of information sources".37 This, again, sounds like an emphasis on the diversity of programs in the interest of the audience, rather than equalizing the speech opportunities for programmers. However, in the same context, the Court found that "[m]ust-carry ensures that a number of local broadcasters retain cable carriage, with the concomitant audience access and advertising revenues needed to support a multiplicity of stations".38 There is, therefore, a direct link between the goal of protecting the audience's access to a plurality of sources and protecting speakers' access to a means of communication. The language of "equality of opportunity" seems to apply quite well to all the policies surveyed above and the aims served by them, because (1) being heard by one's audience is an important social good which many people seek, and (2) we may legitimately and meaningfully aim to reduce the impact of those factors which are outside the speaker's control and affect his access to the means of effective speech. Even in American law, which has been generally unwilling to translate the right of freedom of speech into rights of access to a means of effective speech, we can find traces of recognition that meaningful expressive rights must include some opportunity to actually reach one's intended audience. For example, a few years ago the Supreme Court considered the constitutionality of an injunction imposed upon anti-abortion protesters by a Florida court. The injunction placed several restrictions upon anti-abortion protesters, including a requirement that protesters refrain from physically approaching, within a certain area, any person seeking the services of the clinic. 39 The aim was to prevent so-called "sidewalk counselling" which, as one might suspect, was a rather nasty form of harassment of the patients and staff of the clinic by "pro-life" protesters. The Court decided that this particular provision could not stand because it was difficult "to justify a prohibition on all uninvited approaches of persons seeking the services of the clinic, regardless of how peaceful the contact may be, without burdening more speech than

36. 512 U.S. at 663.

37. Turner Broadcasting System Inc. v. F.C.C., 1997 U.S. LEXIS 2078, *19.

38. Id. at *59.

39. Madsen v. Women's Health Center. Inc., 512 U.S. 753, 129 L. Ed. 2d 593 (1994).

86

CHAPTER 3

necessary to prevent intimidation and to ensure access to the clinic".40 Putting to one side the question of evidence as to what extent such actions can be distinguished from "fighting words" or intimidation, the general idea behind this rejection of bans on uninvited approaches is that one has more than a mere right to speak; one should have an opportunity to be listened to. But, a critic might say, there is a great difference between saying thilt it would be good for people to have an opportunity to be heard by the audience, and granting them a right to be heard. The problem is, of course, that if we decide to design the contours of the proposed right beyond merely a right to speak, and extend it to a right to speak "effectively" (or, a ,right to be heard), then we are making inroads into the prospective listeners' right not to listen, and to be left alone. As a dissenting judge in the Canadian Supreme Court's decision on freedom of expression suggested: "Freedom of expression guarantees the right to loose one's ideas on the world; it does not guarantee the right to be listened to or to be believed".41 According to some First Amendment theorists, the right not to listen is as important, and based on the same grounds, as the right to speak: Charles Black, for one, argued that "the claim to freedom from unwanted speech rests on grounds of high policy and on convictions of human dignity similar to if not identical with those classically brought forward in support of freedom of speech".42 Another influential scholar, Thomas Emerson, thought it to be "a cardinal principle of the system of freedom of expression that no person can be compelled to listen against his will".43 The latter principle seems strong and convincing, though to claim that it is based on the same values as freedom of speech is question-begging unless one shows how freedom from unwanted speech promotes those values which freedom of speech is expected to promote. The most 'obvious candidate here is personal autonomy; however, as suggested in Chapter 1.2, the value of autonomy cannot figure very prominently in the justification for freedom of speech. But then, it is more difficult to show that freedom from unwanted speech is directly instrumental to personal selffulfilment, democratic self-government, or promoting the virtue of tolerance. This is not to say that freedom not to listen is trivial but rather that its grounds may be different from those which support freedom of speech, and so the ranking of those freedoms, in the case of conflict, may not be impossible.

40. 129 L. Ed. at 613. 41. R. v. Keegstra, (1991)2 W.W.R. 1,107 (McLachlin, J., dissenting). 42. Charles Black, "He Cannot Choose but Hear: The Plight of the Captive Auditor", Columbia Law Review 53 (1953): 960-74, p. 967. See, similarly, G. Michael Taylor, "'I'll Defend To Death Your Right to Say It ... But Not to Me' - The captive Audience Corollary to the First Amendment", Southern Illinois University Law Journal (1983): 211-26, p. 216. 43. Thomas Emerson, The System of Freedom of Expression (New York: Random House, 1970), p.710.

SPEECH AND EQUALITY

87

In any event, by positing a right to speak effectively we run the risk of offending the principle that no-one can be compelled to listen against his or her will. We face here what may be called a buttonholer problem: the familiar imagery will be of a mendacious talker who detains us in conversation even after we have made it very clear to him that we want to be let alone. The parade of horribles can be extended: Justice Black once painted a very frightening picture of a situation in which "[c]hurches would be compelled to welcome into their buildings invaders who came but to scoff and jeer" and "homes, the sacred retreat to which families repair for their privacy and their daily way of living, would have to have their doors thrown open to all who desired to convert the occupants to new views, new morals, and a new way of life".44 How can we reconcile the idea of a right to be heard with the intuitively persuasive listeners' rights to have full control over what they want to hear and whom they want to listen to? To be sure, it makes sense to talk about a qualified moral duty to listen to those communications which concern us or which concern the matters on which we make public pronouncements, and the neglect of this moral duty may be morally penalized by criticism from others.45 But this is not an independent duty to listen as it derives from a more general moral duty to speak competently on matters of public interest, and a consideration of other people's views is part of what constitutes "competence". It is important to realize that the putative right to be heard, considered here, is a general right. It is a right to be heard in public discourse in a society, rather than a right to be heard in some special, particular context, such as the right of a patient to have her opinions about therapy considered by the doctor, or a right of parents to be heard regarding the contents of children's school education. These particularized rights to be heard are justified on the basis of specific purposes of certain institutional (and other) contexts, and have nothing to do with the general good of speaking and being heard in public discourse. As a parallel, there may be specific, particular circumstances when a person has a right not to be spoken to, which do not derive from a general, putative right not to listen but rather from some other individual interests which, in a particular context, can only be observed if a person is not talked to against her will (for example, because she wants to sleep, or to read in a library, or to engage in transcendental meditation, etc). What concerns us here is whether there is a general right not to listen which would extinguish this part of a general right to speak which consists of actually being heard by one's audience.

44. Gregory v. Chicago, 394 U.S. 111, 125 (1969). 45. See J.P.Day, '''The Duty of Listening"', Philosophy 71 (1996): 461-64.

88

CHAPTER 3

As far as this latter, general right is concerned, one might perhaps expect that a useful way of reconstructing, and supporting, such a right would be along the lines of the well-known principle that one has a right (and can claim state protection for such a right) to speak to an unwilling or a hostile audience, and that the hostile audience must not be permitted to affect the right to speak.46 As a typical articulation of the doctrine, consider this statement from a decision of the Supreme Court of India: "Freedom of expression which is legitimate and constitutionally protected cannot be held to ransom by an intolerant group of people".47 Or the statement from a decision by the Supreme Court of the United States: "As a general matter, we have indicated that in public debate our own citizens must tolerate insulting, and even outrageous, speech in order to provide breathing space to the freedoms protected by the First Amendment".48 This statement is taken from a decision which struck down a statute which made it unlawful to display signs which would bring a foreign government into "public disrepute" within a certain distance of a foreign embassy. But a closer reading of the decision suggests that the "hostile audience" doctrine does not really follow from a recognition of the right to speak effectively, or to be listened to. Rather, it is based on a prudential reasoning that if hostility of the audience towards the speaker results in a veto of a speech, no right to speak with the purpose of convincing one's audience will survive because the necessary ''breathing space", to which the Boos Court referred, will be eroded. It is not so much that hecklers may render your speech ineffective, but rather that they may prevent you from speaking in the fIrst place: freedom of expression (and not an opportunity to be heard) will be "held to ransom". The "breathing space" argument is based on a calculus of probability of the gravity of two possible harms: the harm of imposing speech upon unwilling listeners (and thus violating their control over what they want to hear) and the harm of preventing the speaker from speaking (and thus violating her or his right to speak). The hostile audience principle is based on an ascertainment that, typically, in the conflict between the two, the latter harm is greater. That is why, in a situation of conflict, it is the hostile (or unwilling) audience which carries the burden of avoiding continued exposure to the speech, rather than the speaker having a duty to desist. Whether this determination of comparative harms is plausible or not, it does not reach the substance of a speaker's right to actually be heard by the audience of her choosing. So perhaps, in the end no strong right "to be listened to" can be established because, to be viable, it would impose obligations upon members of the 46. See David G. Barnum, "Freedom of Assembly and the Hostile Audience in Anglo-American Law", American Journal of Comparative Law 29 (1981): 59-96. 47. The Secretary, Ministry of Information v. Cricket Association of Bengal, IT 1995 (2) S.C., 110, 132. 48. Boos v. Barry, 485 U.S. 312, 322 (1988).

SPEECH AND EQUALITY

89

audience that would clearly be unreasonable. While we (as members of the audience) have a duty to tolerate uninvited and unwelcome speech in public, we have no duty to listen (much less, to listen attentively) to it. But consider this: to make a decision about not listening meaningful, we should at least give the speaker a chance to win our attention; we should perhaps listen to the first words before we look the other way, leave the room, say "Enough of this nonsense!" or switch the radio off. As Justice White once said: "The First Amendment protects the right of every citizen to 'reach the minds of willing listeners and to do so there must be opportunity to win their attention",.49 It is ironic that this maxim accompanied the Supreme Court's opinion which rejected the claim of Krishna followers t
49. Heffron v.lnternational Soc'yfor Krishna Consc., 452 U.S. 640, 655 (1981), quoting Kovacs v. Cooper, 336 U.S. 77, 87 (1949). 50. Franklyn S. Haiman, "Speech v. Privacy: Is There a Right Not to Be Spoken To?", Northwestern University Law Review 67 (1972): 153-99, p. 193. 51. Van Nuys Publishing Co. v. City of Thousand Oaks, 5 Cal. 3d 819, 827 (1971).

90

CHAPTER 3

democratic countries whose general commitment to freedom of speech cannot be plausibly doubted, the law protects the rights of residents against door-todoor solicitation: no-one is compelled to let the Jehovah's Witnesses or salespersons in and engage in their speech. In most legal systems there are laws which protect people against threatening and nuisance phone calls, and which enable people to demand that unwanted mail is not sent to them. Everywhere there are regulations which control the use of public address systems, sound trucks and billboards. Finally, as the Madsen decision mentioned above indicated, it also seems proper to impose strict restrictions upon residential pickets (as in Madsen: pickets by pro-life protesters outside doctors' homes) or pickets outside hospitals where patients may be disturbed. This list is, of course, far from comprehensive, and in various legal systems different regulations to the situations listed here may apply. But all that matters is that these examples are unlikely to cause any strong and principled objections on the grounds of freedom of speech, and yet they seem to suggest that the law takes the side of the audience rather than that of the speaker in cases where rights conflict. But this conclusion does not necessarily follow from these examples. In all these cases, something more than merely a right to be free from unwanted communication argues for the listeners' rights. In the case of door-todoor solicitation, it was the privacy of one's home because, fairly enough, "the home [is] the one place where people ordinarily have the right not to be assaulted by uninvited and offensive sights and sounds".52 In the case of nuisance calls, it was a right not to be put in a state of fear. In the case of loudspeakers and billboards, it was protection from excessive noise and respect for the aesthetic values of the environment and for road safety. And in the case of restrictions upon picketing outside a hospital, it was a concern for the psychic and mental well-being of people in a particularly vulnerable condition. Some of the restrictions on communicative rights in this list can be analogized to laws against stalking; quite apart from the fact that these laws are themselves often controversial (for instance, how do we define "stalking" in a way which does not impose excessive burdens upon one's freedom of harmless action; how can we measure the harm of stalking itself in a way which will justify penalizing a stalker before he inflicts injury on the victim?),53 it seems that if a particular action, communicative or otherwise, causes fear, intimidates or harasses other people, then it can be subject to legal restraint. The regulations against nuisance

52. F.C.C. v. Pacifica Foundation, 438 U.S. 726, 759 (1978) (Powell, J., concurring). The principle stated by Justice Powell is persuasive but, strangely enough, it was applied to uphold a sanction against offensive radio broadcast. All that the homeowner has to do in order to avoid further offence is to turn off the radio - an action eminently efficient and easy. 53. See, e.g., Christine Olle Sloan, "Standing Up to Stalkers: South Carolina's Antistalking Law Is a Good First Step", South Carolina Law Review 45 (1994): 383-427.

SPEECH AND EQUALITY

91

phone calls and threatening calls, or against residential picketing, may be best explained by the protection of people against intimidation. Some cases are unconvincing: should a resident really have a right to stop the flow of unwanted mail into her mailbox?54 The cost of avoiding receiving a communication in this case is minimal. All it takes is to throw unwanted mail into the rubbish bin. In this case, the balancing of the rights of a communicator to try to win the attention of the audience, and the rights of the audience not to be disturbed at home, seems to argue in favour of the former. Similarly in the case of door-todoor solicitation, should residents really have a right to forbid such a solicitation altogether, so that the solicitors are not even allowed to try to make an initial contact?55 Regardless of the results of this specific weighing and balancing of respective interests, what matters for the argument here is that in all the cases listed above, the right not to listen was insufficient to explain the regulation and was amplified by other important rights: residential privacy, freedom from intimidation, etc. This is a weak proposition; a strong proposition (which seems plausible but is not necessary for the argument here) is that it only those other rights, rather than an alleged right not to listen to an unwanted communicator, fully explain the legal regulations. Be that as it may, one thing that these examples do indicate is that the listener's rights to be left alone do not prevail in an absolute manner, but that, even if the law is biased towards unwilling listeners rather than unwanted speakers, the balancing of various rights is a matter of degree. A single call from an irritating speaker is not something that can be prohibited or punished: at what point do the calls from the same person with an increased degree of rudeness or offensiveness or intimidating elements become liable for legal sanction? At what point does the noise from loudspeakers become excessive? When does the following of another person on the street cease being a legal, if unpleasant, innocent action and become "stalking"? In these cases the law must make subtle judgments of degree, about which people may disagree even if they

54. See Rowan v. United States Post Office Dep't, 397 U.S. 728 (1970). In this decision the

55.

Supreme Court upheld a statute allowing a householder to stop unsolicited mail from being sent to him. The Court even analogized the mailing of unwanted materials to a form of trespass, id. at 737, and asserted that if the statute's prohibition "operates to impede the flow of even valid ideas, the answer is that no one has a right to press even 'good' ideas on an unwilling recipient", id. at 738. For indications by the Supreme Court that such blanket bans would be constitutional in the United States, see Village a/Schaumburg v. Citizens/or a Better Environment, 444 U.S. 620, 639 (1980); Martin v. City a/Struthers, 319 U.S. 141, 148 (1943). But see Hynes v. Oradell, 425 U.S. 610 (1976) where the Supreme Court invalidated, for vagueness, an ordinance which required persons desiring to solicit door-to-door to notify the police in writing.

92

CHAPTER 3

agree that "stalking" or "nuisance calls" should be penalized. There is nothing surprising about this. But it indicates that, at a lower level of harm to the unwilling listener, when it is no more than a passing irritation, the speaker's right to try to engage her audience in the communicative interaction may prevail. Another conclusion from the examples listed above is that a response to the conflict of rights between those of unwilling listeners and unwanted speakers greatly depends on the type of action on the part of the listener which is needed to refuse the communication. The harder it is to extract oneself from the unwanted communicator, the higher the degree of protection the law should afford to the members of the audience. The corollary of this is that if the effort needed to avoid communication is trivial: hanging up, chucking out the leaflet, looking the other way, the law may give the prospective speakers a relatively higher level of protection. This resonates with the "captive audience" doctrine in First Amendment theory: a doctrine which says that "captive" listeners should be protected from forced exposure to speech within a physically confined environment.56 Of course, the "captive audience" proposition cannot be taken too literally. As some commentators suggested, we are very rarely really "captive", and almost always an unwilling listener has some alternative means to avoid the message. 57 While for some it may simply mean that it is very rare that the law should take the side of unwilling listeners, a more appropriate conclusion would be that what matters is a rationale for having the "captive audience" doctrine in the first place. The doctrine is mainly based on respect for individual choice: if someone is put in circumstances where he or she cannot choose whom to listen to, or the choice incurs very high personal costs, then the law should support the audience against the speaker. But if the members of the audience can freely choose not to listen to the speaker, there is no reason for the law to interfere with the speaker-listener interaction and provide its own support for the listener. From this point of view, "captivity" is not so much a matter of "confinement" in a physical space as it is a difficulty of avoiding the continued message. Most typically though, there are of course cases where physical confmement and difficulty in avoiding the message overlap, as in cases of targeted residential picketing where the occupants of a residence or household cannot readily move to another residence or household in order to avoid the unwelcome picketing activity being directed at them. But there may also be cases, such as the example of a person pursued by an irritating "sidewalk counsellor" in the street, where a target of speech is not "captive" in the sense of spatial confinement, but

56. See, e.g., Frisby v. Schultz, 487 U.S. 474,487 (1988).

57. See, e.g., Marcy Strauss, "Redefining the Captive Audience Doctrine", Hastings Constitutional Law Quarterly 19 (1991): 85-121, p. 86.

SPEECH AND EQUALITY

93

cannot avoid the unwanted continued communication without a major difficulty or cost. This distinction explains a difference between the majority and a dissenting opinion in a classical captive-audience case, Lehman v. City of Shaker Heights. 58 The majority found that political advertising in the advertising space on rapid transit cars could be refused by a city because "the streetcar audience is a captive audience";59 Justice Brennan, in his dissent, denied that the audience was powerless to avoid the speech: they could simply avert their eyes.60 One of the commentators suggested that the Lehman Court "confused captivity in a place with being captive to speech" and claimed, in agreement with Brennan's dissent, that "[e]ven though individuals were not 'free' to leave the bus, they could still avert their eyes from the ad and look elsewhere".61 This may be true, though perhaps rather artificially so; written messages in places where we do not expect them (that is, which are typically not fora for communication) in a sense ''trap'' us and the reality of a lengthy journey on a train or a bus may be such that the tired travellers stare, willy nilly, at the advertising messages and it takes a lot of force of will to avert one's eyes. So there may be no such great difference, after all, between messages broadcast over loudspeakers (which, as Brennan says, are much more difficult to avoid) and written advertisements. Undoubtedly, judgments of this sort are highly context-dependent and controversial; ultimately, the degree of "captivity" is a matter of empirical evidence about the actual capacity to avoid speech in a given context and a given situation. The problem of avoidability of speech points to one way in which speech can be called "coercive". The coerciveness of speech may consist of the fact that, unless the listener complies with the speaker's views or demands, he will suffer negative consequences or will be deprived of certain benefits. When, for example, an employer expresses certain views about unionization, and the employees fear that unless they comply with the anti-union views of the employer they will face reprisals, then the speech can be characterized as coercive.62 In order to avoid a major setback to their interests, the listeners have to act in a way desired by the speaker and in a way in which they would not have acted in the absence of the speech. The coerciveness of that type of speech

58. 59. 60.. 61. 62.

418 U.S. 298 (1974). Id. at 301-02. Id. at 320-01. Strauss, p. 98, footnote omitted. In the United States, a federal statute which defines unfair labor practices, explicitly restricts an employer's First Amendment right to express certain views due to the coercive effect of threatened reprisals, 29 U.S.c. para. 158 ( c) (1988); see also NLRB v. Gissel Packing Co., 395 U.S. 575 (1969).

94

CHAPTER 3

is not relevant to our present discussion because it is not the fact that the listeners do not want to listen, but rather the fact they do not want to act in accordance with the speech, that constitutes a condition of its coerciveness. The speech is coercive because it announces that its addressees now face a new alternative which did not exist before the speech: either that they comply with the speaker's desired option (something that, otherwise, they would rather not do) or they pay a price which may be even more costly than compliance. 63 They can easily avoid the speech but this is no consolation to them: as long as they know what the reprisals will be, they are put in a coercive situation. But there is another way in which speech may be characterized as "coercive", that is when, in order to avoid the speech itself, the listeners have to incur major costs and act in a way in which they would not have acted in the absence of the speech. The difference from the first form of coerciveness is that here there is no question of acting in accordance with the speaker's preferences: the speech itself compels the unfortunate listener to. act in a way which is unwanted and sets back his interests. The action to avoid the continuous speech is costly, but the cost is judged by the listeners as lower than the cost of receiving the continued speech. As an example, one may think of a picket outside one's residence (in order to avoid running a gauntlet and confronting picketers one avoids leaving one's home) or an anti-abortion picket outside a clinic (in order to avoid the upsetting communication, one abandons one's plan of going to the clinic). In the former example, the main aim of the speakers is not "to disseminate a message to the general public, but to intrude upon the targeted resident". 64 In the latter case, the purpose is not only to communicate the message to the staff and would-be patients, but also to force them not to enter the clinic. This coerciveness of the speech reduces the speaker's claim to an opportunity to speak effectively and the listener's duty not to interfere with the speech. Is the distinction between the two senses of "coerciveness" of speech, suggested in the preceding paragraph, meaningful? After all, one may claim that what the coercive speaker wants to achieve in the second case (where the unavoidability of the speech creates a coercive situation for the listener) is to get her to act in a particular way. For example, by picketing, the pro-life protesters want to compel doctors to stop administering abortions and to force women to stop attending abortion clinics. But the analogy is incomplete. To begin with, there are forms of "coercive" communication which merely aim to harass the unwilling listeners rather to get them to act in accordance with the speaker's

63. This is consistent with Joel Feinberg's concept of coercive threats, see Joel Feinberg, Harm to Self: The Moral Limits of the Criminal Law, vol. 3 (New York: Oxford University Press, 1986), p. 192. 64. Frisby v. Schultz, 487 U.S. 474, 486 (1988).

SPEECH AND EQUALITY

95

wishes. A paradigm example would be that of harassing telephone calls which are not aimed at persuading the listener to do anything: they are aimed purely at creating distress or fear. In order to avoid further harassment, the target of the calls may decide to disconnect her phone, or to get a silent number. Such actions impose certain costs and inconvenience, but are considered to be less costly than being exposed to further calls of this type. Second, even if the coercive speaker aims to compel the hearer to act in a particular way, there is a difference between speech which merely informs the listener about the sanctions for not acting in a particular way (and in certain circumstances it may become a coercive threat, such as when an employer warns its employees that those who join the union will be sacked) and, on the other hand. speech which constitutes a nuisance in itself and the only sanction for not complying with the speaker's demand is that the speech will continue; residential picketing is a paradigmatic example of this form of "coercive" speech. The importance of the distinction is that the second, but not the first, type of coercive speech shares with the captive audience situation an element of difficulty in avoiding the unwanted speech, and thus evokes the value of choice by the listener of what and whom she wants to listen to . . Finally, while the value of individual choice which supports the right not to listen to unwanted speech is very high, it is not absolute. An example is when it is in conflict with the value of broad communication among members of society. Furthermore, for the choice to be rational it must be based on knowledge about what is being rejected. This brings back the idea of protecting the speaker's right to try to engage a listener in communication - at least at an initial stage. In a carefully argued article, Steven D. Smith showed that on the basis of the "tolerance" theory of free speech (contrasted, in his article, with "marketplace" theory), there is nothing paradoxical about individual listeners having a private power to "suppress" communication: "it is reasonable to confer upon individuals the power to decide for themselves after an initial exposure that particular statements, even statements addressing subjects of significance, are without substantial value and are therefore not worthy of further hearing or consideration".65 The italicized words are central to my argument: there must be at least "an initial exposure" in order to make the rejection of further communication meaningfuL But even such a right to speak to an unwilling audience, although limited, as it is, to the "first shot" opportunity, will often be too much of a burden upon the unwilling listener to prevail over his or her right. Two categories of exceptions 65. Steven D. Smith, "Skepticism, Tolerance, and Truth in the Theory of Free Expression", Southern California Law Review 60 (1987): 649-731, p. 728, italics added.

96

CHAPTER 3

come easily to one's mind and, no doubt, there are more. One is when even the initial communication is sufficient to impose a burden or a cost upon a listener which the law is expected to protect the listener against. Consider a harassing telephone call: the fact that the listener can hang up and terminate the conversation is no remedy: the harm has already been inflicted, and the fear has been instilled in the listener's mind. Or if someone is deeply shocked by mous types of offensive pictures, then even an initial, unanticipated exposure to this form of expression may be extremely distressing to the reader, and a quick abandoning of further reading will not erase the distress from his mind. This observation accompanied the United States Supreme Court's.decision to uphold a restriction on offensive radio broadcasts: "to say that one may avoid further offense by turning off the radio when he hears indecent language is like saying that the remedy for an assault is to run away after the first blow".66 But the force of this conclusion derives from the force of the prohibition of a given category of speech in the first place. If we antecedently decide that a given category of speech is proscribable, then, naturally, there is no point in insisting on the right to initiate a conversation where a speaker resorts to proscribable speech. If, as in the Chaplinsky formula, there are certain categories of speech which "by their very utterance inflict injury",67 then the very utterance is synonymous with the infliction of a harm which the law can rightly target, regardless of whether the targeted expression is initial or continued. A second obvious exception to the duty to allow an initial expression, is when the initial expression, regardless of its contents, collides with the important interests of listeners, such as an interest in tranquillity or repose. 68 In contrast to the former exception, where the harm of speech consisted of its proscribable contents, this exception is content-independent. If I want to sleep, and have a right to rest in the privacy of my home or wherever the privacy of a person is taken to be of higher importance than in public spaces, I do not want to be interrupted during my rest by any initial communication whatsoever. Not only do I not have a duty to let the speaker complete the initial stage of his or her communication, but I have a justified claim against allowing the speaker to even try to initiate the communication. This claim can be analogized to a claim against noisy loudspeakers near one's home or office: the interest in tranquillity at certain places and at some times absolutely argues against all communications addressed to us, but most of the time argues against any communications which invade our tranquillity and peace in a particularly drastic way.

66. F.C.C. v. PacifICa Foundation, 438 U.S. 726, 748-49 (1978). 67. Chaplinslcy v. New Hampshire, 315 U.S. 568, 572 (1942) 68. On the right to repose as one possible ground for a right to be free from unwanted speech, see Strauss, pp. 111-14.

SPEECH AND EQUALITY

97

However, these are exceptions, and a rule seems to be that we need to give a speaker at least one shot at her chosen audience, to let the buttonholer at least begin the conversation, even if we will not let him continue interminably. It is not clear how to translate these vague metaphors into a general principle which can be legally operational. It is easier to think about turning this general idea into reality in very specific contexts: for example, those who hand out leaflets in public places have a right to approach any passer-by and commence an explanation of their positions (or offer a leaflet), but they should desist when clearly asked to by a passer-by. The specific design of such rules in particular circumstances is certainly well beyond my purpose here. But the general interpretation of the opportunity to speak effectively (i.e., to speak and be heard) should be clear by now: I have no right to demand that you listen attentively to my arguments, but I should have a right to try to win your attention, rather than be confined to a space from which my voice will not even reach you. As an example of such a "confinement", in both a figurative and literal sense of the word, consider a report by Kathleen M. Sullivan who "once helped Professor Laurence H. Tribe argue [before the Supreme Court] that devotees of Hare Krishna had a First Amendment right to proselytize in the open spaces of the Minnesota State Fair and not be confined to a rented booth".69 As she explains: "We argued that Minnesotans would flock to the booths of the Methodists, Presbyterians, or Episcopalians, but that Hare Krishna, if confined to a booth waiting for customers, would have a long and very quiet day.,,7o A right to proselytize means that there is more than just a right to speak, or just a right to speak to someone who actually volunteers as a listener. It is, as it were, a right of going out of your way, and actively seeking the attention of your potential audience by approaching people and trying to engage them in conversation at a state fair (as in Sullivan's example), at airports, in the streets, etc. Buttonholers can be restrained when they become an actual nuisance, preventing us from doing what we want in tranquillity and peace, notwithstanding our clear signal that we no longer want to listen to their message. However, it does not follow that we have a general right not to be approached, in public fora, and a right not to be talked to. The record of American law with respect to a general right to actively seek the attention of potential hearers is ambiguous and confusing, and many restrictions have been introduced by narrowing down the notion of the "public" 69. Kathleen M. Sullivan, "Free Speech Wars", (1994). 70. Id. p. 208.

SMU Law Review 48 (1994): 203-214, p. 208

98

CHAPTER 3

forum. For example, this is indicated by denying the characterization as a public forum, for First Amendment purposes, of government property such as military bases, sidewalks in front of post offices, and airport terminals.71 Occasionally courts have found that various restrictions on public speech, such as restrictions on picketing, are reasonable ways of serving important social interests without, at the same time, imposing an excessive burden upon free speech, especially if alternative channels of communication are available; as an example one may cite a Supreme Court decision which upheld a town ordinance which made it unlawful to picket in front of the residence of any individual.72 The judiciary have often displayed hostility towards speakers trying to amplify (literally or metaphorically) their voices in public in order to reach unwilling or hostile audiences. Distributors of pamphlets and handbills have even been compared to visitors who "insert a foot in the door and insist on a hearing".73 However, my aim here is not to support my discussion of a right to effective speech by arguing that it is unequivocally recognized in United States law, and, in any event, this right is not absolute. A demarcation between this right and the rights of the members of the audience to a degree of tranquillity is always a controversial matter. But the point that I have tried to establish here is that a right to speak effectively, in a sense which goes beyond a right to speak, and which extends to an opportunity to win the attention of the audience, is embedded both in our popular intuitions and in legal thinking about public expression.

2

Silencing

The upshot of the previous section is that if the ideal of equality of opportunity is applied to speech, the best way of characterizing the opportunity in question is as an opportunity to speak effectively, or to speak and be heard. We have seen various ways in which this opportunity can be implemented and equalized, by various rights of access and by limiting audience rights to prevent unwanted speech. All these ways of equalizing speech opportunities are about enhancing the opportunities of speakers, either because without the enhancement, their access to the communicative fora is insufficient, or because their opportunity to be heard becomes too vulnerable if the potential listeners have unlimited control over what is being communicated to them.

71. See, respectively, Greer v. Spock, 424 U.S. 828 (1976), United States v. Kokinda, 497 U.S. 720 (1990), and International Soc'y for Krishna Consciousness v. Lee, 505 U.S. 672 (1992). On ''public forum" see, generally, Robert Post, "Between Governance and Management: The History and Theory of the Public Forum", UCLA Law Review 34 (1987): 1713-1835. 72. Frisby v. Schultz, 487 U.S. 474 (1988). 73. Kovacs v. Cooper, 336 U.S. 77, 86 (1948).

SPEECH AND EQUALITY

99

But the central question, from the point of view of recent discussions about the equality of communicative opportunities, is whether restricting some speakers can be seen as a way of equalizing speech opportunities for other speakers. Can bans on hate speech or on pornography be seen as a way of equalizing the opportunity to be heard? The substance of the proposals most famously put forward by Catharine MacKinnon is that bans on pornography arid hate speech are necessary in order to equalize speech opportunities. When she, for example, praises "the hate propaganda prohibition and the obscenity law of Canada (properly interpreted)" as "equality laws",74 then it is clear that equality here is seen to result from imposing restrictions upon some speakers, and that these restrictions upon some speakers are seen as resulting in enhanced speech opportunities for the others. How can this be? For such an interpretation to be tenable it would have to maintain that what hate-speech bans or pornography restrictions do is remove from certain people some opportunities to exercise their capacity to speak and be heard (namely, to make racist comments or produce pornography), in order to improve the availability of similar opportunities to others. But offhand it seems that these speech bans do not equalize this particular opportunity. A vilified minority member's opportunity to speak and be heard on public issues is not increased, it would seem, by removing a racist's opportunity to be heard. Something else is arguably protected; namely, a minority member's sense of self worth, dignity, psychic peace, tranquillity, etc. However, promoting equality of opportunity in the realm of speech is not about removing one sort of good from one group in order to protect another sort of good attributable to another group. It was recently suggested in American constitutional theory that restrictions on racist hate speech may be seen as a form of "affirmative action", akin to the elimination of systemic discrimination by positive means. For instance, Kathleen M. Sullivan suggested that the Supreme Court could find that "hate speech" is unprotected by the First Amendment on the basis "that a compelling interest in eradicating racial hierarchy justifies the elimination of targeted expressions of race hatred - a kind of First Amendment 'affirmative action'

74. MacKinnon, p. 103.

100

CHAPTER 3

doctrine".7S Some other writers, too, have analogized such measures to race-

conscious affirmative action programs.76 The standard argument in favour of "affirmative action" regarding such goods as educational and employment opportunities (that is, in favour of paradigm cases of affirmative action), when framed in terms of promoting a genuine, as opposed to a purely formal, equality of opportunity, runs as follows: (1) in a situation of actual inequalities which may be traced back to past discrimination, "formal" equality of treatment (understood as applying the same criteria of selection to all the applicants) does not assure genuine equality of opportunity to attain a desired good; (2) positive measures aimed at equalizing genuine opportunity for minority members require removing some opportunities from non-minority members; (3) this is not unfair because this "surplus" in opportunities, enjoyed by the non-minority members (and subject to removal through the affirmative action), would not have existed had there been genuine equality of opportunity in the first place. This is, of course, a very simplified outline of an equal-opportunity based case for affirmative action. But this is all that is needed for our purposes. I do not find this argument, as applied to affirmative action in employment or education, objectionable in principle; indeed, elsewhere I have expressed my endorsement of the principle and suggested that a proper way of disputing affirmative-action claims is with regard to factual bases underlying the three above-mentioned steps in the specific contexts of particular programs,?7 I am mentioning it here to make it clear that I have no concerns about the very idea of "strong" affirmative action per se, as it seems to me to be a compelling one. But I would dispute the applicability of this idea to the area of freedom of expression precisely because the above three steps, which I believe are crucial to the validity of affirmative action, cannot be utilized in the field of public expression. Step (1) and step (3) may often be valid with regard to effective speech opportunities, that is, opportunities to speak and be heard. Step (1) is validated by an observation that underprivileged groups, suffering present effects of past discrimination, have disproportionately low access to public media and other fora for public .expression. Step (3) relies upon a counterfactual that, were it not for past discrimination, today's privileged social classes would have less access to effective means of public speech than they actually have. However, step (2), which is central for the argument, does not apply to the realm of public speech: reducing effective speech opportunities for advantaged

75. Kathleen M. Sullivan, "The Supreme Court, 1991 Term - Foreword: The Justices of Rules and Standards", Harvard Law Review 106 (1992), 22-123, p. 42, footnote omitted, emphasis added. 76. Akhil Reed Amar, "The Case of the Missing Amendments: R.A.V. v. City of St. Paul", Harvard Law Review 106 (1992): 124-61, p. 159. 77. See Wojciech Sadurski, Giving Desert Its Due (Dordrecht: D. Reidel, 1985), ch.7.

SPEECH AND EQUALITY

101

groups, in the form of removing legal protection for racist or pornographic speech, does not result in increasing effective speech opportunities for the disadvantaged. If step (2) applied to the promotion of access rights, such as those discussed in the previous section, the connection between reducing the access rights of privileged groups and improving the access rights of disadvantaged groups could be established. But if step (2) applies to such a reduction of speech opportunities, which consists of a legal prohibition on some types of speech, then the improvement of speech opportunities for the beneficiaries of the restriction is very tenuous at best. This last proposition can be, and has been, challenged. Indeed, some proponents of hate speech regulations suggest that racists' exercise of their right to make racist comments deprives their victims of their right to free speech, and similarly, some proponents of anti-pornography restrictions argue that pornography is a form of silencing women. In this regard, the argument for the restrictions on hate speech parallels the anti-pornography argument. The silencing argument can be understood in two ways, both of which raise different types of problems. First, it may be said that the unrestricted presence of hate messages actually prevents minority members from publicly expressing their views, or at least restricts their opportunities to do so. Second, it may be said that it devalues their views by demeaning and denigrating the speakers, thus reducing the value attached to such views by the public. I will consider these two arguments in turn. The first interpretation of the "silencing argument" (silencing through intimidation) seems to rely upon a set of empirical conditions that mayor may not support the proposition in particular circumstances, but which make it much more contingent than its proponents would be willing to admit. As an example of such silencing, Mari Matsuda reports that "[i]n order to avoid receiving hate messages, victims have had to ... curtail their own exercise of speech rights".78 Similarly, Cass Sunstein observes that "[p]eople confronted by hate speech may experience a form of 'silencing' in the sense that they are reluctant to speak".19 However, this effect seems to arise more often in the case of face-to-face verbal assaults (or "fighting words") which should properly be restricted, than in the dissemination of general messages through the media, public speeches and the like, where the curtailment of one's speech rights is of no help in "avoiding receiving hate messages". It is not clear how the surrendering of one's right to

78. Marl J. Matsuda, "Public Response to Racist Speech: Considering the Victim's Story", Michigan Law Review 87 (1989): 2320-81, p. 2337. 79. Cass R. Sunstein, Democracy and the Problem ofFree Speech (New York: Free Press, 1983), p.186.

102

CHAPTER 3

speak in public might protect one against racist messages delivered by others. Surely, one can limit one's general liberty of action when the threat of encountering disturbing messages is very great. For instance, to avoid threatening messages one can stop leaving one's home, listening to radio and TV, reading newspapers, etc. But in such extreme cases the reduction of freedom of speech is derived from a general loss of liberty simpliciter. Only an extremely massive, all-pervasive racist campaign against a given minority can be said to produce such effects, and these are the circumstances in which bans on hate speech would be eminently justified. However, in a society so contaminated by racism it is structural racism, rather than particular racist expressions, which produce the silencing effect. As Robert Post says, if indeed "such silencing occurs chiefly through the structural conditions of racism, rather than specifically through the shock of racist speech" then "restraints on racist speech would impair public discourse without at the same time repairing the silence of victim groups".80 It may be claimed, nevertheless, that even if one is not physically prevented from speaking and being heard, one has a substantial disincentive to speak if one is subject to vilifying statements made by others. Conversely, it may be claimed that one has all the more reason and motive to speak if one has to defend one's own group. A set of complex empirical factors affects this outcome, through individual predispositions and, more importantly, depending on how widespread the vilifying comments are. Arguably, in a society flooded by hate messages against a particular minority, there emerges a critical mass which may prevent one from making a case in one's defence. However, the very presence of hate messages, especially at a relatively low level of frequency, cannot be said to prevent one from speaking or even to create a significant disincentive to speak. The reasonableness of prohibitions on racist propaganda therefore seems to depend significantly upon the overall volume of racist messages in the society. The second version of the "silencing argument" (silencing by demeaning the speakers) raises a different problem. Hate messages may adversely affect the credence given by an audience to the views expressed by members of groups who are stigmatized by racist beliefs. Whether and how often it really happens is an empirical matter, and one would hope that the presence of racist views often has the opposite effect to that desired by racists, that is, it may strengthen the general respect accorded to the victims of vilification, and to what they have to say. However, the devaluing effect certainly may occur. It is, incidentally, important to note that the offending and insulting of hatespeech victims is not the relevant harm in this particular argument for "First

80. Robert C. Post, "Racist Speech, Democracy, and the First Amendment", William and Mary Law Review 32 (1991): 267-327, p. 310.

SPEECH AND EQUALITY

103

Amendment affinnative action". Those who propound hate-speech bans as a device of equalizing speech opportunities do not rely on the offensiveness argument, just as feminist proponents of pornography bans do not rely on the offensiveness argument in the "silencing" thesis. This is because offence is a setback to some important goods other than expressive opportunities. From that "offensiveness" perspective, hate speech (or pornography, for that matter) cannot be seen as depriving its victims of those same rights which the proposed laws would deny to the offenders. Consequently, offensiveness-based bans cannot be shown to remove the unfair surplus of expressive opportunities possessed by the dominant group to the victims of their expressions. The problem with the second type of "silencing" (silencing by demeaning) is that it is indistinguishable from the argument that some views should be suppressed on the basis of the concern that they may cause some listeners to adopt the wrong views about minorities. "Silencing" by means of disparaging me and my opinions only operates by convincing other people (that is, the audience) that they should treat me with less respect. In other words, the only way this sort of silencing operates is by persuading other people to hold certain beliefs about me and my group. Therefore, the second type of silencing involves a degree of "mental intermediation". In order for the wrong to occur, speech must be understood, interpreted and believed by some other people (that is, neither the speakers nor the victims) who will then, in effect, adopt a wrongful attitude toward the victim. This sort of mental intermediation is what makes this type of silencing so problematic. We want to hold people responsible for their own views and acts, and if they do something wrong as a consequence of holding their views, it is their responsibility which supersedes that of the people who persuaded them to hold those views. In her book Only Words, Catharine MacKinnon misconstrues a point about "mental intennediation" made by Judge Easterbrook in the Hudnut opinion. 8l She describes his argument as amounting to saying that any act becomes protected under the First Amendment as long as it is accompanied by a point of view, and that a prohibition would amount to viewpoint discrimination. She further attributes to him the arguably implausible view that "no amount of hann of discrimination can outweigh the speech interests of bigots, so long as they say something while doing it",82 and claims that "[n]othing in Hudnut explains why, if pornography is protected speech based on its mental elements, rape and sexual murder, which have mental elements, are not as well".83

81. American Booksellers Association v. Hudnut, 771 F.2d 323 (7th Cir. 1985).

82. MacKinnon, p. 93. 83. Id., p. 94.

104

CHAPTER 3

The reason why this is a non sequitur is that it is not the presence of a mental element, as such, which argues against the regulation of an act, but rather that, for Easterbrook, there is nothing but a mental element in the act of a pornographer that is relevant to the sexual crime. The wrongs which warrant regulation (that is, the conduct by a person other than a pornographer) occur through "mental intermediation" which is something other than a "mental element". Intermediation occurs when a "mental element" is conveyed by the speaker (a pornographer) to the hearer (a sex offender) who is then led to commit an act which warrants calls for a legal ban. Parenthetically, one might add that this description ignores other harms caused by pornography, such as violence against the models used in the pornography's production, but these are not the wrongs targeted (or not primarily targeted) by the anti-pornography ordinance proposed by MacKinnon and Andrea Dworkin, which is at stake in Hudnut. Easterbrook emphasized his point about mental intermediation by saying that the resulting harm "demonstrates the power of pornography as speech". 84 In response, MacKinnon asserts: "To say that the more harm speech does, the more protected it is, is legally wrong, even in this country".8S Elsewhere, MacKinnon stated that the Hudnut decision "takes harm to women as a measure of the power, hence value, of pornography as speech".86 Again, this is an ineffective criticism. The point is not how much harm the speech produces but how this harm occurs. If the harm occurs through persuading others to hold certain views under circumstances in which these views can be exposed, rebutted, and discredited, then even a great harm which results from the actions of the hearers should not warrant the silencing of the speakers. Some people will argue, for example, that advocacy of a one-party state, premarital sex, or unilateral disarmament may lead to very great harm, but the volume ofthe harm does not justify, in itself, suppression of these kinds of speech. By emphasizing the proviso "in itself', I mean to indicate that the suppression of various forms of speech which, if acted upon, would lead to great danger, may be justified without offending freedom of speech principles if additional factors occur, such as the likely imminence of harm and the unlikelihood that counter-speech would be effective. These additional factors, which may be captured by the condition of "clear and present danger", and the elaboration of which has generated a great deal of discussion in legal literature, are extraneous to the volume of harm. MacKinnon's refusal to draw a distinction between "mental intermediation" (as a category which would exempt the speaker from legal sanctions if the

84. 771 F.2d at 329. 85. MacKinnon, p. 93. 86. Catharine A. MacKinnon, "Pornography Left and Right", Harvard Civil Rights-CiVil Liberties Law Review 30 (1995): 143-68, p. 162, footnote omitted.

SPEECH AND EQUALITY

105

actual offence was committed by someone else) and the "mental element" (which does not affect the culpability of the offender to whom the "mental element" can be attributed) is connected to the idea that the actual evil attacked by anti-discrimination law primarily resides in the pernicious message which discrimination conveys. This idea explains a readiness to detect the similarity between anti-discrimination laws and prohibitions of hate speech. On the other hand, once we distinguish between "mental intermediation" and the "mental element", we may properly focus our attention on the actual discriminatory conduct, and on the agent guilty of the conduct. Of course, this is not to deny that discriminatory conduct conveys a pernicious message, or that it is supported by pernicious ideas; rather, the argument is that the evil targeted by legal prohibitions resides in the conduct rather than in the expression of the idea. Discrimination is illegal not because of the wrongfulness of the idea which underlies it or which it conveys. The view that what is really wrong about discrimination is the message it conveys - for example, that "the inequality inheres in the subordinating character of the message delivered by the separation,,87 or that racial segregation is wrong "primarily because of the message segregation conveys,,88 - seems to be a scholarly distortion of the real-life harms of discrimination. Scholars mainly deal with expressions and messages, so it is no wonder that the worst evils they discern in the social world are those which are done through, or to, messages. But for a person who is denied a job or an education because of her race, the evil of discriminatory action can be clearly and painfully identified with the harm inflicted and with the benefits denied, independently of the expressive meanings or consequences of such an act. It is one thing to believe that all conduct "says something" and therefore conveys a certain message to the observers, regardless of the motives of the agent, but it is another to say that the wrongness and harm of certain conduct consists in the message that it conveys, again, regardless of the motive. While the first part of the preceding sentence is trivially true, the second is highly implausible. To identify the evil of a racially motivated action (or an action which may have expressive racist significance) with the evil of the message, may in fact trivialize the evil of racist conduct by rendering it more subjective than is presupposed by anti-discrimination laws. Such rules primarily target the actual harm inflicted upon the victims, regardless of the motive and regardless

87. 88.

Thomas C. Grey, "Civil Rights Vs. Civil Liberties: The Case of Discriminatory Verbal Harassment", Social Philosophy & Policy 8 (1991): 81-107, p. 98, emphasis in original. Charles R. Lawrence, III, "If He Hollers Let Him Go: Regulating Racist Speech on Campus", Duke Law Journal (1990): 431-83, p. 439, emphasis in original.

106

CHAPTER 3

of the message that the illegal action mayor may not convey to the rest of the society. As one commentator noted with respect to hate-crime laws (such as penalty-enhancement provisions): the government is not only regulating acts irrespective of their expressive component, but also has a basis for doing so that is unrelated to suppressing (or preferring) particular views or expression - the interest . . . in preventing conceded harms from falling inequitably on members of a particular racial group.89

The strategy of resting the justification for anti-hate-speech laws or antipornography laws on their analogy with anti-discrimination laws is therefore dubious if the force of anti-discrimination laws is seen as targeting the message conveyed by discrimination. If anti-discrimination (or anti-hate crime) laws were mainly triggered by the wrongness of the message, then there would be a clear analogy with anti-hate speech laws and anti-pornography laws which, after all, seem to be triggered largely by the wrongness of the messages of these expressions. Hence the latter laws would lend much of their justificatory force to anti-hate speech and anti-pornography laws. But if the message-based theory of anti-discrimination laws is not the most convincing explanation for a rationale of these laws, the analogy collapses. If this distinction between targeting a discriminatory message and targeting discriminatory conduct seems artificial, consider the consequences of focusing on the message, as revealed by MacKinnon's policy argument. Writing about pornography and hate propaganda, she declares: "When equality is recognized as a constitutional value and mandate, the idea that some people are inferior to others on the basis of group membership is authoritatively rejected as the basis for public policy.,,90 This is undoubtedly true, but the last words are operative: "as the basis for public policy". There is a non sequitur from this proposition to the further proposition that the government has a mandate to penalize private expressions of racist ideas. The latter proposition would follow from the former only if we believed that the toleration, in the sense of non-prosecution, of any idea present in a society constitutes evidence that this idea is a "basis for public policy". But such a belief is implausible, if only because there is a diversity of mutually conflicting ideas on any matter tolerated by the law in a democratic society. The toleration of many ideas by law is no more a sign of their endorsement "as the basis for public policy" than is the holding of specific books by a library evidence of the endorsement of their contents by the library's management.

89. Elena Kagan, "Regulation of Hate Speech and Pornography After R.A. V.", University of Chicago Law Review 60 (199.3): 873-902, p. 886. 9Q. MacKinnon. Only Words. D. 106.

SPEECH AND EQUALITY

107

Immediately after the last-quoted sentence, MacKinnon continues: "This does not mean that ideas to the contrary [that is, ideas that some people are inferior to others] cannot be debated or expressed. It should mean, however, that social inferiority cannot be imposed through any means, including expressive ones.,,91 The problem is, of course, how to distinguish the "expression" of a,n idea of social inferiority from the "imposition of social inferiority through expressive means". MacKinnon herself confesses that "distinguishing talk about inferiority from verbal imposition of inferiority may be complicated at the edges", but is confident that such a distinction is "clear enough at the center with sexual and racial harassment, pornography, and hate propaganda.,,92 However, the context in which she draws this distinction (an approving discussion of two leading Canadian decisions which upheld prosecutions for anti-Semitic propaganda and for obscenity) suggests that any expression of an idea of social inferiority, regardless of whether it is accompanied by the advocacy of discrimination or by acts of intimidation, can be characterized as an "imposition of social inferiority through expressive means". If James Keegstra's anti-semitic statements to his students93 and Donald Butler's management of a pornography store94 amount to such an imposition, what "expression of an idea of inferiority" will escape such a characterization? The suspicion that the distinction drawn by MacKinnon is really a distinction without difference is supported by her specific examples of how "expressive means of practicing inequality can be prohibited".95 One practical example concerns classroom materials. (T]hose who wish to keep materials that promote inequality from being imposed on students - such as academic books purporting to document women's biological inferiority to men, or arguing that slavery of Africans should return, or that Fourteenth Amendment equality should be repealed, or that reports of rape are routinely fabricated - especially without critical commentary, should not be legally precluded from trying on the grounds that the ideas contained in them cannot be assumed false. No teacher should be forced to teach falsehoods. . . . Teachers who wish to teach such materials should be prepared to explain what they are doing to avoid creating a hostile

91. Id.

92. Id. 93. SeeR. \I. Keegstra, (1990)61 C.C.C. (3d) 1. 94. See R. \I. Butler, (1992) 89 D.L.R. 4th 449. 95. MacKinnon, Only Words, p. 107.

108

CHAPTER 3 learning environment and to provide all students the equal benefit of an education.96

What is particularly significant, from the point of view of our argument, is that MacKinnon presents the "imposition" of racist and sexist materials upon students, through their inclusion in course materials, as her example of "expressive means of practicing inequality". But why is it the case that when racist material is presented to students in the course of the educational process, it becomes an imposition upon them? The language of "imposition" suggests an element of coercion, but how relevant is that language to a selection of course materials which students have to read in connection with their study? Would it make a difference whether the course is optional or mandatory, or whether the readings are prescribed or merely recommended? If the inclusion of such materials amounts to "imposing" them, and if such an "imposition" amounts, in tum, to "practicing inequality" through "expressive means",97 then it becomes virtually impossible to draw the line between "expressing" racist ideas (which MacKinnon claims she would tolerate) and "imposing" social inferiority through expressive means (which she would prohibit). If the mere presence of pernicious texts in course materials which, under contemporary academic conventions, is not usually taken as involving the teacher's endorsement, falls on the "imposition of inferiority" side of MacKinnon's distinction, what then could be characterized as a legally permissible "expression" of an idea of inferiority? The point which I have been hoping to establish so far is that the second interpretation of "silencing" (silencing as demeaning) involves prejudicing other people's views about the victim which demeans and devalues the victim's speech, and such "mental intermediation" is central to this interpretation. However, what still needs to be explained is why governmental suppression of speech aimed at this "mental intermediation" is unacceptable, and why preventing such "silencing" through the suppression of wrong messages is incompatible with a robust conception of freedom of speech. In an important article, David Strauss argued that freedom of speech can be supported by what he calls the "persuasion principle" - a principle that forbids the government from suppressing expressions on the ground that the speech may be effective in persuading people to hold certain views. As Strauss puts it, "harmful consequences resulting from the persuasive effects of speech may not be any part of the justification for restricting speech".98 The general idea is not novel and it resonates with an intuitive view that what is particularly important

96. (d., pp. 107-108. 97. (d., p. 107. 98. David A. Strauss, "Persuasion, Autonomy, and Freedom of Expression", Columbia Law Review 91 (1991): 334-71, p. 335.

SPEECH AND EQUALITY

109

about speech is precisely that it has a capacity to persuade people to the views of the speaker. Therefore, to use the persuasive power of speech as a factor figuring in the justifications for restrictions of speech would undercut many of our intuitive notions about why speech warrants special protection. As another writer suggests, one of the impermissible governmental justifications for restricting even a lower-value speech is the government's "fear that people will be persuaded or influenced by ideas" expressed.99 On the basis of the persuasion principle, Strauss would explicitly deny the government the power to suppress hate speech and pornography on the grounds that these categories of speech may cause people to hold harmful views toward minorities: such restrictions "constitute efforts to restrict speech on the ground that the speech will persuade people to adopt attitudes that the government considers undesirable".IOO Of course, the persuasion principle does not provide a conclusive answer to a general question about the appropriateness of such restrictions; it only precludes restrictions relying on this particular, impermissible rationale. Further, the "persuasion principle" would not prevent the government from suppressing "fighting words", precisely because they are more like assault, and their persuasive effect is nil. But in the case of abstract, non-targeted views which express hate or contempt towards a certain group, for example in books or the media, the fact that these views may prove to be persuasive to the audience is not an adequate reason to forbid their publication under a meaningful notion of free speech. The purported "silencing" which is at work here consists of the fact that, due to the content of hate messages, members of the target group will become convinced that it is not worthwhile for them to make their points publicly and the substantial majority of the audience will become convinced that the members of the target group are not worth listening to. Regardless of the likelihood that these two effects will follow as a result of hate messages (and, arguably, the former effect is likely to occur only if the latter occurs first), it is not the role of government to prevent adult individuals from becoming convinced about whether it is worthwhile to take part in a public debate, or to concern itself with the credibility, or otherwise, of various classes of speakers. To prevent that type of silencing is to deny the autonomy and individual responsibility of the hearers, and it implies that once the hateful speech is uttered, it somewhat mechanically induces negative changes in the hearers' opinions. As C.E. Baker noted, such an approach "is much like the once 99. Sunstein, p. 155. 100. Strauss, "Persuasion", p. 340.

110

CHAPTER 3

fashionable theory of the mass media that saw a public duped into accepting whatever messages the mass media transmitted".lol The corollary of this theory is that hearers are not responsible for the views which they form in their minds about various racial groups because those views are shaped by the speakers. They are being seen as thoughtless receivers of ideas imposed upon them by speakers. Characteristically, this is one of the striking features of MacKinnon's argument about the silencing produced by pornography. The effects of pornography upon its audience, that is, upon the readers and viewers of pornographic materials, are presented as a linear causation in which the ''receiver's'' responsibility vil'tu!lly disappears: [P]ornography makes rapists unaware that their victims are not consenting. . . . it creates 'a person who no longer distinguishes between violence and rape, or violence and sex' .... [By imposing the death penalty upon a sex murderer] we kill a man rather than let his victims stop the pornography that produced him. I02

This is the language MacKinnon consistently uses to describe the way in which pornography works upon its consumers: it produces people who lack a fundamental moral sense, and who perpetrate hideous crimes upon women. In this presentation, the existence of independent minds, brains, emotions, preferences and values among the offenders disappears from the picture. Some feminist critics of the anti-pornography ordinance drafted by Andrea Dworkin and Catharine MacKinnon objected to this simplistic causal view: "The ordinance reinforces a destructive sexist stereotype of men as irresponsible beasts, with 'natural physiological responses' which can be triggered by sexually explicit images of women, and for which the man cannot be accountable.... 03 From this perspective, pornography commits wrongs against women, and the fact that these are done by someone, acquires only a secondary moral relevance. In rejecting such a perspective, one need not deny that there is room for moral and legal culpability in our thinking about those who advocate immoral or criminal actions. If A incites B to commit a crime (or, an immoral action) against C, then surely there are circumstances in which we should hold A responsible for the outcome. However a number of considerations have to be taken into account, such as how authoritative A is in relation to B; how

101. C. Edwin Baker, "Of Course, More Than Words (Book Review)", University ofChicago Law Review 61 (1994): 1181-1211, p. 1206, footnote omitted. 102. MacKinnon, Only Words, pp. 96-97, emphases added, footnote omitted (quoting arguments of counsel in Schiro v. Clark, 963 F.2d 962 (7th Cir. 1992». t03. Nan D. Hunter & Sylvia A. Law, "Brief Amici Curiae of Feminist Anti-Censorship Taskforce, et. aI., in American Boo1csellers Association v. Hudnut", University of Michigan Journal Q{ Law Reform 21 (1987-88): 69-136, p. 127.

SPEECH AND EQUALITY

111

competent B is to evaluate other people's (including A's) judgments; to what extent does A's advocacy provide B with new ideas, as opposed to providing him with ways of fulfilling the desires which he already has; how imminent is the likely action by B against C, so that no counter-influences may be effective, etc. Speech, of course, does influence human behaviour, opinions and attitudes; however, this influence is much more complex than just a one-way, linear impact, whereby words are thought to "produce" certain views, attitudes and behaviour on the part of the audience. Racism and pornography are no exception. Depending on the empirical circumstances of a case, the influence of racists and pornographers on their audiences may be such that they come to accept their views, or that they respond with repulsion, or that they stick to their prior views. Even if, in particular circumstances, it is likely that the impact of speech will be such that it will strengthen prejudiced and bigoted attitudes on the part of the audience, the silencing effect thus produced will directly result from the audience's response to the speech, rather than from the speaker. It is this fact about the role of the audience which should be addressed by the law and policy if "silencing as demeaning" is to be restrained.

3

Asymmetry of Figbting Words

I have argued earlier in this Chapter that the analogy between anti-hate speech laws and affirmative action is inadequate and that restrictions upon racist speech cannot be seen as affirmative-action style equalization of communicative opportunities. But not all restrictions upon racist verbal assaults are equal, and earlier I made an explicit exception for racist "fighting words". Here I wish to explore the issue of fighting words more directly, in the context of one specific idea which may seem like a reflex of the concept of affirmative action. The idea is that anti-vilification laws should only work in one direction; namely, they should protect members of disadvantaged minorities against vilification, but should not apply to vilification expressed by members of such minorities against the majority. Indeed, at least one legal scholar characterized such an "asymmetrical" prohibition on racist expression as a "form of 'affirmative action' for racial minorities".l04 Such an asymmetrical prohibition, I will argue, is justified, even if removed from the context of the "affirmative action" framework. The argument that racial vilification may be punished only when it is directed at members of traditionally disfavoured, disadvantaged minorities,

104. Amar, p. 159.

III

CHAPTER 3

but not when members of the majority are vilified, has a clear egalitarian flavour. But, in the context of "fighting words", it need not appeal to a dubious idea that some speech restrictions amount to a form of affmnative action. Before I tum to this theme, some explanation of the meaning of "fighting words" is required. As I observed earlier, room for the suppression of racist speech exists even within a generous conception of freedom of speech. I have in mind verbal assaults, motivated by racial hatred in face-to-face situations, when the victim has little or no opportunity to avoid the assault, and where the point of the "message" is not to persuade anyone towards the speaker's views about the group depicted in the statement. Their immediate aim is to intimidate the victim or to provoke him or her to fight; these expressions do not purport, as their primary aim, to communicate to the world at large, criticism of the target group to which the victim belongs. No "redeeming" features which apply to abstract communications about a given group (such as a contribution to general debate about public issues or an attempt to persuade people to hold certain views) apply here. It is true that we have no right to be protected from distress resulting from knowing that others have a low opinion of us (or a low opinion of a larger class to which we belong). Indeed, we'-are entitled to know of the fact that some people hate our race, religion, sexual orientation, etc., and government suppression of hate speech would deprive us of this important, if distressing, knowledge. On the other hand, we also have a right to be protected from assaults which provoke us to fight or which intimidate us. Targeted vilifying remarks in face-to-face, personalized situations are better analogized to assaults than to communicative statements, and should be treated as such by the law. To be sure, one should be aware of the dangers of an expansive interpretation of such verbal assaults, and it is important to restrict legal sanctions to those cases where the communicative value of a statement is nil or negligible. One can think of a number of precautions against an overly broad interpretation of this rule, such as the use of a somewhat modified set of the criteria currently dominant in the interpretation of the "fighting words" doctrine in First Amendment jurisprudence where a classical, legal notion of "fighting words" defines them as "those which by their very utterance inflict injury or tend to incite an immediate breach of the peace".lOS A British equivalent can be found in national public order legislation which prohibits the use of "threatening, abusive or insulting" words in public, and which has been interpreted as calling for a determination of ''whether the words are sufficiently threatening, abusive, or insulting to be likely to provoke a breach of the peace

105. Chaplins/CY v. New Hampshire 315 U.S. 568, 572 (1942).

SPEECH AND EQUALITY

113

by any average or ordinary audience to which they are deliberately addressed and to which they specifically refer". 106 The following four criteria seem to establish a set of important precautions against an overly broad interpretation of "fighting words": to be suppressed, offensive epithets must (1) constitute personally abusive insults, (2) be addressed in a face-to-face manner, (3) target a specific individual and be descriptive of that individual, and (4) be addressed to an individual unable to avoid the assaultive message. These criteria roughly correspond to the set of indicia which comprise the constitutionally acceptable "fighting words" prohibition in the United States, as interpreted by Stephen Gard in his important article. t07 The first three criteria merely paraphrase the first three criteria described by Gard. The fourth criterion, absent in Gard's list, describes a familiar "captive audience" requirement and is meant to preclude suppression of personally offensive speech where its victims have ample opportunity to avoid the offensive speaker. The idea, discussed earlier in this chapter (in Section 1), and further in Chapter 6.3(A), is that the burden is normally on readers and listeners to take all reasonable steps to avoid receiving unwanted communications which they can anticipate, or to avert their attention from unwanted messages. To suppress speech on the basis of a negative emotional reaction by members of the audience who could escape exposure to the speech, would amount to making the listener the ultimate judge of the rights of the speaker, or, in other words, recognition of the "heckler's veto". Naturally, as discussed earlier, the very concept of "captivity" of an audience is ambiguous and must not be treated too literally, but rather be seen as a proxy for the "avoidability" of an unwanted message. The fourth criterion in Gard's list - which is missing from my proposed set of indicia of "fighting words" - is that the offending language "must be uttered under such circumstances that the words have a direct tendency to cause an immediate violent response by the average recipient".108 To be sure, Gard is right to conclude that "the failure to limit the fighting words doctrine could have a serious chilling effect on ... political expression"; 109 however, the "tendency

106. Barnum, p. 81. Barnum's characterization is mainly based upon his analysis of a leading British case, Jordan v. Burgoyne, [1963] 2 All E.R. 225. Prohibitions on "threatening, abusive or insulting" words are in the Public Order Act 1936 and in the Race Relations Act 1976. 107. Stephen W. Gard, "Fighting Words as Free Speech", Washington University Law Quarterly 58 (1980): 531-81, p. 563. 108.Id. 109. Id., pp. 562-63.

114

CHAPTER 3

to cause an immediate violent response" test seems to me to be an unduly restrictive requirement. A personal, face-to-face assault which has a tendency to frighten or intimidate the victim should also be liable to suppression, subject, of course, to the presence of the remaining three indicia of "fighting words". The requirement of a ''tendency to cause a violent response" privileges those who tend to respond to offenses in a "macho" manner and disregards an interest in tranquillity among those groups in which "average recipients" tend to abstain from any counter-attack. One commentator notes: "Minorities often flee or withdraw when faced with racial speech instead of being incited to violence. Thus, most racial speech does not reach fighting words exception status".11O Furthermore, to attach that condition to a suppression of "fighting words" is clearly counter-productive from the viewpoint of promoting self-control and the discouragement of violence. Such attitudes are more likely to prevail if we know that the suppression of verbal assaults is not conditional upon the victims' tendency to respond violently. Additionally it raises a fundamental issue of fairness. As Kent Greenawalt asks: "Can the same remark be punishable if directed at the one person able to respond and constitutionally protected if directed at people not able to match the speaker physically?"111 Subject to these qualifications, the general principle that the law may properly restrict verbal, vilifying, targeted assaults in face-to-face situations, seems unobjectionable. The good protected is not one's sensibility but one's sense of safety. This is a major difference from the point of view of the values promoted by freedom of speech. The interest in values promoted by a general regime of free speech is minimal when speech consists of targeted verbal assaults. The aim of "fighting words" is not to open a discussion, invite counterarguments, advocate a view, or to convince one's audience. The availability of efficacious counter-arguments to those who oppose an offensive message, as an alternative to legal suppression, is nil. Neither is there an efficient way of avoiding the verbal injury: part of the definition of "fighting words" is that they come without a warning. "Fighting words" may be based on opinions (and usually are), but their main aim is not to convey those opinions, and correspondingly, the suppression of "fighting words" does not impair the articulation, circulation, and competition of ideas and viewpoints. Justice Scalia compared "fighting words" to a noisy sound truck: "Each is . . . a 'mode of speech,' ... both can be used to convey an idea; but neither has, in and of itself, a claim upon the First Amendment" .112

110. Beth C. Boswell-Odum, ''The Fighting Words Doctrine and Racial Speech on Campus", South Texas Law Review 33 (1992): 261-86, p. 283, footnotes omitted. Ill. See Kent Greenawalt, "Insults and Epithets: Are They Protected Speech?", Rutgers Law Review 42 (1990): 287-307, p. 297. 112. R.A. V. v. City of St. Paul, 505 U.S. 377, 386 (1992), quoting Niemotko v. Maryland, 340 U.S. 268, 282 (1951) (Frankfurter, J., concurring in result).

SPEECH AND EQUALITY

115

The question now arises whether racist assaults which meet all four criteria should be equally suppressed, regardless of the position of a group whose member is attacked. Some writers believe that there should be full symmetry in this respect, suggesting that any special protection for disadvantaged groups would offend the principle of fundamental fairness. They also point to the political and practical disadvantages of a one-way solution: it would involve the government in intergroup hostilities and oblige law enforcers to make often difficult judgments about the identification of disadvantaged groups. Perhaps more importantly, it would be unpalatable to the general public since "[t]o intervene in a manner that protel/ts only one opponent may evoke charges of favoritism or capture, despite the fact that the violent impact of hate speech tends to occur only one-way".113 Other writers, including Mari Matsuda and Charles Lawrence, believe that only members of disadvantaged groups should be protected. I 14 This difference has been reflected, to some degree, in R.A. V. v. City oj St. Paul, a decision of the U.S. Supreme Court which I discussed at length in Chapter 2. The Court, it may be recalled, invalidated a city ordinance which prohibited certain symbolic actions on the grounds that they arouse anger, alarm, or resentment in others on the basis of race, and under which a person was punished for burning a cross on the lawn of a black family. Justice Scalia, writing for the majority, read the ordinance to apply symmetrically against whites as well as blacks: "racial epithets . . . would be prohibited to proponents of all views" .11 5 Justice White, concurring with the judgment but disagreeing with the argument of Scalia, suggested in tum that the case was about protecting vulnerable groups from dominant ones, and he interpreted the "fighting words" prohibition as "a ban on a class of speech that conveys an overriding message of personal injury and imminent violence, . . . a message that is at its ugliest when directed against groups that have long been the targets of discrimination" .116 The conclusions drawn by proponents of a one-way prohibition, such as Matsuda and Lawrence, seem plausible, though not necessarily for the reasons supplied by those proponents. The reason for the difference in the rationale for "asymmetrical" protection is largely connected with the fact tdat those writers would not confine suppression of racial vilification to "fighting words" only: indeed, they make it clear that the "fighting words" doctrine is not an adequate

113. See Toni M. Massaro, "Equality and Freedom of Speech: the Hate Speech Dilemma",

William and Mary Law Review 32 (1991): 211-65, p. 243. 114. See Matsuda, pp. 2357-58, 2361-63; Lawrence, pp. 450 n. 82. 115.505 U.S. at 391. 116. Id., at 408409.

116

CHAPTER 3

paradigm for the conceptualization of the prohibition of racial vilification. 117 Their reasons for a one-way suppression stem from the observation that the degree of harm is much higher when racial insults are directed against members of traditionally oppressed classes. "The harm and hurt is there, but it is of a different degree", says Matsuda about expressions of hatred by subordina,tedgroup members, and she adds: "Because the attack is not tied to the perpetuation of racist vertical relationships, it is not the paradigm worst example of hate propaganda. The dominant-group member hurt by conflict with the angry nationalist is more likely to have access to a safe harbor of exclusive dominantgroup interactions" .118 This is undoubtedly true and, if one accepts the principle of suppression of racial vilification in non-fighting-words situations, one should seriously consider confining it to cases involving a particularly high degree of harm, with use of the position of a victimized class in a society acting as an important indicium of the degree of harm. But the arguments for one-way suppression proposed in this book are somewhat different, and they should be seen in the context of confining suppression of racist speech to "fighting words" only. The first argument is tentative and contingent, the second is more fundamental. The tentative and contingent reason stems from the observable fact that various "offensive language" and "offensive behavior" regulations (to which the "fighting words" controls belong) are, more often than not, enforced in a biased and partial way which tends to exaggerate the seriousness of the insults against majority members (in particular, against enforcement agents themselves) and to undervalue insults against members of disadvantaged minorities. An Australian example provides a good case in point. The New South Wales Summary Offences Act contains sections dealing with "offensive conduct or language" in public places. I 19 Whilst the legislation could be used as an anti-racial-hate law, it has actually been overwhelmingly used by police against Aborigines. 120 It is precisely for this reason that a report by the Australian Institute of Criminology entitled "Aboriginal Justice Issues" strongly recommended that immediate steps be taken to decriminalize offensive behaviour and language in the Summary Offences Act. 12I So the first reason for a deliberate bias expressed in the framing of one-way regulations would be that such an asymmetrical rule would aim at offsetting this social bias in the actual enforcement of such regulations.

117. See Lawrence, pp. 453-56. 118. Matsuda, p. 2361. 119. NSW Summary Offences Act 1988, sections 4 (1) (a) and (b). 120. See Chris Cunneen, "Constructing a Law and Order Agenda: Conservative Populism and Aboriginal People in North West New South Wales", Aboriginal Law Bulletin 2 (38), June 1989): 6-9, p. 6. 121. "Recommendations: Aboriginal Justice Issues Conference", Criminology Australia 4 (2) (October 1992): 29-32.

SPEECH AND EQUALITY

117

The idea is that law enforcers would be legally instructed to pay special (if not exclusive) attention to verbal epithets against minority members, and would be forced to abandon their anti-minority bias. The second, more fundamental reason, concerns the very notion of groupbased insults, such as racist epithets. These insults are social constructs, and they acquire their meaning from social contacts. Their contexts are characterized by past and present discrimination, domination, and stereotyping. What makes a given statement "racist" is that it directly relies upon the resources of discrimination, domination and stereotyping. "Words" themselves do not "wound"; what does wound is the fact that some words come in a package recognizable both by the speakers and the hearers as conveying contempt, hostility and domination. "You bastard!" uttered jokingly or affectionately by one friend to another in a pub is not equivalent to "You bastard!" screamed by a wealthy passer-by to a beggar. This is all the more obvious when the social divisions to which an insult appeals are along racial or other group-based lines in a society which has a history of discrimination and stereotyping along those very lines. It is a striking fact that our language has ample epithets of Blacks or homosexuals, but no corresponding epithets of Whites or heterosexuals. As Thomas Grey suggests in a remark that has wider relevance than just to the public culture in the United States: "It is because, given our cultural history, no such general prejudices strike against the dominant groups that there exist no comparable terms of universally understood hatred and contempt applicable to whites, males, and heterosexuals as such.,,122 But even if such terms existed, they would sound different because they would be uttered in a context which immediately disarms their force. They would not be perceived as insults (not as racist or sexual-orientation related insults, that is), regardless of the intentions of the speaker, but, at worst, as an expression of the speaker's deep dislike of the target. However, this dislike would fail to acquire the significance of racist (or other) insults due to the absence of publicly shared meanings of subordination and inferiority relating to the group whose member had been targeted. From this point of view, the nature of a "racist insult" is analogical to acts of racial discrimination. The latter are not merely constituted by the unequal treatment of a class of people based on race, but also by the fact that this inequality of treatment is consistent with the prevailing pattern of traditional disadvantage, stigmatizing and powerlessness. All these three indicia are important and describe related but distinct phenomena. First, racial

122. Grey, p. 96, emphasis in original, footnote omitted. See also, similarly, Sunstein, p. 186.

118

CHAPTER 3

discrimination (as opposed to mere unfairness of treatment) means that the excessive burdens imposed by a given regulation add to an existing, traditional pattern of injustice suffered by a burdened group; discrimination perpetrates, freezes or strengthens the existing pattern of disadvantages. Second, racial discrimination stigmatizes its victims as inferior, and it reflects and strengthens stereotypes against a victim group. Unfair treatment against members of an otherwise advantaged group is still unfair but at least it does not express hostility, contempt or prejudice by the lawmakers towards the group burdened by a given regulation. Third, racial discrimination occurs when a burden is imposed upon those unrepresented or underrepresented in the political and legislative process, of which a given regulation is a product. In contrast, "[w]hen the group that controls the decision making process classifies so as to advantage a minority and disadvantage itself, the reasons for being unusually suspicious ... are lacking".123 This brief excursus into a theory of discrimination, upon which I have commented at greater length elsewhere,124 is not meant to suggest that hate speech is a form of discrimination, but rather that both racist hate speech and racial discrimination can be defined only in a context-sensitive way, and the relevant context focuses on the place of the victims in the overall pattern of discrimination, stigmatizing and powerlessness. Just as "discrimination against blacks" is qualitatively different from "discrimination against whites" in a society that has a history of anti-black disadvantage and prejudice, so are racial epithets hurled at members of traditionally subjugated groups qualitatively different from racial epithets against those enjoying a comfortable place in their society and in popular culture. As much as individuals against whom violent expression is addressed may be distressed and upset, racist speech expresses contempt and hostility towards a racial group, and contempt and hostility are social constructs, not just products of the speaker's individual intentions. The upshot is that, even if fighting-words regulations are drafted in a neutral manner, their application to group-related epithets, such as racist insults, will be "asymmetrical" in practice. "Epithet" cannot be defined in isolation from the social meanings attributable to certain words, and those meanings are different, and more severe, when uttered against members of disadvantaged, powerless, and stigmatized minorities.

123. J. H. Ely, "The Constitutionality of Reverse Racial Discrimination", University of Chicago Law Review 41 (1974): 723-41, p. 735. 124. See Wojciech Sadurski, "Gerhardy v. Brown v. the Concept of Discrimination", Sydney Law Review II (1986): 5-43, pp. 35-40; also Wojciech Sadurski, "Judicial Protection of Minorities: the Lessons of Footnote Four", Anglo-American Law Review 17 (1988): 163-82. See also, similarly, Tom D. Campbell, "Unlawful Discrimination", Ethical Dimensions of Legal Theory ed. by Wojciech Sadurski (Amsterdam: Rodopi, 1991), pp. 153-71; Ronald Dworkin, Law's Empire (London: Fontana, 1986), pp. 381-99.

CHAPTER 4 DISCRIMINATION AND ILLOCUTIONARY ACTS 1

IDocutions and Perlocutions

In the preceding chapter, I suggested that the equal-opportunity argument in favour of hate-speech restrictions and pornography bans fails because banning hate speech and pornography does not improve communicative opportunities for vilified racial minorities or women. This was because, contrary to the "silencing" argument, the very presence of hate speech in society did not remove the communicative opportunities from the victims of such hate speech, even if it adversely affected their other important interests and goods. But there is another defence of restrictions upon hate speech and pornography which links racial vilification with racial discrimination, and pornography with the subordination of women, more directly than a silencing argument does. The harms of racial vilification and of pornography are often discerned not in the wrongness of the messages these expressions convey, with all their resultant harms, but rather in the wrongness of the acts they constitute. Hate speech is often viewed as a form of subordination, not because it silences its victims, but in and of itself. As Mari Matsuda says, racist speech is "a mechanism of subordination, reinforcing a historical vertical relationship", and it is "the structural subordination of a group based on an idea of racial inferiority." I Similarly, pornography is sometimes viewed as "the graphic sexually explicit subordination of women through pictures and/or words".2 It is important to note that the argument that racist speech is discrimination is different from the argument, considered in Chapter 3, that racist speech, like discrimination, should be banned because it is predicated on a message of racial inferiority concerning a victim group. In understanding the nature of the proscribable speech, the two arguments may be seen as symmetrically converse. The argument considered in Chapter 3 rests on the assumption that the crucial element of racist speech is its message, and that it was the presence of this same message which triggered the prohibition of the discriminatory acts. In other words, the analogy between anti-hate speech and anti-discrimination laws is seen in the fact that anti-discrimination laws are based upon rejection of the message that the discrimination conveys, and the same message is, of course, 1.

2.

Mari J. Matsuda, "Public Response to Racist Speech: Considering the Victim's Story", Michigan Law Review 87 (1989): 2320-81, p. 2358. This is part of the definition of "pornography" in the anti-pornography Model Ordinance drafted by Andrea Dworkin and Catharine MacKinnon, see Catharine MacKinnon, Only Words (Cambridge, Mass.: Harvard University Press, 1993). p. 121, n. 32.

120

CHAPTER 4

present in racist speech as well. In summary, this argument has the following structure: anti-discrimination laws target the message of racial inferiority; the same message is conveyed by racist hate speech, hence those same arguments which lend moral force to anti-discrimination laws also provide support for the laws which restrict hate speech. It was the first premise, namely, that antidiscrimination laws primarily target. an improper message, which I found wanting. The argument to be considered in this chapter is also based on an analogy with anti-discrimination laws which lend moral force to anti-hate speech and anti-pornography restrictions, but the analogy is seen in the fact that both anti-discrimination laws and speech-restrictive laws outlaw discriminatory conduct. The structure of this argument is as follows: anti-discrimination laws rightly outlaw discriminatory conduct; racist speech and pornography constitute a form of discrimination and subordination, hence those same arguments which give moral strength to anti-discrimination laws at the same time support laws against racist speech and pornography. As is clear, it is the second premise which is operative, and controversial. Accordingly, my attention in this chapter will focus on the argument that pornography (and hate speech) constitute acts of discrimination and subordination. In considering that argument, it would be somewhat glib to say that the identification of hate speech with racial subordination is a sleight of hand, a far;on de parler; a vague metaphor at best. This is the same response which has often been made against Catharine MacKinnon's proposition that pornography is a form of discrimination against women and that depictions of violent sex are a form of rape.3 However, anti-pornography advocates who endorse MacKinnon's position, as well as proponents of hate-speech bans who share Matsuda's approach, both insist that these claims are not metaphorical, but literal. MacKinnon herself complains that her proposition that pornography is "an act against women" has been seen as "metaphorical or magical, rhetorical or unreal, a literary hyperbole or propaganda device".4 Her opponents, she claims, assume "that words have only a referential relation to reality".S Certainly, words do more than merely "refer to" reality; they also constitute reality. MacKinnon's thesis about pornography, and a parallel thesis about hate speech, can be best interpreted if rendered into J.L. Austin's framework of "illocutionary" acts, as opposed to "perlocutionary" utterances. Indeed, MacKinnon herself acknowledges the general impact of Austin's philosophy of language on her ideas, though she admits that "Austin is less an authority for my particular development of 'doing things with words' and more a foundational exploration of the view in language theory that some speech can be 3.

See, e.g., Bernard Williams, "Drawing Lines (Book Review)", London Review of Books, 12 May 1994, p. 10; Ronald Dworkin, Freedom's Law (Oxford: Oxford University Press, 1996),

p.230. 4. 5.

MacKinnon, p. II. Id. at II.

DISCRIMINATION AND ILLOCUTIONARY ACTS

121

action".6 A more "particular development" of the ''pornography is subordination" thesis along Austin's line, has been done with a great deal of sophistication by Rae Langton.7 A similar argument, with regard to campus hate speech, has been discussed by Andrew Altman.8 However, 1 will focus my discussion on Langton's work rather than that of Altman because Langton's defence of a position discussed in this chapter is much more forceful thai1 Altman's. Altman is committed to finding a "liberal middle ground",9 while Langton is not constrained by such an aspiration. Much of the argument which follows concerns the discriminatory nature of pornography, more than hate speech. Ultimately, though, the conclusions apply to the illocutionary-act theory as applied to both these categories of speech. I will argue that neither pornography nor hate speech (other than direct, personalized insults) can be easily analogized to illocutionary acts which do subordinate or do discriminate against their victims. Obviously, these two categories of problematical speech raise a host of different issues. However, from the point of view of applying speech-act theory to these two categories of speech, the differences are not that important. It is significant that MacKinnon's book, which has quickly become a locus classicus of this theory, runs the argument for outlawing these two categories of speech together: Only Words is devoted not only to a critique of a liberal approach to pornography restrictions, but also to hate speech prohibitions. MacKinnon herself characterizes racial group defamation as "a practice of discrimination in verbal form".10 For the purposes of our argument, it is more convenient to focus initially on pornography because, in my opinion, the best available exposition of the theory that speech subordinates, as provided by Langton, concerns pornography. The argument can be applied, mutatis mutandis, to hate speech as well. The difference between illocutions and perlocutions is that the perlocutionary effects of an utterance consist of its causal consequences, that is, the effects it has on the hearers, while the inocutionary acts consist of the actions performed in saying something. Langton quotes Austin's example: "In saying 'I do' 1 was marrying; by saying' I do' I greatly distressed my mother" .11 The former statement describes an illocutionary act; the latter statement illustrates its perlocutionary effects. Obviously, our words often have a power ofillocution. It may well be that MacKinnon's propositions about pornography or Matsuda's propositions about hate speech are sometimes taken to be Id. at 121 n. 31. Rae Langton, "Speech Acts and Unspeakable Acts", in Freedom of Communication ed. by Tom Campbell and Wojciech Sadurski (Aldershot: Dartmouth, 1994): 95-130. 8. Andrew Altman, "Liberalism and Campus Hate Speech: A Philosophical Examination", Ethics 103 (1993): 302-17. 9. Id. at 317. 10. MacKinnon, p. 99. II. Langton, p. 100.

6.

7.

122

CHAPTER 4

metaphorical, or plainly wrong, precisely for the reason that J.L. Austin detected a long time ago; namely, our unfortunate tendency to overlook the illocutionary power of words and to consider them only in tenns of their semantic meanings (as locutions) or their effects upon the hearers (as perlocutions). Viewed as ilIocutions, words can do a variety of things, includitig subordinating others. Obvious examples include the sentences: "Blacks are not permitted to vote", "Hispanics need not apply", "Women are not allowed in this pub", etc. These speech-acts constitute acts of discrimination in and by themselves. They discriminate against their targets not by virtue of their effects upon the hearers, that is, not by causing the audience to think one way or the other about the target group, and certainly not by depicting a situation of inequality. Rather, they are speech-acts which constitute unequal treatment. Langton describes three ways in which speech acts subordinate people: (1) they rank members of the target group as having an inferior worth; (2) they legitimate the discriminatory behaviour of others; and (3) they deprive members of the victim group of some important powers. They do all this, she correctly observes, because the speaker has the authority to subordinate others: ''The speech acts of apartheid that legitimate discriminatory behavior and unjustly deprive blacks of certain rights have an exercitive force that would be absent if they were made by speakers who did not have the appropriate authority".12 This is a crucial point which underpins the argument classifying hate speech or pornography as a form of subordination. Can the speakers (that is, the pornographers or racists making public derogatory remarks about minorities) be seen as having the degree of authority necessary to make an illocution which subordinates in the same manner that the illocution "Blacks are not permitted to vote" subordinates? This obviously occurs when the government utters words of racism or of gender inequality, but that is not an issue in the context of racist hate speech or pornography. Obviously, authority which may render a speech act discriminatory or subordinating is not exclusively held by the government. There may also be forms of official non-governmental authority (say, authorities ofa private university, or an umpire in a sporting match) and of nonofficial authority (say, parents over their children), and this suggests that the phenomenon of authority behind an illocutionary utterance (and in particular, those illocutions which are capable of discriminating or subordinating) has to be considered more carefully.

12. Id. at 103.

DISCRIMINATION AND ILLOCUTIONARY ACTS

2

123

Authority in Discriminatory mocudons

The issue of authority is very complex, but for the purposes of illocutionary acts I would suggest that it can be reduced to two main ingredients: a normative ingredient of "legitimacy" and an empirical ingredient of "control". The former occurs when a large percentage of the intended audience of the speech act recognizes the right of the speaker to make a ''verdictive'' utterance, that is, an utterance dictating the way in which people should behave. For example, most of the players and spectators believe that an umpire is entitled to perform her role in a given sporting event. The second ingredient occurs when the speaker's verdicts are actually enforced, and she has the means of imposing her decisions upon those to whom they are addressed. An ideal model of authority requires that these two conditions be fully met. In real life, however, and especially in the context of non-governmental authority, the under-attainment of one of these factors can be compensated by a higher attainment of the other. For example, although parents may be rather unsuccessful in actually enforcing their adopted rules of familial conduct upon their children, there is a general conviction (shared by the children themselvesand by the rest of the community) that they have a right to issue such rules, and, thus, their illocutions have a capacity to subordinate. Conversely, the credentials of a given regime to govern may be rejected by many people, but if it manages to enforce its will upon a large fraction of the popUlation, its speech will have the authority needed to make its ilIocutions effective. Obviously, a total failure in either of these dimensions will result in a non-authoritative utterance. What "authority" do racists and pornographers have, in countries such as the United States or Canada or Australia, which would make their speech acts "verdictive"? Do they have authority in the domain in which they operate? Is their speech understood to command actions that rank their targets as having inferior worth, to legitimate discriminatory behaviour, and to deprive the victims of some important powers - the three features that Langton considers characteristic of illocutionary acts of subordination? It would be unrealistic to say that they never have such authority. An association, group, club or fraternity which announces that African-Americans are not welcome at its functions does exercise illocutionary authority by performing a speech act which is, in itself, an act of discrimination that is no different from any other act of "private discrimination". Just as the refusal by a private landlord to rent his apartment to particular prospective tenants because of their skin colour is an act of discrimination, so too is an announcement that certain people, defined by their skin colour, will not be admitted to a particular organization or function, an identical act of racial discrimination. In fact, both these types of acts are committed by words, can be motivated by the same

124

CHAPTER 4

prejudice and produce the same type of hann. However, this is not the sort of speech we are concerned with when we discuss racist hate speech. It is important to see that acts such as those denials of services, membership, admissions, etc., are, or should be, legally proscribable precisely because specified conditions of "authority" - both in the sense of recognition of the legitimacy and of power of enforcement - are met. However, here we are concerned with statements made in speeches, articles, publications, the mass media etc., and we examine whether the required elements of authority of the speakers are met therein. It is perhaps instructive to see how the analogous argument applies to pornography. According to Langton, ''pornographers' speech acts" may be seen as meeting the condition of authority if we agree "that pornography's voice is the voice of the ruling class".13 She goes on: Just as the speech of the umpire is authoritative within a certain domain - the game of tennis - so pornographic speech is authoritative within a certain domain - the game of sex. The authors of pornographic speech are not mere bystanders to the game, they are speakers whose verdict counts. . . . Pomography tells its hearers which moves are appropriate and permissible: it tells them that certain moves are appropriate because women want to be raped, it legitimates violence. 14

The problem with this argument is that it exploits ambiguity in the proposition that someone's ''verdict counts", and in the proposition that someone tells his or her audience ''which moves are appropriate and permissible". Ambiguity arises because someone's "verdict" may "count" in either an illocutionary or perlocutionary fashion. An umpire's decision in a game is an example of the former while an art connoisseur's judgment about a painting may be an example of the latter. I will return to this distinction below. My general point is that an umpire's ''verdict counts" because the twin conditions oflegitimacy and control are met to a relatively high degree. It is much less credible to say that there exists a combination of these two conditions in a pornographer's speech. It is hardly credible to say that people who buy pornography (much less, the general audience) recognize a pornographer's authority to declare the appropriate and permissible ways for men to approach women in the domain of sex; hard-core pornography is more a masturbation device than a code of conduct. It is even less credible to talk about a pornographer's effective power to enforce the behaviour which he might consider "appropriate and permissible". This is not to say that pornographic speech does not affect its readers' behaviour and values, although, incidentally, whether pornography does cause negative thoughts, attitudes and acts towards women is a matter on which 13. Id. at 109. 14. Id.

DISCRIMINATION AND ILLOCUTIONARY ACTS

125

considerable doubt has been raised by a number of empirical studies. At best, what can be said is that there is a degree of empirical evidence for the existence of an association between pornography and negative acts towards women, and against such a link. However, even putting these empirical uncertainties aside, the way pornography may affect male audiences is not equivalent to the way in which the decisions of an umpire, whose "verdicts count" by virtue of her authority, affect the players. Rather, pornography may be effective in convincing its readers. This is a par excellence perlocutionary effect of pornography, and, as such, it escapes the analogy with acts of subordination or discrimination. The ambiguity of Langton's statement that the "verdict counts" (or that people are "told" which moves are appropriate, etc.) stems from the fact that her explanation of the way in which the verdict counts is in perlocutionary terms, while her conclusion is in illocutionary terms. This requires further explanation. When a speaker "tells [his or her] hearers which moves are appropriate and permissible", to use Langton's words, the statement may be seen as an illocution or a perlocution. An example of the former may be a set of rules issued by an instructor to her students in the classroom. Because the twin conditions of legitimacy and control are met, the utterance is authoritative. An example of the latter is a message emerging from a popular commentator whose texts are effective enough to affect people's thinking and behaviour. The effectiveness may be (and often is) due to the fact that the message is shared by the ruUng power, and hence by those who enjoy privileged access to mass media, but this is a factor which makes the message more pervasive, not more "authoritative" in the sense of illocutionary force. The ambiguity of the proposition that someone's "verdict counts" is parallelled by the ambiguity of the common uses of the notion of "authority". We often say that someone's opinions are "authoritative", in the sense that he or she is an authority on a certain matter, but this is different from saying that someone is in a position of authority, or has authority over others. Joseph Raz distinguishes between the two meanings by saying that "to be an authority on a certain matter is to be an authority about what to believe rather than about what to do" .15 It is important, for the purposes of our discussion, to note that someone who is "an authority about what to believe" typically affects her hearers through perlocutionary utterances. For example, I consider X to be an authority on house renovations because it is my experience that her judgment and advice on these matters is usually sound: it is not the case that my belief about the soundness of X' s advice is based on my prior recognition of X as an authority. However, perhaps it can be argued that this is an over-simplified and, hence, distorted description of the notion that someone is "an authority on a given matter". Therefore, to the degree to which real-life situations depart from this 15. Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986), p; 64.

126

CHAPTER 4

simplified description, it might be thought that the gap between the two notions of "authority" is often bridged. In particular, it can be realistically claimed that people who are "authorities" on some matters derive their status not merely from the fact that we often consider their judgments to be sound. First, I may consider X an authority on house renovations because many people around. me think this way, and I have been influenced by them to recognize X's "authority" prior to having an opportunity to subject any of X's advice to my own critical scrutiny. Additionally, even if my initial recognition of X' s authority was solely motivated by my belief in the soundness of her judgment and advice, with time, the authoritative status of X becomes an independent ground for trusting her advice. Increasingly, I value her opinions because I consider her to be an authority on the subject rather than establishing her "authority" on a case-bycase basis whenever I find her judgments wise or useful. For these reasons, the recognition of someone's authority on a given matter becomes an independent ground for accepting her beliefs - independent, that is, from the actual soundness of her judgments. But if the "authority", in the sense of being an "authority on a certain matter", is independent of the authority's soundness of judgment, doesn't this undermine the thesis that this notion of "authority" is insufficient to supply an iUocutionary force to a statement? First of all, it is important to realize that this independence (or the separation of the authority from the substance of a judgment) is always qualified. If X's advice on some particular matters turns out to be systematically unsound, she will no doubt lose her status as an "authority" on those matters. More importantly, authority which makes illocutionary force possible has certain features which "an authority on a certain matter" lacks. An example will help demonstrate the difference. In his essay "Performative Utterances", Austin gives the following illustration of one type of "infelicity" which causes an intended authoritative illocution to misfire: Suppose that you are just about to name the ship, you have been appointed to name it, and you are just about to bang the bottle against the stem; but at the very moment some low type comes up, snatches the bottle out of your hands, breaks it on the stem, shouts out 'I name this ship the Generalissimo Stalin', and then for good measure kicks away the chocks.1 6

It is clear, Austin observes, that the ship certainly has not actually been named the Generalissimo Stalin, and at least one good reason for this conclusion is that the authority to name the vessel has been invoked by the wrong person. To see that pornographers are more like this "low type" in terms of their authority to perform effective illocutions, or "performative utterances", rather than like the person who was originally supposed to name the ship, it is useful to invoke two 16. J.L. Austin, Philosophical Papers (Oxford: Clarendon Press, 1970, 2nd ed.), pp. 239-40.

DISCRIMINATION AND ILLOCUTIONARY ACTS

127

interrelated distinctions made by another important philosopher of language, John Searle. In the process of describing illocutionary acts as rule-governed and conventional, Searle introduces a distinction between "constitutive" and "regulative" rules. 17 Regulative rules apply to a pre-existing activity (examples include a set of rules of etiquette or rules of fishing), while constitutive rules create the possibility of a form of behaviour, so that to say that a particular ruie governs a particular behaviour (say, a game of chess, or a practice of promising) is an analytical truth. In the latter case, one simply cannot describe a given practice without stating its rules. A typical form that a constitutive rule takes is "X counts as Y in context C' .18 Building upon this initial distinction, Searle goes on to further distinguish between ''brute'' and "institutional" facts. The latter presuppose the existence of certain institutions, that is, of systems of constitutive rules,19 and the former are typically accessible simply through our empirical knowledge. Performing speech acts, Searle concludes, is an institutional fact because speaking a language consists of performing acts according to constitutive rules. Applying this proposition to our discussion, one may say that the rules of felicity, the compliance with which makes the illocutions effective (including the rules regarding the authority of a speaker), are constitutive, and illocutionary acts, including the acts of subordination, are institutional facts. Referring back to Austin's example, the "low type" who tried to name the ship was unsuccessful because, under the rules governing the procedure for naming ships, he had no authority to do it. This is an explanation given purely in illocutionary terms, and references to any possible influence of his statement upon the audience's beliefs (whether they were shocked or amused, surprised or satisfied) are irrelevant in explaining why his action did not constitute the naming of the ship. It must be admitted that, in reality, the distinction between illocutions and perlocutions is not as sharp as the preceding discussion might suggest, and one way of illuminating the borderline between them is by referring back to the distinction between constitutive rules and regulative ones. After all, we have just seen how this distinction is instrumental in supporting Searle's concept of illocutions. But, as some theorists have noted, the distinction between constitutive and regulative rules may be questioned; as Schauer points out in his book on the theory of rules, "many constitutive rules have their regulative side".2o This is because even rules which establish the very possibility of certain practices or institutions, such as games, further serve to regulate behaviour within these practices and institutions. As Schauer explains, "these constitutive 17. John Searle, Speech Acts: An Essay in the Philosophy of Language (London: Cambridge University Press, 1969), pp. 33-42. 18. Id. at 35. 19. Id. at 50-53. 20. Frederick Schauer, Playing by the Rules (Oxford: Clarendon Press, 1991), p. 7.

128

CHAPTER 4

rules lose their constitutive character within those institutions, serving instead to regulate antecedently defmed behavior".21 To take again the example supplied by Austin, those same rules which establish the practice of "naming the ship", without which the very practice cannot exist, further regulate proper ways of conducting the practice. A person otherwise authorized to "name the ship" may make certain wrong moves which will render the whole ceremony less than accomplished but will not necessarily deprive it of the character of the "naming of the ship". In this case, the rules that the namer of a ship violates will function as regulative rules. Similarly, the constitutive rules of christening a child will, within the ceremony, act as regulative ones: even if a priest offends some of the detailed rules, the ceremony will not necessarily lose its character qua the christening, even though the disgruntled parents may demand that it be repeated, perhaps by a more experienced priest who will perform it lege artis. Schauer's remark helpfully indicates that those same rules may operate at two levels: at one level they establish certain practices, in which case they operate as the constitutive ones, but once the practice has been established, they may then operate at a lower level, within the practice, and their function then becomes regulative. But it should be observed that not all rules are equally capable of performing these two functions. Some constitutive rules are of such an all-or-nothing character that they are substantially confined to a constitutive function. Hence their violation results in a total failure to perform a practice rather than in an improper practice. Typically, rules about authority, contrary to the rules concerning procedure, have such a categorically constitutive character. A person who has no authority to name the ship fails to even begin the naming; his action does not count as belonging to the practice at all. The same can be said about a non-priest trying to christen a baby, a non-marriage celebrant performing an act which looks like a marriage ceremony, or someone who is not an umpire calling a penalty in a soccer match. These acts, regardless of the procedural and substantive standards on the basis of which they are performed, do not even meet the threshold criteria for successful practices; they are hopelessly far from meeting the felicity conditions for these particular illocutions. The upshot is that, while in many cases the sharp distinction between constitutive and regulative rules may be misleading, this proviso does not apply to the subject which is currently of interest for this discussion, namely rules governing the authority of speakers. The authority rules have a fairly categorical constitutive character. And just as the man in Austin's example failed to meet the threshold rules of authority to name the ship, the same applies to pornographers and to racist speakers in the contexts relevant to our discussion, that is, not to face-to-face verbal assaults, etc. Their speech acts do not count as verdicts on matters of sex or race, regardless of whether or not they have an 21. ld.

DISCRIMINATION AND ILLOCUTIONARY ACTS

129

important impact upon the audience. It is an "institutional" fact (rather than a ''brute'' fact) that they do not have such authority, and our conventions disqualify their utterances from counting as verdicts. This is not to say that they do not influence a number of people into thinking about racial matters or sex in a way which is degrading to minorities and women, and which may lead those people to commit acts of subordination and discrimination. But this influence can be characterized as a perlocutionary function of hate speech and pornography, rather than being attached to illocutions which constitute acts of subordination in themselves. At this stage, a critic could make a rather obvious objection to the above argument by asking: ''How do we know whether the pornographers indeed lack the authority, under the conventional rules which make up the institutional facts pertaining to the codes of sexual conduct in our societies, to issue illocutions that subordinate women?" A similar question could be asked about the authority of racist speakers. The points made above concerning pornographers' and racists' lack of authority can be seen merely as assertions, not as arguments. It is useful at this point to consider a preliminary question about what such an argument would be like; in other words, what sort of evidence would convince us whether or not to view pornographers as authoritative in this regard. It should be clear by now which arguments are not convincing. In her essay on the subordinating functions of pornography, Langton uses evidence which pertains to perlocutionary effects of pornography, and which addresses what the readers of pornography (and how many of them) learn about sex from pornography. In support of her proposition about pornography being "authoritative" toward its consumers, she cites the alarmingly high number of male survey respondents who say various outrageous things about women and about sex. She concludes: "In this domain, and for these hearers, it may be that pornography has all the authority ofa monopoly".22 If her evidence is indeed only about the effectiveness of pornography's impact upon readers' beliefs, what evidence do we need to establish the authoritative position of pornographers? Undoubtedly, such evidence would collapse into empirical arguments about prevailing community perceptions. To anticipate a possible objection: this does not blur the line between ''brute'' and "institutional" facts (and, for that matter, does not collapse into arguments about the way those speech acts affect the beliefs of hearers). If the rules which underlie illocutionary acts are of a constitutive nature, then one way to find out about the existing conventions of a particular practice is to ascertain the prevailing perceptions about these conventions. There may, of course, be other, simpler, ways to find out about various institutional facts; sometimes we only need to read the constitution, a statute, a prayer book, or a set of written rules of chess or football. However, not all conventional practices are based on rules as 22. Langton, p. 110, footnote omitted.

130

CHAPTER 4

formally codified as these, and it would be wrong to attempt to reject the thesis that pornography is a fonn of discrimination against women merely by showing that there is no statute which authorizes pornographers to issue codes of discriminatory conduct. In inquiring into these less fonnalized institutional facts (and into constitutive rules which make certain illocutions possible), we need to detennine the prevailing social perceptions about authority. In Austin's shipnaming example, even if we were unable to identify any fonnal source of a rule (something which we may assume arguendo), prevailing perceptions about the legitimacy of who can and who cannot name a ship allowed us to dismiss the "low type's" attempt. Whose perceptions should count? And perceptions about what? There are two possible answers to the first question; either the perceptions of society at large, or those of the actual hearers of a given type of speech. The fonner answer, of course, would greatly facilitate our conclusion, but Langton may be correct in saying that the relevant constituency for judging the authoritative nature of pornography illocutions consists only of consumers of pornography: "What is important is whether [the speech of pornographers] ... is authoritative for the hearers that count: people, men, boys, who in addition to wanting 'entertainment,' want to discover the right ways to do things, want to know which moves in the sexual game are legitimate".23 I will accept Langton's definition of the relevant audience - not because it is uncontroversial, but for reasons of argumentative economy. If under this test the argument about pornographers' authority does not succeed, then this is all the more reason why it cannot succeed under the broader test. . The second question is more difficult. What is the relevant subject-matter of perceptions which serve to identify constitutive rules of authority? Surely, these cannot be perceptions about what is right and what is wrong in matters of sex, and, therefore, whether the ideology of pornographers is accepted by the readers of pornography. The argument arising from such evidence would be either perlocutionary or circular. We may assume that many of those readers do accept pornography's viewpoint; indeed, it is crucial for anti-pornography advocates that most readers accept it. Note, incidentally, that at times MacKinnon suggests that pornography does not express a viewpoint, and does not put forward any argument: "Pornography consumers are not consuming an idea any more than eating a loaf of bread is consuming the ideas on its wrapper or the ideas in its recipe. . . . [p]ornography does not engage the conscious mind in the chosen way the model of 'content,' in terms of which it is largely defended, envisions and requires.,,24 However for the argument about pornography's illocutionary force it is crucial to accept that pornography proclaims, as Langton says, "the right way to do things", and that it declares ''which moves in the sexual games 23. Id. 24. MacKinnon, p. 16.

DISCRIMINATION AND ILLOCUTIONARY ACTS

131

are legitimate", etc.25 Clearly pornography without a viewpoint cannot discriminate. If pornography consumers accept the pornography viewpoint (for example, that "violence is sexy and coercion legitimate"),26 because they are so influenced by the pornographic materials, then this is a perlocutionary argument. But if they accept these viewpoints because they had already been predisposed to accept them, then the argument about pornographers' authority is question-begging. Further, if the actual reality of pornography's impact upon the consumer's mind is halfway between conftrming the predisposition and implanting a new viewpoint, as is probably the case with most consumers of pornography who accept a viewpoint depicted by Langton, then we are no closer to an illocutionary model than either of these alternatives is. In paradigm illocutionary situations, satisfying ourselves about the authority of speakers does not involve asking whether the audience accepts the speech on its merits. In Austin's example, the fact that we did not like the name given to the ship by a "low type" did not ftgure among the reasons why the illocution misftred. Sources of authority are not, or at least are not critically, related to substantive agreement with the contents of speech-acts. Instead, sources of authority are typically related to two types of considerations. The ftrst concerns the actual recognition of a speaker's legitimate right to issue a verdict qua a verdict. It is a necessary condition of the felicity of illocution that we (that is, the addressees of an illocution) see a speaker as having a right to issue a particular type of illocution, such as an order, a warning, or the christening of a baby. A clerk "giving an order" to his superior, a baby "warning" her parents, and an accidental passer-by "christening" a child all fail to perform an illocution because they do not meet this condition of felicity. Second, it is also a necessary condition of an illocution that it be issued with a special kind of intention peculiar to this kind of act, and with the purpose that this intention be recognized as such by a hearer. The illocution is effective when that intention is recognized by a hearer. Searle explains this using the example of the illocutionary act of greeting people: Uttering 'Hello' and meaning it is a matter of (a) intending to get the hearer to recognize that he is being greeted, (b) intending to get him to recognize that· he is being greeted by means of getting him to recognize one's intention to greet him, (c) intending to get him to recognize one's intention to greet him in virtue of his knowledge of the meaning of the sentence 'Hello'.27

The combination of these two necessary conditions constitutes a sufftcient set of conditions for speakers to be in a position to issue illocutionary statements. We 25. Langton, p. 110. 26. Id. 27. Searle, p. 49.

131

CHAPTER 4

must now apply these conditions to putatively discriminatory illocutions issued by pornographers. We have previously accepted that the perceptions by consumers of pornography are decisive in this regard. As a result, the two questions we must ask are: (1) Do regular readers of pornographic magazines, viewers of pornographic movies etc, recognize the pornographer's right to ~ illocutionary statements regarding "proper" conduct toward women in sexual matters? And (2) do these consumers of pornography recognize that it is the pornographer's intention to issue such rules of conduct? If these two questions have an air of absurdity to them, per~ps it is because even the very supposition that the answer to either of these questions could be affirmative sounds so deeply counterintuitive. It seems very plausible that people who read pornographic magazines, watch pornographic movies, etc., do so not because they believe that the authors, publishers and producers of pornography have special insight into matters of sexuality, which privileges them in defining the code of sexual conduct. Rather they consume pornography in order to become sexually aroused. It is not easy to confirm this point by reference to empirical studies. While the number of empirical studies concerned with the relationship between pornography and behaviour is very large, it is difficult to find evidence of research directly concerned with pornography consumers' opinions about producers of pornography. Indeed, it seems that there has been hardly any research into the motivations for individual decisions to reach for pornographic materials. The surveys that have been done seem to confirm that pornography is hardly seen as a credible source of expertise on matters related to sexuality. One survey of the impact of pornography as a source of adolescent sexual information found that both male and female adolescents ranked pornography as their least important source of sexual information.28 Recognition of the pornographer's special "authority", or of his intention to issue a set of rules of conduct in sexual matters, does not seem to be a plausible description of the attitude held by pornography's consumers toward its producers. To be sure, in the process they may become convinced that what they see is an accurate depiction of female sexuality (for instance, that women enjoy pain in sex), but this is a matter ofperlocutionary effect. In these last several pages I have focused mainly on the issue of pornography because it was the subject-matter of Langton's important essay which made a relevant argument from a theory of speech acts. The same reasoning applies to much racist hate speech, and, in particular, to expressions of racism in a generalized, abstract form, as opposed to specific denials of opportunities in verbal form or person-to-person verbal assaults. Racists may well belong to a 28. P.G. Tjaden, "Pornography and Sex Education", Journal of Sex Research 24 (1988): 208212; see also W.A Fisher and A. Barak, "Pornography, Erotica, and Behavior: More Questions Than Answers", International Journal ofLaw and Psychiatry 14 (1991): 65-83, p.

70.

DISCRIMINATION AND ILLOCUTIONARY ACTS

133

dominant social group, and so constitute part of a "ruling power" (in Langton's words), just as pornographers do. By itself, inclusion in the dominant social group does not make all of an individual's pronouncements authoritative or verdictive. Whether speakers affect the way in which hearers treat members of a disadvantaged class is a matter of the persuasive force of these pronouncements upon the hearers, rather than of a combination of legitimacy and control. This may actually be clearer in the case of racist hate speech than in the case of pornography if MacKinnon's observation that racist speech "works more through its content" than pornography does is true. 29 As she further explains, "there is no evidence that consumers of racist propaganda aggress against the target of the literature whether or not they agree with the positions it takes",30 which, she claims, makes it different from pornography where there is no question of any argument at all. The argument of this chapter does not rest upon the proposition that racism ''works through thought,,31 more than pornography does; as noted earlier, if one accepts MacKinnon's thesis that pornography is viewpoint-free, then it is very hard to maintain that pornography teaches its consumers wrongful lessons about sex, relations between men and women, etc. But if MacKinnon's thesis is indeed correct, and racism is more thought-based than is pornography, then it would be a good reason for viewing the subordinating effects of hate speech (but not necessarily of pornography) in terms of perlocutionary, rather than illocutionary acts. Resolution of this problem may hinge upon complex empirical studies about the actual motivations of those who seek to receive racist speech and pornography and, to this degree, is beyond the frames of this book. What matters more for us here is the structure of the argument about the illocutionary character of speech-acts which may render these speech-acts discriminatory per se; rather than a judgment about which particular types of speech can be so characterized in contemporary societies.

29. MacKinnon, p. 56. 30. Id. at 62, footnote omitted. 3 J. Id. at 63.

CHAPTER 5 VIEWPOINT NEUTRALITY AND ITS RATIONALES 1

Two Types of Neutrality

The notion of ''viewpoint neutrality" may partly owe its impact to the popularity of the notion of ''neutrality'' in other contexts of liberal political and legal theory. For example, in the context of the relationship between state and religion, the notion of state neutrality towards religion has often been thought to be the best interpretation of the complex set of rules governing the relationship between law and religion in liberal-democratic states. As Sidney Hook wrote some years ago, "a genuinely democratic state, especially one which contains a plurality of religious faiths, should be neutral in matters of religion, and regard it as essentially a private matter".l ill the United States, the Supreme Court has repeatedly asserted its commitment to "a scrupulous neutrality by the State, as among religions, and also between religious and other activities".2 To be sure, this ideal of state neutrality towards religion is far from being uncontroversial. ill a 1985 decision, the then Justice Rehnquist opined that "nothing in the Establishment Clause requires government to be strictly neutral between religion and irreligion".3 Nevertheless, it is certainly a well entrenched idiom within this strand of doctrinal and judicial opinions which emphasize the need to maintain as strict a separation as possible between the secular state and religious beliefs and institutions, and which demand that the law should be prevented "from entering the purely private domain of religious expression and belief'.4 More generally, the ideal of a liberal state which is neutral towards moral conceptions of the good is often seen as the best interpretation of a liberal state committed to the protection of individual liberties. The underlying idea resonates with the "harm to others" principle, namely, that the liberal state must not, as a matter of principle, enforce official views about immoral (as opposed to harmful) conduct, and must not protect individuals against self-inflicted

1. Sidney Hook, Religion in a Free Society (Lincoln: University of Nebraska Press, 1967), p. 27. See also Wojciech Sadurski, "Introduction" in Law and Religion ed. by Wojciech Sadurski

(Aldershot: Dartmouth, 1992), pp. xi-xxiii. 2. Roemer v. Board of Public Works, 426 U.S. 744, 746-47 (1976). For other examples of judicial appeals to the concept of neutrality, see Zorach v. Clauson, 343 U.S. 306, 314 (1952); Epperson v. Arkansas, 393 U.S. 97, 103 (1968); Gillette v. United States, 401 U.S. 437,449 (1971); Jones v. Wolf, 443 U.S. 595, 602 (1979). 3. Wallace v. Jaffree, 472 U.S. 38, 113 (1985) (Rehnquist, J., dissenting). 4. Jonathan Weiss, "Privilege, Posture and Protection: 'Religion' in the Law", Yale Law Journal 73 (1964): 593-623, p. 623.

136

CHAPTERS

harms. While there is no room here to explore this idea any further,s I wish to note that the idiom of neutrality has a strong foothold in the general liberal theory oflaw, and in some areas other than freedom of speech. In a 1972 decision of the United States Supreme Court concerning the validity of an anti-picketing city ordinance, Justice Marshall wrote these words which became part of the canon of First Amendment literature: "[A]bove all else, the First Amendment means that government has no power to restrict expression because of its message, its -ideas, its subject matter, or its content".6 Taken literally, these words are so exaggerated that they collapse into nearabsurdity. Virtually all restrictions (and note that the above-quoted principle announced by Marshall applies not only to outright bans but also to less prohibitive restrictions) which are commonly accepted by even the most orthodox civil libertarians would have to be invalidated under this formula: the government would not be permitted to restrict blackmail, insider trading, disclosure of military secrets, incitement to crime or malicious defamation. The only restrictions which would survive would be those analogous to limits upon the size of roadside billboards or upon the level of noise emitted by loudspeakers. That Marshall might actually have meant what he said seems to be confirmed by another extraordinary hyperbole which follows soon after the above-quoted formula: "Any restriction on expressive activity because of its content would completely undercut the 'profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide open,,,.7 There are at least three reasons why Marshall could not, and did not, mean to announce a complete ban on content-related restrictions on expression. These reasons are evident not only in the First Amendment jurisprudence which directly preceded and followed Mosley, but also in Marshall's opinion in Mosley itself. First, as I noted in Chapter 2.2, any system of protection of freedom of expression must accept a hierarchy of categories of speech which call for varying degrees of legal protection. It is clear that a law which attached equal importance to a political speech in an election campaign and to a graphic depiction of sexual intercourse with a minor by a paedophile, would be morally unacceptable. This is not to say that the latter expression must necessarily be banned but rather that the harm-related arguments in favour of the restriction would at least have greater weight in the latter case than in the former. One of S. I have discussed the relationship between the "harm to others" principle and the idea of moral neutrality of the liberal state in my Moral Pluralism and Legal Neutrality (Dordrecht: Kluwer 1990), chap. 4, and in "The Moral Neutrality of the Liberal State", Synthesis

Philosophica 9 (1994): 101-14. 6. Police Department of Chicago v. Mosley, 408 U.S. 92, 9S (1972). 7. 408 U.S. at 96. quoting New York Times v. Sullivan, 376 U.S. 2S4, 270 (1964), emphases added.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

137

the arguments put forward in Chapter 2.2 above, was that an attempt to disregard a hierarchy of categories of speech and apply a uniform level of protection to all speech would inevitably result in a lower level of protection across the board. In the words of Justice Blackmun: "If all expressive activity must be accorded the same protection, that protection will be scant". 8 In any event, First Amendment doctrine relies upon a crucial distinction between "protected" and "unprotected" speech and, within the former class, upon the ranking of different categories of speech which deserve varying degrees of protection. The ranking is based upon various criteria of "content" and, for this reason, Marshall's wholesale condemnation of all content-related restrictions cannot be taken literally. Indeed, the special position of speech related to "public issues", mentioned in Marshall's own quotation from New York Times v. Sullivan, is a prima facie indication that at least some restrictions of expression on "non-public" issues (an arguably content-related description) may be tolerable under the First Amendment. The second reason why the sweeping Mosley formula cannot be taken literally has nothing to do with speech itself, but rather with the strength of the prohibition upon governmental restrictions. The sentence reads as if it contained an absolute ban on such restrictions, and this is reinforced by Marshall's declared hostility towards "{aJny restriction on expressive activity,,9 and by his announcement that a regulation which is not neutral but is content-related "is never permitted".10 But no such absolute ban has ever been in operation under the principle of freedom of speech in any legal system committed to liberty, and certainly not in the United States. Invalidity of content-related restrictions is only presumptive, and the presumption may be rebutted by sufficiently strong arguments about the importance of the governmental purposes served by the restriction and the close fit between the restriction and these purposes. What Marshall presents as an absolute prohibition is merely a manner of describing a very exacting scrutiny of governmental regulations. By its very nature, strict scrutiny suggests the lessthan-absolute character of a principle against content-based regulations. It would be somewhat deceptive to postulate a "strict scrutiny" if one knew in advance that no governmental restriction could ever survive it. Indeed, Marshall's opinion in Mosley explicitly states: "Because picketing plainly involves expressive conduct within the protection of the First Amendment ... discriminations among pickets must be tailored to serve a substantial governmental interest".l1 The specific distinction under challenge in Mosley, R.A. V. v. City 0/ St. Paul, 505 U.S. 377, 415 (1992) (Blackmun, I., concurring). See also Cass R. Sunstein, "Low Value Speech Revisited", Northwestern University Law Review 83 (1989): 555-61, pp. 557-8. 9. Police Department o/Chicago v. Mosley, 408 U.S. 92,96 (1972), emphasis added. 10. Id. at 99, emphasis added. 11. Id. at 99, references omitted.

8.

138

CHAPTERS

that is, the distinction between labour and non-labour peaceful picketing, has not survived such a test but, by positing the test, Marshall envisages that under some circumstances a content-related regulation may be upheld. The third reason for not taking the Marshall formula literally is directly connected to the central topic of this chapter. Almost immediately after announcing an absolute prohibition of content-related regulations, Marshall goes on to say: Necessarily, then, ... government may not grant the use of a forum to people whose views it finds acceptable, but deny use to those wishing to express less favored or more controversial views. And it may not select which issues are worth discussing or debating in public facilities. 12

He then describes two ways in which content may improperly enter into legislators' reasons for prohibition. First, legislators may exclude some speakers because of a dislike for their message; second, legislators may exclude some issues as less worthy of public discussion than others. This is, of course, a distinction between viewpoint-based restrictions and subject-matter restrictions, which we initially canvassed in Chapter 2.4. But while Marshall's equal condemnation of both these types of regulation may be valid in the context of a "public forum" (which is a term of art with a narrower meaning than "public facilities"), as was the case in Mosley (a street), it does not apply to those "public facilities" which constitute a so-called nonpubIic forum even though they still are public facilities. For instance, with regard to access to a school mail system, or the advertising spaces on city buses, or military reservations, or the after-hours use of public school premises, the subjectmatter distinctions of United States law are permissible but viewpoint-based distinctions are not. 13 The meaning of this dichotomy, and the reasons for upholding it, constitute the focus of this chapter. I will argue that the distinction is useful but only in so far as it illuminates different types of moral deficiencies associated with the two types of speech restrictions. Once we identify the underlying grounds for those 12. Id. at 96. 13. See Perry Educ. Ass'n v. Perry Local Educators Ass'n, 460 U.S. 37 (1983) (school mail system); Lehman v. City ofShalrer Heights, 418 U.S. 298 (1974) (city buses); Greer v. Spock, 424 U.S. 828 (1976) (military reservations); Lamb's Chapel v. Center Moriches Union Free School District, 508 U.S. 384 (1993) (public school buildings). However, note that as late as 1979, in a case concerning the placement of politically controversial inserts in envelopes containing bills sent by a public utility (therefore, somewhat resembling a non-public forum over which the government has legitimate regulatory interests), the Court proclaimed that "[t]he First Amendment's hostility to content-based regulation extends not only to restrictions on particular viewpoints, but also to prohibition of public discussion of an entire topic", Consolidated Edison Co. v. Public Service Comm 'n, 447 U.S. 530,537 (1980).

VIEWPOINT NEUTRALITY AND ITS RATIONALES

139

restrictions (and my suggestion will be that viewpoint restrictions are typically based on intolerance, and subject-matter restrictions, on paternalism), only then can we recognize the unequal moral concern raised by these restrictions. In order to help organize the discussion in a clear way, I will use as an illustration a United States Supreme Court decision, handed down in June 1995, in which the difference between these two types of content-related considerations acquired a crucial constitutional importance. In Rosenberger v. Rector & Visitors of the University of Virginia,14 the question was whether the denial of funding for a particular religious student newspaper was a subject-matter regulation (as Justice Souter claimed in his dissent) or a viewpoint-based regulation (as Justice Kennedy claimed in the opinion of the Court). This marked the difference between the validity and invalidity of an action under the First Amendment's free speech clause. But it should be emphasized that a critique of a particular court decision is not my purpose here. Rather, I will use it to elucidate the meaning and rationales for a general principle of viewpoint neutrality which is often taken to express the central aim of the principle of freedom of speech. The underlying belief is that, properly understood, a reluctance to regulate speech in a way which disadvantages some viewpoints more than others is a powerful and persuasive indicator of a robust system of freedom of expression. Indeed, the rationale for this idea is so substantial that it is important to keep it distinct from the rationale for content-neutrality, contrary to the Court's decision in Rosenberger which largely blurred the distinction between the two types of neutrality. It should be stressed that the formula of "viewpoint neutrality" is of universal applicability and is not only suited to the specifically American pattern of argument under the First Amendment. Recently a New Zealand author used terms identical to the "viewpoint neutrality" discourse to criticize a New Zealand High Court's decision about a swastika display on a residential house. 15 While agreeing with the Court's decision (which upheld a ban on swastikas), Bede Harris claimed that a similar conclusion could be reached "on considerations of time and place", and without any regard to the content or the viewpoint of the expression. "The issue of viewpoint neutrality is central to freedom of expression, yet was not addressed in the judgment" deplores Harris, who adds: "The result was a decision which, by restricting the ambit of permissible expression to what conforms to prevailing notions of ideological acceptability, quietly legitimated viewpoint censorship". 16

14. 115 S.Ct. 2510(1995). 15. Zdrahal v. Wellington City Council, [1995] 1 NZLR 700 (HC), discussed in Bede Harris, "Viewpoint Neutrality and Freedom of Expression in New Zealand", Otago Law Review 8 (1996): 515-38. 16. Harris, pp. 537-38.

140

Z

CHAPTER 5

What Is "A Viewpoint"? A Case Study of Rosenberger

What does it take for a given expression or a set of expressions to constitute a ''viewpoint"? The question is important because if the principle of freedo~ of speech is taken to dictate a higher hostility toward vie'vpoint-based restrictions than to many other restrictions, we must have an idea about what a ''viewpoint'' is. In particular, we need to know how to distinguish restrictions based on viewpoints from those based on subject matter and from those based on criteria which are altogether content independent, such as the time, place or manner of the expression. At first blush, the notion of viewpoint seems so obvious as to require no further clarification. But this is an illusion, and the Rosenberger decision is a good illustration of the ambiguities and complexities of the distinction between a "viewpoint" and a "subject matter". Rosenberger, which has been hailed as an important step in the Court's "increasingly libertarian and uncompromising line on freedom of speech", 17 was at the intersection of the principle of freedom of speech and freedom of religion. At issue was a policy of the University of Virginia which denied a subsidy to the religiOUS student newspaper Wide Awake: A Christian Perspective at the University of Virginia. The University defended this exclusion under the Establishment Clause of the First Amendment ("Congress shall make no law respecting an establishment of religion',), while the newspaper complained that the refusal of the subsidy violated its right to freedom of speech under the same constitutional amendment. The University subsidized a wide range of student newspapers, however, under its guidelines for granting subsidies it was prohibited from supporting religious organizations and activities, understood as those which ''primarily promote or manifest a particular belief in or about a deity or an ultimate reality". On this basis, the University denied the subsidy to Wide Awake. The Supreme Court (divided five to four) decided that the refusal to subsidize the Christian newspaper indeed amounted to a denial of the right to free speech. Can such a denial constitute an infringement of freedom of speech? An important part of the majority's affirmative conclusion was the characterization of the University's denial of funding as viewpoint discrimination. Rather than openly challenging the legal significance of the viewpoint/content distinction, the Rosenberger Court reinterpreted and broadened the concept of ''viewpoint'' in a way which made it virtually indistinguishable from "content". Kennedy rejected the argument ''that no viewpoint discrimination occurs because the Guidelines discriminate against an entire class of viewpoints".18 He claimed 17. Charles Fried, ''The Supreme Court, 1994 Term - Foreword: Revolutions?", Harvard Law Review 109 (1995): 13-77. p. 68. 18. 115 S.Ct. at 2518.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

141

that this is a fallacious argument, which is based upon "an insupportable assumption that all debate is bipolar and that anti-religious speech is the only response to religious speech".19 But is the former conclusion necessarily based upon the latter "insupportable assumption"? And why deem the assumption "insupportable"? Consider a case where a Muslim view on a given subject is restricted while a Christian perspective is allowed. Surely this would constitute objectionable viewpoint discrimination? And, from the point of view of freedom of speech, the restriction would be as speech-threatening as the prohibition of a religious response (which may comprise "an entire class of viewpoints") to an atheistic point of view. This is uncontroversial, but how does it defeat the "insupportable assumption" of "bipolarity"? Kennedy's example does not help solve the puzzle: "If the topic of the debate is, for example, racism, then exclusion of several views on that problem is just as offensive to the First Amendment as exclusion of only one".20 Yet this is not a proper characterization of the alternative to the view defended by Kennedy. The alternative is not so much an "exclusion of several views on that problem" but the removal of the subjectmatter from the permissible agenda. If the University decided that the discussion of race would not have the financial backing of the University, then it may impoverish public debate at the University but it would not involve the sort of partisanship or censorship which is the evil targeted by the principle of viewpoint neutrality. The objectionable position would be if, for example, the University expressed its disapproval and thus refused funding of a newspaper which argued for differential treatment on the basis of an alleged genetic inferiority while supporting (by funding) 'all other views on the issue. What Kennedy probably means is that religious and anti-religious viewpoints are not the only viewpoints one can have on a given issue, and if there are some other viewpoints left, then an even-handed restriction upon religious and anti-religious viewpoints still constitutes viewpoint discrimination because it does not cover the entire subject matter. But obviously a dispositive issue here is the characterization of the "subject matter" in this particular case. If the subject matter is religion itself, then religious and anti-religious (and agnostic) viewpoints do cover the field. If the subject matter is race, then racist and anti-racist viewpoints cover the entire subject matter. The nature of the regulation (that is, whether it is a content- or a viewpointregulation) seems to depend upon whether all viewpoints on the topic are equally prohibited from the forum or whether some can be expressed while others are disfavoured or suppressed. As will be shown below, this is more difficult to establish than it sounds. But the difficulty has nothing to do with

19. Id. 20. Id.

142

CHAPTERS

the alleged "bipolarity" of the view which equates an even-handed exclusion of all the main viewpoints with a restriction on content. However, the issue at hand is religion, and it is far from irrelevant to the discussion. Kennedy returns to the topic directly after the just mentioned racism example and states: "It is as objectionable to exclude both a theistic .and an atheistic perspective on the debate as it is to exclude one, the other, or yet another political, economic, or social viewpoint".21 Note, however, that "exclusion of both a theistic and an atheistic perspective" is. equivalent to the suppression of all debate about religion. This is another manner of speaking about the removal of an entire topic from the agenda. Whether it is objectionable depends on two issues: (1) the wisdom of the decision that a particular topic lends itselfto harmful consequences (such as divisiveness) and, (2) the institutional authority of a particular body to make and enforce such a judgment. In the specific setting of the Rosenberger debate, the crucial issue is, of course, whether the financial sponsorship of a debate would implicate the Establishment Clause of the First Amendment which prohibits "an establishment of religion" by law. But these are issues which we can put aside for the purposes of our current discussion because one thing that is clear is that such an "exclusion", if it is objectionable, is objectionable for reasons other than "the exclusion of one, the other, or yet another political" viewpoint. Depending upon the context, suppression of the subject matter may be more or less objectionable than skewing the debate by disfavouring one particular viewpoint or a number of viewpoints which do not exhaust all the viewpoints on the subject that are being expressed. But, in any event, an altogether different harm is inflicted on freedom of speech by shrinking the agenda. Consider, for example, a debate about sex. A university announces that it will not subsidize student newspapers which have, as their primary agenda, the promotion of debate about sexuality. The restriction would strike us as patently silly, but it is useful to compare this hypothetical decision with a hypothetical refusal to support periodicals advocating a "permissive" perspective on sex while at the same time supporting journals manifesting a conservative approach. The difference is clear. While the former regulation may result in the impoverishment of public debate in the given community (in this example, a university), the latter restriction actually skews the debate in a particular direction: an authoritative body allocates public resources to support one side of the controversy thus facilitating its success in the battle for the minds of readers. As Geoffrey Stone says, "by effectively excising a specific message from public debate, [a viewpoint discriminatory law] mutilates 'the thinking

21. Id.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

143

process of the comrnunity"'.22 A university where publications avoid any mention of sexual issues (in order not to lose eligibility for funds) may be an anomaly, but the anomaly does not favour any of the parties to the controversy. It merely deprives all the parties of one possible forum for a debate in which they would otherwise engage. This is a manifestation of paternalism on the part of the university because its underlying notion is that the restriction of the scope of subsidized speech is for the benefit of those whose speech is regulated in this way. But (as will be discussed in more detail in Section 4 of this chapter) it is not a manifestation of intolerance because no restriction of disliked viewpoints takes place. Martin Redish refutes the "skewing of the public debate" argument by claiming that content-neutral restrictions may have a differential impact upon the competing views in a particular controversy: Individuals who have heard one side of an issue may well be precluded from learning the other by content-neutral restrictions. This is because we have no way of ascertaining in what ways and at what times individuals are exposed to different facts or opinions. While such regulations do not focus upon the expression of a single opinion, or impede expression of the same viewpoint at all times, erratic and unpredictable distortion of the marketplace is no less a distortion. 23

The argument misses the point. When a restriction limits public expression in a content-neutral manner, the pattern of distortion is "unpredictable": it may just happen that viewpoint A will be hit harder than viewpoint B. But if such a disproportionate impact is a matter of random effect, if it could not be anticipated, and if it does not result from a deliberate attempt to discriminate against a viewpoint through a facially neutral restriction, then it is perhaps unfortunate but not unfair. There are, however, circumstances in which we do have a way "of ascertaining in what ways and at what times individuals are exposed to different facts and opinions". If we then tailor a content-neutral restriction to fit this knowledge in order to achieve a specific result in terms of favouring one side of the controversy, then it is an example of an indirectly viewpoint-discriminatory measure. Consider a restriction on leafleting, supported by arguments about reducing litter on the streets. It may occur that, when leaflets are intended to be circulated on a particular controversy, more people who are likely to be persuaded to viewpoint A will just happen to 22. Geoffrey R. Stone, "Content Regulation and the First Amendment", William

& Mary Law Review 25 (1983): 189-252, p. 198, quoting Alexander Meiklejohn, Political Freedom (New

York: Harper, 1960), p. 27. 23. Martin Redish, "The Content Distinction in First Amendment Analysis", Stanford Law Review 34 (1981): 113-51, p. 131. Note that, in this particular argument, Redish lumps viewpoint-based and subject-matter-based distinctions together, and compares them with content-neutral distinctions.

144

CHAPTERS

receive leaflets with viewpoint A expressed, and be persuaded. This is the scenario Redish describes. But it is indistinguishable from a scenario of no restrictions because, where there are no restrictions, it may also just occur that more people will be exposed to viewpoint A with the result of being persuaded to it. This is significantly different from a scenario in which a leafleting ban can be anticipated to have a differential effect, for example, because one of the parties to a current public controversy cannot afford other means of propagating their cause and may be much more adversely affected by this particular ban than the party which espouses an opposing viewpoint. The discussion so far has assumed that no outright prohibition of a given expression is envisaged: the favouring of the expression is through a subsidy and the disfavouring is "merely" through denial of a subsidy. In consequence, although there may be a debate on the disfavoured subject, the speakers will not benefit from the support (such as a financial subsidy, or making public premises available) that speakers on other subjects may get. This fact takes away much of the gravity of the limiting of the agenda. At the same time, skewing the debate by imposing a viewpoint constraint strikes us as reprehensible because the deplorable aspect of viewpoint discrimination is the symbolic (as much as any other) endorsement by the authority of one side of the controversy. The upshot is that, when no outright prohibition is involved (as in Rosenberger), the contrast between the low moral significance of content discrimination and the high moral significance of viewpoint discrimination is more acute. This partly explains why Kennedy has a rhetorical stake in blurring the line between these two types of speech regulation. Kennedy suggests that "discrimination against one set of views or ideas is but a subset or particular instance of the more general phenomenon of content discrimination" and that "the distinction is not a precise one".24 The rhetorical gain entails embracing the moral outrage caused by viewpoint discrimination against a regulation which, superficially at least, appears to be based on subject matter. But the strategy of blurring the borderline between these two types of regulation has its costs and Kennedy's position is rendered unusually hard to defend because Rosenberger is a case about more than freedom of speech. As noted earlier, Rosenberger is at the intersection of the principle of neutrality of the state towards religion (with the Establishment Clause as its constitutional expression) and the principle of freedom of speech, and these two principles pull in opposite directions. I have already indicated why Kennedy insists on the uncertain character of the content/viewpoint distinction. If a regulation which is prima facie contentbased can be characterized as being viewpoint-based because of the uncertain distinction between the two, then all the serious moral objections generated by 24. Rosenberger, 115 S.Ct. at 2517, emphasis added.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

145

viewpoint discrimination, which are largely absent in cases of content restriction, apply. But here is the conundrum faced by Kennedy: the more he insists on the viewpoint-based character of a regulation, the more likely it is that the regulation will be mandated by the Establishment Clause. If the denial of a subsidy to a student journal was in fact viewpoint based, then it must mean that the journal expressed an identifiable viewpoint on religion. But expressing a clear viewpoint on religion is equivalent to proselytizing, that is, advocating one "correct" perspective on religious matters, in either a pro-religious or an anti-religious manner. A subsidy to a periodical characterized as advocating a particular religious perspective obviously clashes with the principle of religious neutrality of the law. The dilemma for Kennedy is that he cannot have it both ways. He cannot characterize a student periodical as being merely about religion, for the purposes of the Establishment Clause analysis, so that a direct subsidy would not destroy the religious-neutrality credentials of the subsidizer, whilst, at the same time, characterizing the journal as advocating a particular religious viewpoint so that refusal of a subsidy would make the would-be subsidizer vulnerable to the charge of supporting one viewpoint in a religious debate. One possible lesson to be drawn from this conundrum is that there is an inherent tension, alluded to above, between the principle of non-establishment of religion and the principle of freedom of speech. This is not surprising. It is a reflex of an internal tension between the two Religion Clauses in the First Amendment: the Establishment Clause and the Free Exercise of Religion Clause. 25 The Free Exercise Clause demands that the state facilitates an individual's pursuit of his or her religious activity if secular duties hinder such an exercise, while the Establishment Clause disables the state from allocating its resources, or imposing rights and duties, based upon the religious status of an individual. By analogy to the Free Exercise Clause, the Free Speech Clause may demand that the government facilitates religious speech to the same degree as any other speech and this is precisely what the Establishment Clause prohibits. But the prohibition makes sense if the speech is characterized as expressing a religious viewpoint rather than simply discussing religious issues. Government support of speech which is merely about religion need not implicate the Establishment Clause as it does not raise the spectre of favouritism. That is why it is so important to see whether the contrast between 25. For a discussion of this clash, see Sadurski, "Introduction", pp. xi-xxi; see also, generally, Gerald Gunther, Constitutional Law (Westbury, N.Y.: Foundation Press, 1991, 12th ed.), pp. 1501-02. The Court of Appeal's decision in Rosenberger v. Rector & Visitors of the University of Virginia rested upon the apprehension of a collision between the Establishment Clause and the Free Speech Clause, and the judgment that the University'S refusal to subsidize Wake Awake "serve[d] the compelling interest of preventing the University of Virginia from an excessive entanglement with religion", 18 F.3d 269, 288 (4th Cir. 1994).

146

CHAPTERS

the two types of speech restrictions can be maintained. Kennedy's uncertainty about the distinction is evident from his ambiguity on the "religion as a viewpoint" issue. On the one hand, he ftrmly declares that a religious perspective is a viewpoint when this characterization serves his aim to invalidate the University's decision to deny funding to Wide Awake. But, on the other hand, he admits that "[i]t is, in a sense. something of an understatement to speak of religious thought and discussion as just a viewpoint. as distinct from a comprehensive body of thought".26 He would conclude. nevertheless. that "viewpoint discrimination is the proper way to interpret the University's objections to Wide Awake", because the University did not "exclude religion as a subject matter" but rather "select[ed] for disfavored treatment those student journalistic efforts with religious editorial viewpoints".27 This sounds like an ad-hoc judgment. We are told that religion is not "just a viewpoint" but we are also told that it can be characterized as a viewpoint for the purposes of this decision. But, as we will see below, religion can be represented either as a subject-matter or as a viewpoint depending on the aims one wishes to achieve, and keeping in mind the greater likelihood of a legal prohibition of the latter but not of the former. This is because even the most directly proselytizing talk about religion (as a subject matter) can always be characterized as being about something else (salvation, peace of mind, morality, human perfection, child rearing. community duties), with religion merely providing a "perspective". Therefore, what could otherwise be properly regulated under the "subject matter" heading, will easily obtain protection under the prohibition of "viewpoint" regulation. This actually happened in Grossbaum v. Indianapolis-Marion County Building Authority,28 a United States Court of Appeals for the Seventh Circuit decision, partly based upon Rosenberger, in which religious displays in a city hall were not characterized as pertaining to the "subject matter" of "religion", but as providing a "viewpoint" on the "subject matter" of the "holiday season". (I will return to that decision in Section 5 of this chapter). A useful analogy can be drawn with politics. Can political speech be characterized as viewpoint or subject matter related?29 For example, it would seem that "political speech" describes a subject matter because one would be hard-pressed to identify a distinctive "viewpoint" of politics. The exclusion of the whole category of "political speech" would impoverish debate but would not necessarily skew it in the sense that some actual protagonists in actual 26. 115S.Ct.at2517. 27. Id. 28. 63 F.3d 581 (7th Cir. 1995). 29. See, similarly, Luba L. Shur, "Content-Based Restrictions in a University Funding System and the Irrelevance of the Establishment Clause: Putting Wide Awake to Rest", Virginia Law Review 81 (1995): 1665-1720, pp. 1698-99.

VffiWPOINT NEUTRALITY AND ITS RATIONALES

147

discussions and controversies would gain an unfair advantage over others. In any event, the Supreme Court of United States has characterized the exclusion of political speech from various fora as a subject-matter regulation when it declared such exclusions constitutional. For example, it upheld a municipal public transit system ban upon political advertisements on its trains, while an array of other advertisements were allowed; it upheld a ban on entry by political speakers into a military base, while non political speakers, including clergymen, were allowed; and it upheld an exclusion of political speech from a charity drive in a federal workplace.30 But if "political" describes a subject matter, then why would "religious" describe a viewpoint? As a result of such a characterization, political speech would receive inferior protection compared to religious speech. It would sharply conflict with numerous declarations that political speech ranks highest on the hierarchy of speech under the First Amendment, notwithstanding that the Establishment Clause seems to mandate at least some avoidance of governmental support for religious expression. Obviously the two descriptions, "the political" and ''the religious", are open to manipulation. Just as is the case with ''translating'' religious contentrestrictions into viewpoint-restrictions (as mentioned above), one may also turn an exclusion of political speech into a viewpoint exclusion by saying that this is speech about something other than politics (for example, about health care or military defence) but viewed from a ''political'' perspective. The conclusion one might draw is not necessarily that the distinction is worthless but rather that it cannot be drawn unless one compares the evil captured by the "content regulation" category with that described by the "viewpoint regulation" category. This is the burden of Justice Souter's argument in the section of his dissent which addresses the Free Speech Clause. He is at pains to emphasize that the relevant inquiry in Rosenberger is "not merely whether the University bases its funding decisions on the subject-matter of student speech" but, rather, whether the University is "impermissibly distinguishing among competing viewpoints".31 He states that, in distinguishing between the two forms of regulation, "the government's purpose is the controlling consideration".32 He also implies that the disproportionate impact upon different viewpoints does not render a regulation viewPointdiscriminatory, as long as the motive for the regulation· is not related to a viewpoint: "the issue ... turns on whether the burden on speech is explained by reference to viewpoint".33 What matters is the rationale for the distinction. But how does the rationale serve to draw the line between viewpoint- and 30. See e.g. Lehman v. City of Shaker Heights, 418 U.S. 298 (1974) (public transit); Greer v. Spock, 424 U.S. 828 (1976) (military base); Cornelius v. NAACP Legal Defense & Ed. Fund, 473 U.S. 788 (1985) (charity drive). 31. 115 S.Ct. at 2548. 32. Id. (quoting Ward v. Rock Against RaCism, 491 U.S. 781, 791 (1989». 33. Id.

148

CHAPTER 5

content-regulations? We will tum to this question in Section 4 of this chapter, but at this stage, we need to consider a preliminary question: what does it take for those regulations to be ''reasonable''?

3

Reasonableness and Viewpoint Regulations: A Case Study of Lamb's Chapel

lbroughout his dissent in Rosenberger, Justice Souter makes a distinction between the constitutional weight, or significance, of content discrimination and viewpoint discrimination. In the context of Rosenberger's facts, Souter states that when a public university's funding determinations are made "on the basis of a reasonable subject-matter distinction, but not on a viewpoint distinction", there is no violation of the Free Speech Clause.34 This does not imply, he says, that every viewpoint-based distinction by a public university violates the Free Speech Clause, but rather, that no ''reasonable'' content-based regulation is ever violative of the Clause. This may seem a rather cryptic suggestion, and the force of the distinction between the constitutional significance of the two types of speech regulation is uncertain. If all reasonable content regulations are permissible and, in addition, if some viewpoint regulations may be permissible as well, then a great deal depends upon the criteria for determining the reasonableness of content regulations. After all, one may suspect that the criterion for the permissibility of viewpoint distinctions in this particular setting (that is, a public university's funding decision) does meet the test of "reasonableness". For example, the suppression of racist, bigoted remarks may be seen as "reasonable" from the point of view of the University's educational mission. But in this case, it would not be a viewpoint/content distinction which would be crucial for the permissibility of the regulation of speech, but a much more open-ended "reasonableness" criterion. This would undermine much of the content/viewpoint distinction because the operative reasonableness criterion would cut across the two. The only hint given by Souter about the reasonableness requirement for a content-based distinction is his citation of a doctrine which he refers to as the "limited-access forum". In order to pass constitutional muster, a restriction must be "reasonable in light of the purpose served by the forum".3S In the classification of various "fora" this had, perhaps misleadingly, been labelled a "non-public forum" in order to distinguish it from the so-called "traditional public forum" and "designated public forum". A brief explanation of the main 34. Id. n. 12. Souter declines to elaborate on how the "reasonableness" of the content distinctions would be evaluated because, as he says, "petitioners have not challenged the University's Guideline as unreasonable". 35. Id. n. 12 (quoting Cornelius, 473 U.S. at 806).

VIEWPOINT NEUTRALITY AND ITS RATIONALES

149

terms of the "forum analysis" is necessary here. The Supreme Court of the United States has developed a "forum analysis" under which "the extent to which the Government can control access depends on the nature of the relevant forum,,}6 In a "traditional" public forum (a place that "by long tradition or by government fiat [has] been devoted to assembly and debate")37 and in a "designated" public forum (one "created by government designation of a place or channel of communication for use by the public at large for assembly and speech"),38 regulations affecting access only survive if they are narrowly drawn to achieve a compelling state interest. But with respect to public property that is neither a "traditional" nor a "designated" public forum, access regulations can be based on subject matter as long as they are reasonable and viewpoint-neutral. The distinction between the second and third category is not very precise. Nor is the terminology. The second category (public forum by government designation) has sometimes been labelled "a limited public forum".39 This is perhaps because, as Justice Blackmun explains in Cornelius, some subject-matter restrictions which would not survive strict scrutiny in a traditional public forum may be necessary to serve a compelling purpose in a public forum created by government designation. 4o But the "limited-access forum" of Souter's dissent in Rosenberger seems to apply to a third category (a non-pUblic forum), as it is clearly indicated by his application of "reasonableness" and viewpoint neutrality tests to such fora, and by the citation of that dictum in Cornelius which describes a test for a nonpublic forum. One should therefore be warned against confusing the "limited public forum" (in Blackmun's dissent in Cornelius, which is the second category in forum analysis), with the "limited-access forum" in Souter's dissent in Rosenberger (which is the third category). The reasonableness of a restriction is directly judged by the relevance of the distinction to the particular purpose of the forum, that is, the purpose which justifies the rendering of the access to the forum "limited". Assuming that the analogy with the limited-access forum holds (and Souter refuses to say whether "the reasonableness criterion applies in speech funding cases in the same manner that it applies in limited-access forum cases"),41 how would the test of reasonableness apply to content-based and viewpoint-based distinctions? Lamb's Chapel v. Center Moriches Union Free School District, the case that Souter cites approvingly in this context, does not go beyond reciting the Cornelius principle that "control over access to a nonpublic forum can be based on subject matter and speaker identity so long as the distinctions drawn 36. Cornelius, 473 U.S. at 800. 37. Perry Educ. Ass'n v. Perry Local Educators Ass'n, 460 U.S. 37,45 (1983). 38. 39. 40. 41.

473 U.S. at 802. See 473 U.S. at 818-819 (Blackmun, J., dissenting). Id. 115 S.Ct. at 2548 n. 12.

150

CHAPTERS

are reasonable in light of the purpose served by the forum and are viewpoint neutral',.42 Since Justice White (who delivered the opinion of the Court in Lamb's Chapel) found the exclusion of a religious group from school premises to be viewpoint-based, there is no elaboration of the ''manner'' in which the ''reasonableness'' test operates. Earlier, in Perry Education Association v. Perry Local Educators' Association, the Court established that in public property which is neither a traditional nor a "designated" public forum, [i]n addition to time, place and manner regulations, the State may reserve the forum for its intended purposes, conununicative or otherwise, as long as the regulation on speech is reasonable and not an effort to suppress expression merely because public officials oppose the speaker's view.43

It further explained that differential access to the interschool mail system provided to rival teacher unions was reasonable because it was "wholly consistent with the [School] District's legitimate interest in 'preserv[ing] the property ... for the use to which it [was] lawfully dedicated. ,,044 Notwithstanding these uncertainties and ambiguities which suggest that viewpoint-based and content-based regulations may in fact be subjected to a uniform standard of "reasonableness", the difference in the weight of viewpoint and content regulations still stands. By analogy to the limited-access forum, one may say that a content regulation is reasonable if it is rationally related to the purpose of the forum (or, as the Court once said, if the topic is "encompassed within the purpose of the forum,,).4s This means that once the purpose is correctly ascertained, the judgment about the reasonableness of a restriction is a relatively straightforward means-ends relationship. A noticeboard in a library has a different purpose to a noticeboard at a police station and what "reasonably" belongs on the former may be properly denied access to the latter. Of course, the more comprehensive the purposes of the forum are, the higher the incidence of the correlation between the content and the purpose. But what would constitute a reasonable viewpoint-based regulation? The relationship to the purpose would mean that the purpose of the institution is defmed by reference to a particular ideology, point of view or opinion. In such cases, the purpose would determine whether an exclusion of particular viewpoints is ''reasonable''. The reasonableness of the exclusion would, in other words, signify hostility by the forum-controlling authority to a particular 42. 508 U.S. 384,392-93 (1993) (quoting Cornelius, 473 U.S. at 806). 43. 460 U.S. 37,46(1983). 44. Id. at 50-51 (quoting United States Postal Service 11. Council of Greenburgh Civic Ass'ns., 453 U.S. 114, 129-130 (1981), quoting Greerv. Spock, 424 U.S. 828, 836 (1976». 45. Cornelius, 473 U.S. at 806.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

151

viewpoint. Such hostility is at odds with the character of fora such as public universities which, under the prevailing liberal ideology, are devoted to neutrality between viewpoints. The restriction, however, would be perfectly reasonable if the institution had an explicit ideological purpose, as is the case with religious colleges. The exclusion of anti-religious teaching from such colleges would be a reasonable viewpoint regulation. But this is no model fot public institutions in a libera1-democratic state. The upshot is that hardly any viewpoint-based regulations would survive the test of reasonableness in public fora and so the distinction between content and viewpoint regulations, as far as their constitutional weight is concerned, is unquestionable. The main moral rationale behind the hostility towards viewpoint regulations is this: "the prohibition on viewpoint discrimination serves that important purpose of the Free Speech Clause, which is to bar the government from skewing public debate".46 Souter further explains that "viewpoint discrimination occurs when the government allows one message while prohibiting the messages of those who can be reasonably expected to respond", and adds that "[i]t is precisely this element of taking sides in a public debate that identifies viewpoint discrimination and makes it the most pernicious of all distinctions based on content".47 Governmental partisanship is thus the main First Amendment sin targeted by the principle of viewpoint neutrality. It is one thing for the government to compress the scope of the debate, though this may be lamentable and speechthreatening because of a reduction in the.overall amount of speech. It is another thing to favour one side of the debate through silencing, or disadvantaging in more indirect ways, likely respondents to the favoured speakers. Yet the problem is that, at times, it may be difficult to tell whether a particular regulation affects the scope of the agenda or the competitive chances of particular viewpoints. The difficulties are displayed in Souter's own characterization of the funding restriction under challenge in Rosenberger. It should be remembered that the University guidelines denied funding to any newspaper which "primarily promotes or manifests a particular belief in or about a deity or an ultimate reality". In the earlier part of his dissent, dealing with scrutiny under the Establishment Clause, Souter quite convincingly showed that Wide Awake was engaged in vigorous proselytizing: it was a preaching newspaper rather than being involved in a "merely descriptive examination of religious doctrine".48 The question is, how is this evidence of the newspaper's distinctive viewpoint to be reconciled with Souter's

46. liS S.Ct. at 2548. 47. Id. at 2548-49. 48. Id. at 2535.

152

CHAPTERS

conclusion that "[t]here is no viewpoint discrimination in the University's application of its Guidelines to deny funding to Wide Awake"?'9 Souter's argument is that while Wide Awake was denied funding because it manifested a religious viewpoint, it was not denied funding because of that particular viewpoint. Any religious viewpoint would disqualify it from a subsidy. He suggests that the funding limit would apply "to Muslim and JeWish and Buddhist advocacy as well as to Christian" and, since the wording of the guidelines mentions not only the belief in a deity but also about it, the inference is that it would also apply to agnostics and atheists.50 But does this render the restriction content-based? A clear implication of this argument is that a "subject-matter" is constituted by a sum of "viewpoints". That is, if we put together all the viewpoints on a given issue, we produce an entire "subject-matter". A restriction which affects all viewpoints effectively becomes a content-based, but viewpoint-neutral, restriction. This may seem like a mechanical and simplistic picture. It may also be bard to implement. For a start, it calls for a test of a hypothetical nature which may be difficult to apply. Such a test would question whether the same restriction would apply to a publication which represented a different viewpoint. In addition, the definition of "subject-matter" as a sum of "viewpoints" relies upon the questionable distinction between a ''neutral'' discussion of religious matters and the advocacy of one or another view about these matters. Arguably, a publication which exhibited a neutral and descriptive approach (such as a journal concerned with the history of religious thought or a discussion of contemporary religions from a sociological standpoint) would not, for this reason, be disqualified from subsidization. One might speculate that such a neutral publication would not come under the definition of "religious activity" as an activity "which primarily promotes or manifests a particular belief in or about a deity or an ultimate reality". A sociological survey of contemporary religions or a historical discussion does not, it appears, primarily "promote or manifest" any particular belief in the deity, etc. But how is this "subject matter" different from the subject matter emerging from the adding together of several viewpoints? One possible response is that the subject matter properly targeted by the rule that denies official support for religious activities is "hortatory speech that 'primarily promotes or manifests' any view on the merits of religion" and that the denial of funding is "for the entire subject matter of religious apologetics". S1 But "hortatory speech" and ''religious apologetics" are loaded notions. Can an agnostic argument, which would be one of the ''viewpoints'' on Souter's list to be disqualified from funding, be properly characterized as 49. Id. at 2549. SO. Id. 51. Id.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

153

''religious apologetics"? This is more than a semantic problem: if the line between an impermissible subject matter (which consists of several "viewpoints", and which also includes atheistic and agnostic speech) and a permissible one (which would dispassionately describe religious matters) is that the latter involves no viewpoint at all, then it seems that the distinction is untenable. This is because it would seem that the "subject matter" in both these cases is exactly the same, and that the constitutional test is the presence or absence of a viewpoint. After all, how many viewpoint-free writings about religion have you seen lately? The difficulty in Souter's dissent is somewhat compounded by the way he distinguishes Rosenberger from Lamb's Chapel. Under challenge in Lamb's Chapel was a school board's prohibition of the after-hours use of school premises for religious purposes, even though the premises were otherwise open for a variety of social, civic and recreational purposes. Souter characterizes that restriction in the following way: "'Religious' was understood to refer to the viewpoint of a believer, and the regulation did not purport to deny access to any speaker wishing to express a non-religious or expressly antireligious point of view on any subject".52 Hence the ban can be presented as a case of viewpoint-based discrimination. But how does it differ from the exclusion of the religious publication in Rosenberger? This is the point of Kennedy's argument on the analogy between the two cases: Kennedy concludes that "here [in Rosenberger], as in Lamb's Chapel, viewpoint discrimination is the proper way to interpret the university's objections to Wide Awake".53 This is because, he claims, "the University does not exclude religion as a subject matter but selects for disfavored treatment those student journalistic efforts with religious editorial viewpoints".54 But isn't this just the same as in Lamb's Chapel where the school board "select[ed] for disfavored treatment" those groups which wanted to address various questions (in this case, about child rearing) from a religious viewpoint? In contrast to Kennedy's arguments, Souter claims that this analogy does not hold. It is important to ascertain which criteria he uses to distinguish the two regulations because these criteria seem to constitute a crucial constitutional distinction between impermissible viewpoint regulation and permissible content regulation. Therefore, the question is: why (in Souter's eyes) was "religion" a viewpoint in Lamb's Chapel and a subject matter in Rosenberger? One indication offered by Souter can be found in his suggestion that under the school board's regulation in Lamb's Chapel an "expressly antireligious point of view on any subject" would not have been denied access to the 52. Id. at 2550. 53. Id.at2517. 54. Id.

154

CHAPTERS

premises.55 Souter even cites an excerpt from the transcript of the oral argument in Lamb's Chapel as evidence that anyone ''with an atheistic or antireligious message would be permitted to use school property".56 This may be a strong indication of viewpoint discrimination behind the regulation, but it is not clear how easy it is to apply this test in practice. Imagine a particular religious group which wants to address an issue from a religious perspective. Should the validity of its exclusion depend upon whether an atheistic group would be permitted or excluded? What if no other such group ever applies? The use of a hypothetical test to judge an actual request is notoriously unreliable, and there are additional complications. If an atheistic group is excluded, how do we know that the exclusion is based on its subject matter rather than on its viewpoint, unless the topic of the atheistic group's meeting is exactly the same as that of the religious one? Because, of course, if the topic is somewhat different, its exclusion may still be content-based, but this would not be the same content as in the case of the religious group's exclusion. The suspicion may still linger that the latter exclusion was viewpoint-based. For example, using the facts of Rosenberger, imagine that Wide Awake discussed a number of social matters from a religious perspective and no atheistic newspaper ever discussed the same issues. How would we know that Wide Awake's exclusion was content-based? Under the "treatment of antireligious application" test, this is unverifiable. This is not the only, and perhaps not the most important, criterion for Souter in distinguishing the two cases. For him the main distinction seems to be that in Lamb's Chapel the access was denied to ''those who discuss issues in general from a religious viewpoint" while in Rosenberger the funding was denied "to those engaged in promoting or opposing religious conversion and religious observances as such".57 Souter clearly places great weight on this distinction. If the latter denial of funding did amount to viewpoint discrimination, he warns, it would mean that the Court would have "all but eviscerated the line between viewpoint and content".58 But this distinction between viewpoint and content seems too thin to carry such a great constitutional weight. It seems that what is really crucial for Souter is that, in Lamb's Chapel, a group tried to discuss a "primarily nonreligious" topic from a religious perspective while, in Rosenberger, a newspaper had religion as its subject matter. This must be the distinction between "discussing issues in general from a religious viewpoint" (Lamb's Chapel) and ''promoting or opposing religious conversion and religious observances" (Rosenberger). The regulation burdening the former is content55. 56. 57. 58.

Id at 2550. Id. at 2550 n. 13. Id. at 2550. Id.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

155

based while the regulation affecting the latter is viewpoint-based. The significance of this to Souter is clearly emphasized by his hypothesis that, if Kennedy's argument was right (that is, if the exclusion of funding for Wide Awake was viewpoint-based) then "a university that funds private publications about any primarily nonreligious topic [would also have] to fund publications primarily espousing adherence to or rejection of religion". 59 This consequence would be absurd, since it is not the case that for every secular discussion there is an opposing religious speech. For example, it would not be viewpointdiscriminatory for a university ''to fund a magazine about racism, and not to fund publications aimed at urging repentance before God".60 The problem is that such a clear-cut distinction can rarely be made, as is evidenced by the confrontation between Rosenberger and Lamb's Chapel. The two cases do not neatly correspond to a distinction between ''promoting religious conversion" and "discussing issues in general from a religious viewpoint". Indeed, it is difficult to imagine a promotion of religious conversion other than through a discussion of various issues (not ''primarily religious") from a religious perspective: the proselytizers know that the best way to gain adherents to their faith is to show people that with the religious perspective they will gain a better insight into the various issues which matter to them. As Souter's brief discussion of Wide Awake's content suggests, apart from articles on purely religious themes, there were also "essays on facially secular topics", such as racism or eating disorders, written from a Christian perspective. On the other hand, the group in Lamb's Chapel sought to discuss child-rearing from a Christian perspective. But how do we know that they were not trying to use this opportunity in order to promote religious conversion of non-Christians among the audience, or to strengthen their own religious observance by their adherence? Indeed, it would be strange if they did not. In fact, on the reading of Lamb's Chapel it seems clear that this was precisely one of the aims of the evangelical church called Lamb's Chapel, in the community of Center Moriches, which applied for permission to use school facilities to show a film series dealing with family and child-rearing issues from a Christian perspective. One of these films, as described in a brochure entitled "Overcoming a Painful Childhood", featured a Mrs Shirley Dobson who "recalls the influences which brought her to a loving God who saw her personal circumstances and heard her cries for help".61 It is very doubtful if the degree of ''religious apologetics", hortatory religious speech and straightforward proselytizing involved there was qualitatively lower than in Wide Awake's articles about racism or eating disorders viewed from a Christian perspective. 59. Id. at 2551. 60. Id. 61. 508 U.S. at 389 n. 3 (1993).

156

CHAPTERS

The upshot is that the distinction between viewpoint regulations and content regulations, drawn along the lines of the difference between a discussion of a "secular" topic from a religious perspective and speech aimed at promoting religious conversion, is rather unreliable. Souter's own example can hardly be squared with this criterion. But this is not fatal for the distinction. It only shows the unhelpful character of the specific criteria used by Souter. What does seem to be crucial is the reason for, and the effect of, a regulation, rather than the character of the speech itself. The primary motive for, and/or effect of, some regulations is to "skew" the debate. On the other hand, there are regulations which are mainly motivated by, and/or result in, the removal of some items from the agenda. These two types of regulation have distinctive and different prima facie moral defects, and the distinction between viewpoint and subject-matter regulations is shorthand for this moral distinction. At this point this is just an assertion, and it is the task of the next section to discuss in more detail the moral bases of hostility to these two types of regulation.

4

Rationales for Viewpoint Neutrality and Subject-Matter Neutrality

In a recent article, Nadine Strossen links "viewpoint-neutrality" and "contentneutrality" (two concepts which she blends into one) with the general idea that the government may not suppress any opinions merely on the basis that they are unpopular, offensive to the community, or repugnant to the majority of people. 62 But this general idea of not suppressing an opinion on the basis of its unpopularity gives effect to the principle of viewpoint-neutrality, not to the principle of content-neutrality. Opinions 'and viewpoints can be repugnant or offensive, but topics or subject matters cannot. Significantly, the general idea endorsed by Strossen (again, that the unpopUlarity of an idea does not provide sufficient grounds for suppression) is not anti-paternalistic. Consider the matter from the opposite angle: suppose someone advances the principle rebutted by Strossen, and believes that the majority's dislike of an opinion is sufficient to warrant a prohibition of that opinion. There is nothing paternalistic about such a principle. Paternalism is typically involved when someone's liberty is restricted on the basis of an argument about that person's own good. However, protecting the majority against offence restricts the speaker's liberty for the alleged benefit of the hearer - another person who, it is argued, wants to be protected against the unpleasant sensation of being offended. Paternalism, therefore, is not a typical motive behind viewpoint regulation (or, at least, the categories of viewpoint regulation which are triggered by the finding that an opinion is unpopular, offensive, or repugnant to the majority). In contrast, paternalism may be involved in content discrimination because it 62. Nadine Strossen, "Civil Liberties", George Mason University CiVil Rights Law Journal 4 (1994): 253-73, pp. 267-68.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

157

might be claimed (as a possible argument for a content restriction) that discussing certain matters is bad for the discussants themselves, regardless of the viewpoint they wish to express. This was the case in my hypothetical example, in Section 2 of this chapter, of a university's paternalistic policy which would deny support to student publications about sex. The general principle of not allowing the majority's repugnance to warrant prohibitions on speech is a principle activated by intolerance, not by paternalism. The example Strossen has given to illustrate this principle, namely, the cross-burning ban in R.A. v.,63 is certainly not a case of a paternalistic regulation. The St. Paul's ordinance in R.A. V. was not paternalistic because it did not restrict speakers' expressive liberties for their own good. If anything, it was a case of regulation based on an intolerance of racism. Whether such intolerance should properly be enforceable through punishment is another matter. For our purposes it is important to realize that the moral principle at work behind viewpoint neutrality is distinct from, and often viewed as more serious than, the moral principle at work in content neutrality. Strossen is wrong in lumping content neutrality and viewpoint neutrality together, when both the general principle and the example she supplies are relevant to viewpoint discrimination only. We now need to consider more carefully the proposition that the distinction between content-based restrictions and viewpoint-based restrictions corresponds to the distinction between paternalism and intolerance. The first claim is that the most plausible reason for legally restricting an agenda (content regulation) is the belief that the exclusion of the prohibited subject matter would be beneficial to the speakers themselves. But this is obviously not the only reason why the government would engage in agenda restriction. There are reasons which are less, and some which are more, invidious than this paternalistic justification. On the less invidious side, issue suppression may be justified by an attempt to facilitate the debate. Justice Stevens once noted: "As is true of many other aspects of liberty, some forms of orderly regulation actually promote freedom more than would a state of total anarchy".64 "Gag rules", a phenomenon analyzed by Stephen Holmes, are often a way of making the debate orderly, and indeed, of making it possible. Holmes concludes that "[c]ommunities, like individuals, can silence themselves about selected issues for what they see as their own good", and adds that "strategic self-censorship seems to be an almost universally employed technique of self-management and self-rule".65 Holmes 63. R.A.V. v. City of St. Paul, 505 U.S. 377 (1992). 64. Consolidated Edison Co. v. Public Service Comm'n, 447 U.S. 530, 546 (1980) (Stevens, J.), footnote omitted. 65. Stephen Holmes, "Gag Rules or the Politics of Omission", in Constitutionalism and Democracy, ed. by Jon Elster & Rune Siagstad (Cambridge: Cambridge University Press, 1988), p. 57.

IS8

CHAPTER 5

explored, at length, the various benign uses of "gag rules", the most obvious of which is to serve the purpose of facilitating public discourse. At any public meeting, the agenda must be restricted to some extent in order to get the discussion going in an orderly fashion. Of course, these restrictions may be excessive and irrational but there is no reason to believe that an ordinary reasonableness-based scrutiny is unable to separate unreasonable from reasonable agenda limitations. On the more invidious side, there may be agenda restrictions motivated by the government's hostility to a given topic. An obvious example is when a government does not want to have any political issues discussed in a particular forum, or by a particular group of speakers, because it fears that the views expressed will be unfavourable to the government. Prevention of a critique of those in power is certainly a more invidious motive than a paternalistic one. But such a motive is highly unlikely to be behind a subject-matter restriction: hostility to a topic is usually an irrational attitude. Indeed, it is hard to ascertain what "hostility to a topic" (as opposed to hostility to an idea, or to a viewpoint) might mean. The avoidance of an idea disliked by the government is much more easily achieved through a viewpoint-based restriction. To be sure, one might imagine a government which precludes the discussion of "politics" from a forum in order to preclude an expression of views critical of the government. But in a democratic country, such a move by the government would also preclude the expression of viewpoints criticizing the opposition and one would think that such viewpoints are something that a government would wish to encourage. Hence, a subject-matter exclusion of political debate would be, even from the egoistic point of view of the rulers, irrational. In any event, a subject-matter restriction motivated by, or having the effect of, suppressing only critical views, would constitute a case of indirect viewpoint discrimination. The second main claim is to show that viewpoint-based restrictions may be explained by reference to a government's hostility to the idea expressed. But this is perhaps not the only plausible explanation. Can viewpoint-based restrictions be based on a paternalistic motive? It is highly unlikely. An important exception to this proposition will be discussed below, but ftrst it is necessary to emphasize again, at the risk of becoming repetitive, that paternalism occurs when a government restricts the actions of those very people whose interests are alleged to be served by the restriction. Alternatively, there may be cases of "impure" or "two-party" paternalism, where the government restricts the actions of people other than those in whose interests the legislation is passed but who would act with the permission of the legislation'S beneftciaries.66 It is hard to see how viewpoint-based restrictions 66. For a distinction between these two forms of paternalism, see Joel Feinberg, The Moral Limits o/the Criminal Law, vol. 4: Harm to SeIf(New York: Oxford University Press, 1986),

VIEWPOINT NEUTRALITY AND ITS RATIONALES

159

can be based upon a genuine concern for the speakers rather than the hearers. If expressing a view is seen as harmful to the speaker (a condition of a paternalistic intervention), then it must be either: (1) because the rule-makers believe that acting upon a given view will be harmful to the speakers; or (2) because they believe that the view is so immoral that the very act of expressing it causes a self-inflicted moral harm. The former motivation seems irrational: we may well protect agents from self-inflicted harms by prohibiting an action (for example, riding a motorcycle without a helmet) but not by prohibiting an expression of support for an action. In tum, the second motivation is not genuinely paternalistic but moralistic. To say that we can prevent self-inflicted "moral harm" is only another way of saying that an action which is considered to be immoral can be prohibited. This is because the only sense that can be given to a notion of self-inflicted "moral harm" is that the person failed to attain what we consider to be a morally good character, or because he or she manifested his or her wickedness. But moral flaws do not "harm" a person in a way analogous to harm to one's interests, unless one has a prior wish to attain moral goodness. Unless this is the case, "moral harm" is just a manner of speaking about a morally reprehensible character. 67 Restrictions on expressions of a particular view, in order to prevent harm resulting from acting upon that view, may be rational in so far as we want to protect the hearer (who may be persuaded to take a harmful course), but not the speakers. It is obvious that if we prohibited the advocacy of riding without a helmet, we would have in mind the good of those who might be convinced to act on the basis of these arguments rather than the welfare of those who advocate the action. However, a restriction upon the speakers, motivated by our concern for the hearers, is not paternalistic unless we have a good reason to believe that the hearers are willing and consenting receivers of the messages against which we want to protect them (in which case it could be characterized as "impure" or "two-party" paternalism). However, this is rarely a reasonable belief. In order for a hearer to "consent to" receiving a message (much less, to request a message), first she must know what the content of the message is. Otherwise it is not an informed consent. But once she has acquired knowledge of the message, it is often too late for the authorities to protect her against it. The need to avoid imprecise use of the notion of "paternalism" with respect to speech restrictions is illustrated by Justice Brennan's dissenting opinion in Kokinda. Brennan considers, among other things, the Post Office's argument for a ban on the solicitation of money on the sidewalk outside a post office, and discusses the claim that such solicitation causes access difficulties for pp. 9, 172-176 and Gerald Dworkin, "Paternalism", in Paternalism ed. by Rolf Sartorius (University of Minnesota Press: Minneapolis, 1983), p. 22. 67. See Joel Feinberg, The Moral Limits of the Criminal Law, vol. 1: Harm to Others (New York: Oxford University Press, 1984), pp. 65-70.

160

CHAPTERS

customers of the post office, in particular, because it may annoy or embarrass some people. He considers that "[a]lthough the Service's paternalism may be well intended, it is axiomatic that a listener's reaction to speech is not a content-neutral basis for regulation".68 Regardless of whether the regulation is content-based or not, and regardless of whether it is well intended,. the intention attributed by Brennan to the Post Office is not paternalistic. Protecting the hearers against annoyance or embarrassment caused by the speakers cannot be characterized as paternalistic because the class of people whose liberty is being restricted is not the same as the class of people whose interests justify the restriction - and such an identity is an indicium of truly paternalistic intervention. The intervention might be characterized as indirectly paternalistic if the suggestion was that the hearers knowingly made themselves vulnerable to the risk of embarrassment by, for example, going to places about which they had been warned they may encounter something that would embarrass them. But certainly no such case can be made with regard to the patrons of post offices. There seems to be one significant exception to the proposition that viewpoint restrictions are not typically paternalistic. That is when the restriction is motivated by a desire to prevent harm to a speaker which would probably result from the audience's anger in reaction to the speech. Such a restriction seems to be genuinely motivated by paternalistic considerations. When a true motive for the suppression of a particular opinion is neither to prevent the harm to others resulting from acting upon the opinion, nor to protect the hearers against feelings of hurt, but rather to protect the speakers from the reaction by the audience, there occurs a correlation between the identity of the agent whose liberty is restricted and whose interests motivate the restriction. It is, of course, an established First Amendment principle that fear of disruption flowing from others' opposition to the speaker's viewpoint is not a good reason for restricting the speech. Thus, it is important to canvass reasons for the exclusion of this ground for suppression. In one of the classical statements explaining this exclusion, the American Supreme Court in Terminiello said: [A] function of free speech under our system of government is to invite dispute. It may indeed best serve its high purpose when it induces a condition of unrest, creates dissatisfaction with conditions as they are, or even stirs people to anger. Speech is often provocative and chaIlenging. It may strike at prejudices and preconceptions and have profound unsettling effects as it presses for acceptance of an idea.69

68. United States v. Kokinda, 497 U.S. 720,754 (1990) (Brennan, J., dissenting). 69. Terminiello v. Chicago, 337 U.S. 1,4 (1949).

VIEWPOINT NEUTRALITY AND ITS RATIONALES

161

This is a typical example of reasoning showing why the anger of hearers against speakers should not count as a reason for restriction. It has also been said that to allow such anger to trigger a prohibition would amount to a "heckler's veto", and that the burden in such situations should be upon the hearers to restrain themselves, to avert their eyes, etc., rather than upon the speakers to moderate their speech. A few examples will demonstrate the popularity of this precept in the United States' free speech doctrine. In Cohen v. California, a decision concerning offensive and vulgar words on a shirt worn by a man in the courthouse, the Supreme Court squarely placed the burden upon viewers to "avoid further bombardment of their sensibilities simply by averting their eyes".1° In a Nazi-parade decision, the Supreme Court of Illinois reminded the residents of a Jewish-populated suburb of Chicago that it is incumbent upon them "to avoid the offensive symbol [a swastika displayed by during a march by neo-Nazis] if they can do so without unreasonable inconvenience".7) Perhaps the most eloquent defence ofthe doctrine was given by the Supreme Court of New York in Rockwell v. Mo"is: [T]he unpopularity of views, their shocking quality, their obnoxiousness, and even their alarming impact is not enough. Otherwise, the preacher of any strange doctrine could be stopped; the anti-racist himself could be suppressed, if he undertakes to speak in 'restricted' areas; and one who asks that public schools be open indiscriminately to an ethnic groups could be lawfully suppressed, if only he choose to speak where persuasion is needed most. 72

And, more generally, the Supreme Court stated in Erznoznik v. City of Jacksonville that "the Constitution does not permit government to decide which types of otherwise protected speech are sufficiently offensive to require protection for the unwilling listener or viewer". 73 It is therefore clear that the force of the principle that hostile reaction against the speaker should not justify a restriction of the speech does not rest upon any fundamental dislike of paternalism, that is, it is not based on a principled objection to the idea of protecting the speakers against themselves. Rather, it rests upon two other rationales: (1) an unwillingness of the legislators to give effect to hearers' intolerance; and (2) a general belief that conflict and disagreement is a good thing in the realm of speech. The first rationale boils down to a general presumption against intolerance as a motive for legislation, and it further confirms that hostility to viewpoint restrictions is based upon hostility to intolerance. The word "intolerance" is used here in a broad sense which encompasses also intolerance for truly intolerable ideas: 70. Cohen v. California, 403 U.S. 15.21 (1971). 71. Vii/age of Skokie v. National Socialist Party ofAmerica, 373 N.E.2d 21 72. Rockwell v. Morris, 211 N.Y.S.2d 25.35-36 (1961). 73. Erznoznik v. City ofJacksonville, 422 U.S. 205, 210 (1975).

(1978).

162

CHAPTERS

ideas which should be condemned and rejected, such as fascism or racism. Still, the principle is that a dislike for the speakers and for their views, without more, is not sufficient reason to demand the use of the state's coercive powers against them. To allow such a reason for suppression would, in the words of the Supreme Court, "effectively empower a majority to silence dissidents simply as a matter of personal predilections".74 The matter is of course not so simple; as we will see in Chapter 6, if we replace the words "majority" with "traditionally denigrated minority" and "dissidents" with "bigots and racists". But, as a general principle that in the absence of a particular speech-related harm a "mere offensiveness" does not count as a legitimate ground for suppression, it is hard to impeach. Apart from everything else, the principle is supported by a Millian philosophy of the beneficial role of moral and intellectual conflict. As Jeremy Waldron argues, for Mill "ethical confrontation was indispensable for moral and social progress" and "moral progress depends ... on the struggle and confrontation between opposing views of the good life".7S This Millian argument, so clearly captured in the Terminiello passage quoted above, reflects a general conception of the benefits of diversity in the realm of expression. For our purposes, it is sufficient to establish that neither rationale has a significant paternalistic ingredient. By implication, if hostility to viewpoint restrictions is not based on anti-paternalistic sentiments, then this might suggest that viewpoint restrictions themselves are not usually motivated by paternalism.

5

The Content/Viewpoint Distinction and the Level of Generality

One reason for the difficulties in characterizing a regulation as content-neutral, content-based or viewpoint-based is that the conclusion largely depends on the level of generality of the question asked. Consider first a distinction between content-neutral and content-based restrictions. Whether a regulation is contentneutral (that is, in First Amendment parlance, whether it constitutes a "time, place or manner" regulation) or content-based, depends upon the generality of the "baseline" against which we judge the nature of the regulation in question. A good example is provided by the restriction challenged in Kokinda where members of a political advocacy group were not permitted to solicit contributions on a sidewalk near the entrance to a post office. They were prosecuted under a rule which prohibited solicitation on postal premises. The Court upheld the rule on the basis of the "non-public forum" doctrine. The majority found that the regulation did not discriminate because of content or viewpoint, and that it passed the test of reasonableness applicable in such 74. Cohen, 403 U.S. at 21. 75. See Jeremy Waldron, "Mill and the Value of Moral Distress", Political Studies 3S (1987): 410-23, pp. 416 and 420.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

163

circumstances. But Justice Kennedy, in his concurrence, found it unnecessary to resort to the "non-public forum" doctrine because he asserted that the restriction in question was merely a "time, place or manner" regulation which would normally warrant a reasonableness test. The only speech prohibited in this case was an "in-person solicitation for immediate payments on the premises",16 said Kennedy, and the restriction had nothing to do with the content of the speech. However, Justice Brennan disagreed with this characterization. The restriction, he claimed, was not content-neutral: [I]ndeed, it is tied explicitly to the content of speech. If a person on postal premises says to members of the public, 'Please support my political advocacy group,' he cannot be punished. If he says, 'Please contribute $ 10,' he is subject to criminal prosecution. His punishment depends entirely on what he says.77

The disagreement between Kennedy and Brennan concerns what it means to say that a regulation is "tied . . . to the content of speech". In one sense, Kennedy is right: regardless of the subject matter, all solicitation is equally prohibited. But "solicitation" itself is a term which applies to the content of speech (in a way in which, for example, loudness does not) and, in that sense, Brennan is right: the punishment depends on what the speaker says. The ambiguity arises as a result of different perceptions of the baseline for comparison. Brennan compares the restriction with a regulation in which the meaning of speech is irrelevant to the restriction; Kennedy compares it to a restriction in which the meaning is relevant, but the sphere of life to which the meaning applies is irrelevant to the restriction. Consider an analogy with blackmail. Is a prohibition on blackmail a subject-matter restriction? If the baseline is placed at the highest level of generality, at which we have to compare the regulation with a prohibition which applies regardless of the meaning of the words uttered, then obviously the prohibition of blackmail is content-based - we cannot tell whether the utterance constitutes blackmail unless we understand it. But if we place the baseline of comparison at a lower level of generality and compare the restriction with a regulation which only defmes the particular sphere of life to which it applies, then it is content-neutral because any blackmailing words (regardless of the sphere to which they apply) will be captured by the prohibition. Since a solution to the disagreement cannot be found in the very nature of the speech, a more fruitful method of deciding the meaning of "content neutrality" is by reference to the justification of the general rule against 76. Kokinda, 497 U.S. at 739 (Kennedy, J., concurring). 77. Id. at 753 (Brennan, J., dissenting).

164

CHAPTERS

content-based restrictions. One judicial formulation makes it explicit that it is the justification, rather than the nature of speech, which is crucial. In Ward v. Rock Against Racism78 the Court stated: "[E]ven in a public forum the government may impose reasonable restrictions on the time, place, or manner of protected speech, provided the restrictions 'are justified without reference to the content of the regulated speech ....79 Interestingly, Kennedy quotes this formula in his concurring opinion in Kokinda,80 but he does not. offer any justification for the restriction. He only asserts that the no-solicitation rule does not refer to the content of speech - but this is not what the Ward test demands. What it demands is to see whether the regulation is justified without reference to the content. What sort of justification would suit Kennedy's purposes? Arguably he would have to show that there is a convincing rationale for the rule - a rationale that does not refer to a distinction between the solicitation of money for political causes and for any other causes. In other words, any solicitation, no matter what its subject matter, would have to be included in the description of the rationale for the restriction. The arguments about "convenient and unimpeded access for postal patrons" or "facilitating [the] customers' postal transactions..81 seem to fit the prohibition reasonably well. Admittedly, they seem to be under-inclusive because other forms of speech may impede access for postal patrons as well, but it is undeniable that there is no reference to content in the rationale for a restriction. In tum, in Brennan's argument the emphasis is on the communicative impact of the expression upon customers as the rationale for the restrictions. As Brennan says, the fear on the part of the government is that "solicitation is bothersome because of its content: The post office is concerned that being asked for money may be embarrassing or annoying to some people, partiCUlarly when the speaker is a member of a disfavored or unpopular political advocacy group". 82 The alleged danger that "solicitors might annoy postal customers and discourage them from patronizing postal offices", 83 which Brennan discerns as the genuine rationale for the restriction, can only arise as a result of the content of the speech heard by postal patrons (or even, as Brennan's mention of "disfavored or unpopular" groups suggests, because of the viewpoint manifested by the speech). It is not necessary for our purposes here to weigh the merits of Kennedy's and Brennan's respective arguments about the grounds of customers' 78. 491 U.S. 781 (1989). 79. Id. at 791, quoting Clark v. Community for Creative Non-Violence, 468 U.S. 288, 293 (1984), emphasis added. 80. 497 U.S. at 738 (Kennedy, J., concurring). 81. Id. at 737 and 739. 82. Id. at 754 (Brennan, J., dissenting). 83. Id.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

165

annoyance in post offices around the United States. My point is that the best way of making sense of the "content-based" nature of the restriction is not by looking at the character of the speech restricted (because at this level, characterization depends crucially upon the level of generality), but rather at a rationale for the restriction, and whether this rationale refers to the content of the speech. Can a similar conclusion be drawn with respect to the distinction between content-based and viewpoint-based restrictions? Consider again the disagreement in Rosenberger between Souter and Kennedy about whether the denial of funding for a religious newspaper was viewpoint based. If the baseline for comparison is a restriction to which a viewpoint is not at all relevant, then Kennedy's claim that this restriction is not viewpoint neutral, is correct. This is for the obvious reason that, on any topic, views with a religious perspective may be disfavoured while views without such a perspective (but at the same time not activated by an anti-religious perspective) will not. However, as Souter suggests, compared to restrictions which favour or disfavour specific sectarian viewpoints within a general category of religious speech, this particular restriction in Rosenberger is viewpoint-neutral. This is because, in this case, no single religious viewpoint will receive less favourable treatment. The most fruitful way of distinguishing the two types of restrictions may be to appeal to a rationale which asks whether the restriction is justified with or without reference to a viewpoint. Under this interpretation, Kennedy's argument would have to be read as showing that the refusal of funding cannot be justified except by reference to the viewpoint of the speakers. But this is merely an assertion. What would a specific viewpoint-related rationale be? The rationale produced by the University was about avoiding conflict with the Establishment Clause - a Clause which does not disfavour any religious viewpoint. Even if, as Kennedy believes, the University has a mistaken understanding of what the Establishment Clause demands, he has not suggested that there is a viewpoint-based justification at work. In contrast, Souter presents a convincing case for the argument that the University's rationale for excluding funding for Wide Awake is viewpointneutral, claiming the denial of funding is not based on any hostility towards religion but only on what (the University believes) would be an impermissible violation of the neutrality of state towards religion, as mandated by the Establishment Clause. As Souter makes it clear, the rationale of the rule against viewpoint discrimination is to prevent the government from skewing public debate - the harm to be avoided is the situation "when government allows one message while prohibiting the messages of those who can reasonably be expected to respond".84 No such effect is likely to occur here because the 84. 115 S. Ct. at 2548-49.

166

CHAPTERS

rationale for the general distrust of viewpoint distinctions does not apply to these circumstances. A similar analysis which focuses on the question of the level of generality can be made with respect to a recent United States Court of Appeals for the Seventh Circuit decision, ss which largely relied upon Rosenberger. Under challenge in Grossbaum was a policy adopted by the City-County Building's proprietors in Indianapolis which proclaimed that no religious displays or symbols were to be permitted in the lobby of the building. The lobby was otherwise open to a number of other public and private speakers. As a result of this policy, a display of a menorah during the Jewish Chanukah holiday was not permitted. Rabbi Grossbaum and a Jewish Lubavitch group, who complained about the policy, characterized it as viewpoint-based. They argued that the "holiday season" was a subject matter for "discussion in the City-County Building lobby". This enabled Grossbaum to present an even-handed ban upon all religious displays as a viewpoint-based prohibition. But the building'S authorities claimed that the "holiday season" was not a ''meaningful'' subject matter: "By creating a policy which prohibits all religious displays and symbols, the Building Authority submits, it properly exerted its right to exercise control over the . . . lobby in a reasonable and viewpoint-neutral manner."S6 In effect, they claimed that it was ''religion'' which was the proper subject matter affected by the prohibition, and consequently that all the viewpoints comprising the subject matter of religion were equally affected. Again, it is a controversy which is unresolvable unless we settle upon a proper level of generality, which is a largely arbitrary decision. Grossbaum identifies as "subject matter", a theme upon which both religious and nonreligious ideologies may offer opinions. As a result of this enterprise, all religions, taken together, are represented as viewpoints. But the building's authorities draw their conceptual picture at a lower level of generality, so that "subject matter" is constituted by all religions, and ''viewpoint'' by anyone particular religion. Again, the best way of deciding between these two characterizations is by inquiring into whether the ban was likely to have been triggered by hostility towards religious messages or, rather, by a justified fear on the part of the decision-makers that permitting religious displays would run them into conflict with the Establishment Clause of the First Amendment. A number of factors may help to ascertain which of these two explanations is more plausible. Among them (though this is not an exhaustive list) are: Has the complaining group been subjected to a history of persecution and discrimination by a ruling majority in this community? Is the complaining group deprived of equally effective means of popularizing its viewpoint? Is it 85. Grossbaum v. Indianapolis-Marion County Building Authority, 63 F.3d 581 (7th Cir. 1995).

86. Id. at 588.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

167

highly improbable that the display of the symbol would be taken by some members of the public as an endorsement of the viewpoint by the authority? And so on. Positive answers to these questions support the hostility-towards-aviewpoint explanation while negative answers support an explanation that the denial of access was triggered by a viewpoint-neutral reluctance to get embroiled in Establishment Clause problems.

6

Indirect Viewpoint Discrimination

A regulation may be viewpoint discriminatory even if, on its face, it regulates an entire subject matter, and even if it picks up content-neutral characteristics of speech such as the status of speakers or the manner of communication. If a regulation is viewpoint discriminatory in its effect, then all the reasons which justify hostility to open viewpoint discrimination apply.

6.1 Subject-Matter Restrictions A prohibition upon a subject matter may in fact disguise a ban upon a viewpoint. This situation has to be distinguished from cases in which there is controversy as to whether a particular regulation is really based upon a subject matter or upon a viewpoint, as the Rosenberger case illustrates. There may be bans which are unquestionably addressed at a particular subject matter, although they in fact disadvantage only one point of view within the subject matter. In Mosley, that locus classicus for the argument in favour of content neutrality, the United States Supreme Court invalidated a city ordinance prohibiting all picketing near a school, except for peaceful labour picketing. The Court decided that the distinction between labour and nonlabour pickets (arguably, a clear subject matter distinction) did not further any of the goals cited as justifications for the ban. It stated: "If peaceful labor picketing is permitted, there is no justification for prohibiting all nonlabor picketing, both peaceful and nonpeaceful".87 Justice Marshall explicitly characterized the ordinance as describing impermissible picketing "in terms of subject matter,,88 and he made no use of the principle of viewpoint neutrality to condemn the ordinance. And yet, the facts of the case suggest that the ban might have had a clear viewpoint-discriminatory effect. The suit was brought by Earl Mosley, a postal employee who, for seven months prior to the enactment of the ordinance, had frequently picketed Jones High School in Chicago carrying a sign protesting against an allegedly racially discriminatory policy of the school. Under the ordinance, teachers at the school were allowed to express views about their 87. Police Department of Chicago v. Mosley, 408 U.S. 92,100 (1972). 88. Id. at 99.

168

CHAPTERS

conditions of employment (labour picketing), but others who wished to picket the school on issues not characterized as "labour disputes" could not. But who would want to picket the school, apart from currently employed teachers? Most probably, these other picketers would include parents, prospective teachers and other people concerned about aspects of the school's policy. These people, in contrast to the currently employed teachers, would be more likely than not to criticize school policies which favoured the existing teachers at the school. This was indeed the situation of Mosley himself. The exclusion of nonlabour picketing, therefore, disadvantaged viewpoints critical of the status quo which favoured the currently employed teachers.89 The lesson from this is that the distinction between content- and viewpointregulation is highly context-sensitive. It depends upon the specific facts which may suggest to us whether, due to the relative wlnerability of certain viewpoints, content regulation may have such an adverse effect upon some viewpoints that it will amount to viewpoint discrimination. But it does not follow that any content regulation can be presented as a viewpoint regulation, as the effective blurring of the line between two regulations by Justice Kennedy in Rosenberger would suggest. A Harvard Law Review comment praising Kennedy's reasoning and hailing Rosenberger for its broadening of the traditional concept of viewpoint discrimination, concludes that virtually all "bans on religious content" amount to viewpoint regulation.90 The weight of this conclusion rests upon the following analogy to a restriction based on the topic of sexual orientation: For example, a prohibition of literature discussing sexual orientation might be deemed content-based but not viewpoint-based because it forbids discussion of any viewpoint about sexual orientation. But because mainstream culture revolves round heterosexuality, heterosexuals tend not to discuss the subject of their sexual orientation at all; thus, only those offering a homosexual viewpoint have their speech suppressed. 91

The conclusion does not follow from its premises. It is a matter of context whether the motives and/or the effects of a ban on the discussion of sexual orientation are indeed such that "only those offering a homosexual viewpoint" will have their speech suppressed. Consider the case of a bigoted, homophobic milieu with a strong tradition of general acceptance of verbal attacks on gay men and lesbians (say, a small college in a very conservative area). A decision to suppress (or otherwise regulate) discussion of sexual orientation may have 89. See, similarly, Paul B. Stephan III, ''The First Amendment and Content Discrimination", Virginia Law Review 68 (1982): 203-51, pp. 233-34. 90. Note, ''The Supreme Court, 1994 Term: Leading Cases", Harvard Law Review 109 (1995): 111-335, p. 215 n. 43. 91. Id. at 214-15, footnote omitted.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

169

nothing to do with intolerance of homosexuality. To the contrary, it may be motivated by an intention to protect those holding an unpopular view on sexual orientation. While they will be denied the right to raise the issue in that particular limited forum, so too will their opponents. On balance, the minority may benefit from the restriction. The restriction would admittedly be somewhat paternalistic (which would confirm a suggestion made earlier, in· Section 4 of this chapter, that content regulations are typically paternalistic), but not intolerant. The argument that "because mainstream culture revolves round heterosexuality, heterosexuals tend not to discuss the subject of their sexual orientation at all" is a non sequitur. Precisely because "mainstream culture revolves round heterosexuality", heterosexuals tend to discuss the topic of sexual orientation, though not necessarily their own but that of sexual minorities. Hence, there is no reason to believe that a subject-matter restriction of speech based on sexual orientation will inevitably, or even typically, constitute a viewpoint restriction favouring a heterosexual perspective. Consequently, the analogy that the Harvard Law Review comment draws between the sexual orientation example and the Rosenberger situation does not prove the point that "the same could be said of bans on religious content".92 According to the author of the comment: "A ban that includes atheistic and religious viewpoints is virtually identical to a ban on religious viewpoints alone because, rather than pointedly questioning the existence of a Supreme Deity, agnostic and atheist writers in general ignore the question and focus on specific secular controversies.,,93 But agnostic and atheist authors often do directly discuss the question of religion. They discuss it from their own perspective, and such discussions would be suppressed under a subject-matter regulation based on religion. These suppressions may in principle work equally against religious viewpoints on one hand, and atheistic and agnostic viewpoints on the other. Indeed, in a community saturated with humanistic, rationalist or sceptical traditions, the primary effect of a ban on the discussion of religious issues will be to protect believers against attacks on their religious convictions. But when religion (or the existence of a Supreme Being, etc.) is not the subject matter of a discussion, it is not the case that the existence of a religion-based subject-matter regulation will disfavour those with religious perspectives because "agnostic and atheist writers in general· ignore the question [of the existence of a Supreme Deity] and focus on specific secular controversies".94 There is no such thing as "specific secular controversies" allegedly defined by the irrelevance of the issue of Deity to a discussion. There is nothing secular or non-secular in the discussion of a number of issues to which religion is irrelevant. To presume otherwise would be to adopt a "bipolar" view of public 92. Id. at 215 n. 43. 93. Id. 94. Id.

170

CHAPTER 5

discourse - a view criticized by Kennedy in Rosenberger whom, ironically, the Harvard Law Review note praises for his opinion in that case. The perception that, in Rosenberger, subject matter regulation is a disguised viewpoint regulation may partly stem from a belief about the relative powerlessness of the proponents of a viewpoint affected by a subject matter regulation. If one of the viewpoints adversely affected by a subject matter regulation is socially wlnerable and unpopular, then even if other viewpoints comprising the subject matter are also regulated, the adverse effect upon the unpopular viewpoint should raise our concern that the regulation may be motivated by intolerance towards that viewpoint. That is why the example about a ban on the discussion of sexual orientation may be plausibly represented in many contexts (though not in all, as suggested above) as viewpoint based. However, to apply the same moral concern to a religionbased ban, and to view a Rosenberger type of regulation as discrimination against a religious viewpoint, assumes that the religion affected by the ban is an unpopular and wlnerable perspective in this context. This, rather incredibly, has indeed been suggested by the Harvard Law Review note. The University of Virginia's regulation is given as an example ofa ban ''used to silence speakers with disfavored viewpoints".9S The note complained that the University was willing ''to provide funding to journals with religious perspectives, as long as that perspective was not Cbristian",96 and the Court's decision is hailed as restricting the University officials' discretion "to decide which perspectives are sufficiently politically popular to warrant funding".97 If the diagnosis is that religion in general, and Christian religion in particular, is disfavoured and politically unpopular in the United States, or in some specific forum such as the University of Virginia, then a subject-matter regulation limiting the discussion of religion may raise concerns analogous to those which are triggered by viewpoint discrimination (namely, that it is yet another exercise in intolerance). But it would take some explaining to convince one about the accuracy of such a characterization of the position of Christian religion in American public life today.

6.2 Speakers' Status Restrictions

In Perry Education Association v. Perry Local Educators' Association, the Supreme Court of the United States reviewed a restriction on access to the interschool mail system and teacher mailboxes in Perry township schools. Under a collective-bargaining agreement between the Board of Education and Perry Education Association (PEA), PEA (as the exclusive bargaining 95. Id. at 216. 96. Id" footnote omitted. 97. Id. at 217.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

171

representative for the teachers) was granted access to the internal mail system while access was denied to any rival union. Justice White, who delivered the opinion of the Court, claimed that the internal mail system in public schools was not a "public forum" and, therefore, it could be reserved "for its intended purposes, communicative or otherwise, as long as the regulation on speech [was] reasonable and not an effort to suppress expression merely because public officials oppose[d] the speaker's view".98 As to the first criterion, White accepted the School District's arguments about the reasonableness of the exclusion of any teacher's union other than PEA, such as a submission about the special responsibilities of PEA and ensuring labour peace within the schools. As to the requirement of viewpoint neutrality, White claimed that it was "more accurate to characterize the access policy as based on the status of the respective unions rather than their views".99 This is, however, often a distinction without a difference. If, as often happens, particular viewpoints can be reliably attributed to the "status" of a particular soCial actor, then a status-based exclusion is equivalent to viewpoint discrimination. According to Justice Brennan, who wrote a dissenting opinion, this was the case in Perry. Brennan quotes with approval the Court of Appeals' decision in the Perry case, based, inter alia, on a judgment that the exclusion under challenge "favors a particular viewpoint on labor relations in the Perry schools" because "the teachers inevitably will receive from [PEA] selflaudatory descriptions of its activities on their behalf and will be denied the critical perspective offered by [an alternative union]".IOO Further, Brennan states persuasively that "access is denied to the respondents because of the likelihood of their expressing points of view different from the petitioner's on a range of subjects". 101 But in many cases the characteristics of the speaker do not have to be so strongly correlated with the viewpoint as to raise the fear of indirect viewpoint discrimination. It is therefore a mistake to simply equate regulations based on a speaker's identity with viewpoint regulations. Justice Stevens is guilty of such an equation in his concurrence in Consolidated Edison when he says: A regulation of speech that is motivated by nothing more than a desire to curtail expression of a particular point of view on controversial issues of general interest is the purest example of a 'law. " abridging the freedom of speech, or of the press.' A regulation that denies one

98. 460 U.S. 37,46 (1983). 99. Id. at 49. 100. Id. at 65 (Brennan, J., dissenting). 101. Id.

172

CHAPTERS group of persons the right to address a selected audience on 'controversial issues of public policy' is plainly such a resulation.l02

Furthermore, even if there is such a likelihood of discrimination, the reasons for restricting access to some speakers because of their status may be compelling. Think of a group of non-faculty members claiming the right to speak at a faculty meeting of a public university. Clearly, in a ''non-public forum", the test for a justification of exclusion needs to be less rigorous than in public fora such as streets and parks. This conclusion is reflected in the lenient requirement of reasonableness which is applied in non-pUblic fora under the First Amendment forum doctrine. Nevertheless, one should be aware of the risk of viewpoint discrimination resulting from a prima facie neutral, statusbased distinction, as is illustrated by Perry.

6.3 Means-of-Communication Restrictions Regulations which close off an entire medium of communication mayor may not have a disproportionate impact upon a particular viewpoint. A law prohibiting all leafleting would have a more severe impact upon those who do not have easy access to more effective means of mass communication, and often these will be the critical viewpoints of the current social status quo. A ban on anonymous leaflets, such as the one struck down by the United States Supreme Court in 1995,103 would understandably disadvantage those who might have a reason to fear readers' reactions to the content, and hence adversely affect those who may reasonably expect their views to be controversial and unpopular. As Justice Stevens mentioned obiter, "the disclosure requirement places a more significant burden on advocates of unpopular causes than on defenders of the status quo".I04 It should be added that in this particular case Stevens, who delivered the opinion of the Court, did not base his decision to invalidate the disclosure requirement upon the argument of viewpoint discrimination. Rather, the invalidation resulted from his strict scrutiny of the regulation of the speech covered by the disclosure requirement, namely, speech on public issues. However, the suggestion of indirect viewpoint discrimination explicitly figured in the footnoted remark just cited, and also implicitly in the concluding remark that "[a]nonymity is a shield from the tyranny of the majority", and, therefore, it "exemplifies the purpose behind the Bill of Rights, and of the First Amendment in particular: to

102. Consolidated Edison Co. v. Public Service Comm 'n, 447 U.S. 530, 546 (l9S0) (Stevens, J., concurring) (quoting U.S. Const., amend. I), footnote omitted, emphases added. 103. See McIntyre v. Ohio Elections CommisSion, 115 S. Ct 1511 (1995). 104. Id. at ISIS n.S.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

173

protect unpopular individuals from retaliation - and their ideas from suppression - at the hand of an intolerant society". 105 It is uncertain whether such a correlation between a restriction and the type of a viewpoint could have been established in Lehman v. City of Shaker Heights}06 In that decision, the Supreme Court of the United States upheld a city ban on political ads in advertising spaces on public transit vehicles. It would be hard to say that such a ban affects one particular political group or ideology more than others, as the advertising spaces could have been used by aU sorts of parties and groups. They are, one would imagine, neither so expensive as to be available only to the wealthiest groups (to which one might attribute a by·and·large conservative attitude), nor so very cheap as to be attractive to very poor groups (which, one might speculate, tend to represent more radical viewpoints). One may perhaps detect, as did Brennan in his dissent, the dangers of viewpoint discrimination inherent in the very distinction between political advertising and commercial advertising. Brennan offers the following hypothetical: "a commercial advertisement peddling snowmobiles would be accepted, while a counter·advertisement calling upon the public to support legislation controlling the environmental destruction and noise pollution caused by snowmobiles would be rejected".107 But there is no reason to believe that political advertisements would systematically take an anti·consumerist line. While nearly all commercial ads implicitly rely upon a "viewpoint" that the consumption of material goods is worthwhile, and in this sense rely upon a non-critical attitude towards free markets and capitalism, there is no reason to believe that the category of expression excluded under the ordinance (that is, political advertising) would generally take an opposing viewpoint. For this reason, the charge of indirect viewpoint·discrimination is rather doubtful in this particular case.

7

Paternalism and Intolerance

In his dissenting opinion in Perry, Brennan insists that the government must not discriminate among viewpoints on a particular topic, even if it "may entirely exclude discussion of the subject from the forum" .108 This raises an apparent paradox: "the greater power does not include the lesser". The government may prohibit more but not less speech. This is, Brennan says, "because for First Amendment purposes exercise of the lesser power is more threatening to core values".109 But why is viewpoint discrimination more 105. Id. at 1524. 106.418 U.S. 298 (1974). 107. Id. at 317 (Brennan, J., dissenting).

108. Perry Education Assn. v. Perry Local Educators' Assn., 460 U.S. 37, 62 (1983) (Brennan, J., dissenting). 109. Id. at 62.

174

CHAPTERS

threatening to the core values of freedom of speech than the exclusion of a subject from a forum? If the argument in Section 4 of this chapter is correct, this question translates into another question: why is paternalism less dangerous than intolerance in the area of restrictions on freedom of expression? Put at this level of generality, the question does not allow for any satisfactory answer. Obviously, there may be some paternalistic restrictions which are intolerable and, on the other hand, there may be restrictions which express an intolerance so mild, or so justifiable, as to be acceptable. As a general and abstract proposition, it may therefore be the case that subject-matter distinctions are just as objectionable as viewpoint distinctions. As Frederick Schauer argues: "If the first amendment is in fact designed in theory to protect discussion over a wide area, a subject matter restriction is no more justifiable than a viewpoint restriction".ttO According to Schauer, similar theoretical underpinnings of the free speech principle to those which support hostility to viewpoint distinctions also argue against subject-matter regulations. ttt Likewise, the Supreme Court once announced: "The First Amendment's hostility to content-based regulation extends not only to restrictions on particular viewpoints, but also to prohibition of public discussion of an entire topic".tt2 This proposition about equal hostility towards subject-matter and viewpoint-distinctions is given the following rationale in Consolidated Edison, the decision from which the last sentence was quoted: "If the marketplace of ideas is to remain free and open, governments must not be allowed to choose 'which issues are worth discussing or debating .... ' To allow the government the choice of permissible subjects for public debate would be to allow that government control over the search for political truth".tt3 This would sound persuasive if the proposed restrictions were to affect "public debate" in its entirety; however, the issue is usually about restrictions in particular fora, contexts or circumstances. The question then is whether, everything else being equal ("everything else" including the forum, the speaker, etc), a subjectmatter restriction is equally objectionable as a viewpoint restriction. It is 110. Frederick Schauer, "Categories and the First Amendment: A Play in Three Acts", Vanderbilt Law Review 34 (1981): 265-307, p. 285. III. One should note that Schauer precedes this conclusion by a proposition: "However loose the boundaries may be, distinctions drawn on the basis of viewpoint are importantly different from distinctions drawn on the basis of subject matter", id. at 284, footnote omitted. However, in the two paragraphs that follow he seems to strongly repudiate the proposition just quoted, and argues for equivalence in the seriousness of these two types of regulations, on the grounds that subject matter regUlations are rarely viewpoint neutral, that they freeze the existing knowledge about a certain subject by suppressing a discussion of the subject, and on the basis of lack of "theoretical foundations" for the distinction between viewpoint- and subject-matter regulations under the first amendment. 112. Consolidated Edison, 447 U.S. at 537. 113. Id. at 537-38, quoting Mos/ey,408 U.S. at 96, references omitted.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

175

characteristic that, immediately after having produced the "marketplace of ideas" rationale for hostility towards subject-matter restrictions, Justice Powell discusses, in Consolidated Edison, two earlier decisions by the Supreme Court in which subject-matter regulations were upheld. Powell is at pains to emphasize that Greer l14 and Lehmani IS are to be viewed as narrow exceptions to the general prohibition against subject-matter distinctions, but the rationale given to these "exceptions" is very telling. In both these cases, as Powell correctly states, the Court "concluded that partisan political speech would disrupt the operation of governmental facilities even though other forms of speech posed no such danger".1l6 But such a conclusion can be, mutatis mutandis, plausibly made in a number of different contexts, and the availability of an unrestricted agenda for public speech may bring havoc to the operation of any forum or institution. However, the ban on political speeches in a military base (upheld in Greer) is less objectionable than a ban on left-wing political speeches in the same military base, or a ban on any political speeches other than those that approve of the current government's military strategy. This latter ban would probably carry very little risk of "disrupt[ing] the legitimate governmental purpose for which this facility was dedicated".1l7 Indeed, permission for political speeches on the condition that they praise the politics of the government would probably be conducive to "the purposes" for which a military base "was dedicated". And yet, such a ban would more obviously clash with the values underlying freedom of speech than would a prohibition on all political speech. But "disruption to the purposes of a facility" does not need to collapse into the managerial or governmental purposes of the institution; it may consist of a disruption to the activities of those who participate in a given activity, such as a public meeting, university, or workplace. This smacks of paternalism, and one can see why subject-matter restrictions can be considered to be as potentially dangerous as viewpoint restrictions: paternalism, in general terms, is an attitude essentially foreign to the core values of freedom of speech. These central values have been acknowledged in First Amendment jurisprudence as, among other things, the principle of individual autonomy and self-mastery ("people will perceive their own best interests if only they are well enough informed"), 118 and as the principle of democratic self-government ("the people in our democracy are entrusted with the responsibility for judging and evaluating the relative merits of conflicting arguments,,).119 These statements reject the possibility of restraining people's speech for their own good; subject 114. Greer v. Spack, 424 U.S. 828 (1976). 115. Lehman v. Shaker Heights, 418 U.S. 298 (1974). 116.447 U.S. at 539. 117. Id. at 538.

118. Virginia Pharmacy Board v. Virginia Consumer Council, 425 U.S. 748, 770 (1976). 119. First National Bank a/Boston v. Bel/otti, 435 U.S. 765, 791 (1978), footnote omitted.

176

CHAPTER 5

to a proviso that they are well informed (and minimizing speech control is a prerequisite of meeting this requirement), people decide their own interests on the basis of independent judgment and the evaluation of beliefs and views put to them. But there are some cases where paternalism is triggered not by the legislators' distrust in individuals and the perception of a "danger that· the people cannot evaluate ... information", 120 but by the opinion that, for some reasons other than their al/eged incapacity to evaluate information, participants in a given forum will be better off if a certain subject matter is excluded from the debate. Consider Rosenberger again. Whatever the reasons for the public university's decision to deny financial support to a religious newspaper, distrust of the audience based on suspicions that students would not be able to evaluate the information, did not figure among them. If, arguendo, the University was right in interpreting the Establishment Clause as not permitting the direct funding of a religious newspaper, and if such a denial amounted to a regulation of subject matter rather than of a viewpoint (as it was argued in Section 2 of this chapter), then the rationale behind that regulation is different from the rationale which would make a paternalistic restriction morally offensive. A truly reprehensible paternalistic rationale would be one related to a decision-maker's suspicion that the audience is incapable of evaluating the information and arguments. The university's decision to deny funding, whether right or wrong under the Establishment Clause, did not stem from that paternalistic motive. Rather, the denial might be justified by the fear of divisiveness stemming from religious co~troversies in a given forum. In his essay on "Gag Rules", Stephen Holmes states: "No issue is more frequently classified as 'worthy of avoiding' than religion. Sectarianism, is understandably considered to be divisive, a serious threat to communal cooperation.,,121 This threat of divisiveness has been explicitly cited by the Court as a major evil targeted by the Establishment Clause: "[P]olitical division along religious lines was one of principal evils against which the First Amendment was intended to protect. . . . The potential divisiveness of such conflict is a threat to the normal political process". 122 Strictly speaking, a speech restriction motivated by a wish to avoid divisiveness may also be seen as paternalistic because it limits individuals' expressive freedom in order to advance their own good. So if all that matters in order to characterize an act as "paternalistic" is that someone is coerced for her or his own good, then subject-matter restrictions, as discussed above, are paternalistic. But it is not the offensive sort of paternalism which presumes that 120. Id. at 792. 121. Holmes, p. 23. 122. Lemon v. Kurtzman, 403 U.S. 602, 622 (1971), references omitted.

VIEWPOINT NEUTRALITY AND ITS RATIONALES

177

people are incapable of properly using their freedom, including their free access to information and to different arguments. An act which imposes restraints upon a person's liberty for his or her own good, but which can be recognized by that person as necessary in order to achieve the goal shared by that person, is not "paternalistic" in a stronger, more objectionable sense. Rather, it is the sort of paternalism which imposes constraints upon individual action in order to produce a collective good or to avoid a collective harm encapsulated, in this case, by the concept of "divisiveness". Or, to put it somewhat differently, it consists of the foreclosure of certain individual actions which, if taken by many, would make everybody or nearly everybody worseoff, and which people have an incentive to take unless assured that others will also abstain. This variety of "paternalism", resulting as it does from an attempt to solve the Prisoner's Dilemma just described, has been recognized as not exactly fitting the usual premonitions against forcing people for their own good. 123 In a genuine Prisoner's Dilemma, foreclosure of one avenue of action is a condition for the attainment of the optimal solution for every participant; in the absence of such a foreclosure (in our example, a restriction on a subject matter) a suboptimal result will be attained by everybody. It is therefore very likely that most, if not all, people, whose freedom of action is restricted by a regulation, would recognize that they are all better off thanks to universal compliance with the rule. So while the constraint is paternalistic in one sense (because its effect is to restrict the choices open to individuals for their own good), its motivation is nevertheless not paternalistic. That is, it does not depend on the argument that those people who are controlled by the rule cannot recognize what is really good for them. Instead, it gives effect to their actual choices which can be implemented only where there is mutual compliance enforced by the existence of a legal constraint. One major argument of this chapter has been that this feature explains why subject-matter regulations may be seen as morally less objectionable than viewpoint regulations. The general argument can only be upheld if three more specific arguments can be sustained: (1) that a meaningful distinction can be drawn between viewpoint neutrality and subject-matter neutrality in speech regulations; (2) that subject matter restrictions are often based upon paternalism while viewpoint restrictions are usually triggered by intolerance; and (3) that paternalism is less problematic as a legislative motivation than intolerance. None of these three arguments allows for a confident, unambiguous and exceptionless defence. But with all the qualifications, reservations and caveats put forward in the earlier parts of this chapter, the arguments can hopefully be sustained. If so, they illuminate the moral force of 123. See Gerald Dworkin, p. 23; Cass R. Sunstein, "Legal Interference with Private Preferences", University of Chicago Law Review 53 (1986): 1129·74, pp. 1140-45.

178

CHAPTERS

the principle of viewpoint neutrality as a major regulative principle of freedom of speech.

CHAPTER 6 RACIAL VILIFICATION AND FREEDOM OF SPEECH 1

The Context of the Hate-Speech Controversy

The legal systems of the United Kingdom, Canada, Australia, New Zealand and most other modern states prohibit the dissemination of ideas hostile to racial groups. For example, the 1965 Race Relations Act made "incitement to racial hatred" an offence in the United Kingdom in circumstances where the accused intended to incite racial hatred and the language used was threatening, abusive or insulting and was actually likely to stir up racial hatred. l In a number of other European countries, legal systems go further and criminalize the expression of views which "merely" vilify or insult racial groups. For example, in the Netherlands it is a criminal offence to "deliberately give public expression to views insulting to a group of persons on account of their race, religion or conviction or sexual preference";2 in Sweden, the Freedom of the Press Act, one of Sweden's constitutional documents, prohibits the expression of contempt for a population group "with allusion to its race, skin colour, national or ethnic origin, or religious faith".3 In Australia, at the federal level, the Racial Hatred Act of 1995 prohibits public behaviour that is likely "to offend, insult, humiliate or intimidate another person or a group of people", if the act is done "because of the race, colour or national or ethnic origin" of the other person or a group;4 in addition, a number of Australian states prohibit acts of racial vilification. Also major international human rights treaties authorize, and sometimes even require, governments to punish the advocacy of hatred on national, racial or religious grounds. The requirement of a legal prohibition of hate speech is included in the International Covenant on Civil and Political Rights and the American Convention on Human Rights, while the European Convention on Human Rights and the African Charter on Human and People's Rights permit, although I.

2. 3. 4.

Section 6 of the Race Relations Act. In 1976 a revised offence was inserted as Section SA into the Public Order Act 1936 of the new Race Relations Act. In 1986 a new Public Order Act created two separate crimes of incitement to racial hatred, one by using written material and the other by using words or behaviour; also, a new offence of possession of racially inflammatory material was introduced. See 1. Oyediran, "The United Kingdom's Compliance with Article 4 of the International Convention on the Elimination of All Forms of Racial Discrimination", in Striking a Balance: Hate Speech. freedom of Expression and Nondiscrimination, ed. by Sandra Co liver (London: Article 19, 1992), pp. 246-47. Section 137 of the Criminal Code. See I. Boereffijn, "Incitement to National, Racial and Religious Hatred: Legislation and Practice in the Netherlands", in Striking a Balance, p. 202. Freedom of Press Act, Chap. 7, Art. 4, para II. Section 3 of Racial Hatred Act 1995 (Cth) which constitutes an amendment to the Racial Discrimination Act 1975.

180

CHAPTER 6

they do not expressly require, governments to ban hate speech. Perhaps the strongest formula is in the International Convention on the Elimination of All Forms of Racial Discrimination which requires, in Article 4, state parties to declare as criminal offences, "all dissemination of ideas based on racial superiority or hatred, incitement to racial discrimination" and participation in "organizations, and also organized and all other propaganda activities, which promote and incite racial discrimination". The United States is an exception with its doctrine that speech may not be prohibited, regardless of its offensiveness, unless there is a clear and present danger that it will incite imminent unlawful action. The last time a statute outlawing defamation of racial and religious groups was upheld was in 1952, in a decision taken by the United States Supreme Court by a margin of five to four. s The case concerned the prosecution of a white supremacist under a state law which prohibited any publication exposing citizens to "contempt, derision, or obloquy" on the basis of their race, colour, creed or religion. While the decision has never been formally overruled, it is generally considered not to be valid any more. The decision has not been relied upon in any subsequent federal decision; it was described in a Supreme Court decision (though by a dissenting judge) as one that "should be overruled as a misfit in our constitutional system and as out of line with the dictates of the First Amendment",6 and has been understood by the lower courts to have a questionable vitality. Back in 1968, Judge Skelly Wright observed that "far from spawning progeny, Beauharnais has been left more and more barren by subsequent First Amendment decisions, to the point where it is now doubtful that the decision still represents the views of the Court".7 More recently, another appellate judge, Frank Easterbrook, noted that recent Supreme Court cases "had so washed away the foundations of Beauharnais that it could not be considered authoritative".8 Accordingly, all attempts to prohibit hate speech (other than "fighting words") have been consistently held unconstitutional. A federal court invalidated a university's limitations on hate speech on the basis that what the University could not do . . . was establish an antidiscrimination policy which had the effect of prohibiting certain speech because it disagreed with ideas or messages sought to be conveyed. . . . Nor could the University proscribe speech simply

S. 6. 7. 8.

Beauhamais \I. Rlinois, 343 U.S. 250 (1952). Ga"ison \I. Louisiana, 379 U.S. 64, 82 (1964) (Douglas J, dissenting). Anti-Defamation LeagueofB'Nai B'Rith \I. FCC, 403 F.2d. 169, 174 n.5 (D.C. Cir. 1968). American Booksellers Ass 'n \I. Hudnut, 771 F.2d 323, 331 n. 3 (7th Cir.1985). One of the very few commentators who claim that Beauharnais has not been effectively overruled and is still good law, is Kenneth Lasson, see his "Group Libel Versus Free Speech: When Big Brother Should Butt In", Duquesne Law Review 23 (1984): 77-130, pp. 111-14.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

181

because it was found to be offensive, even gravely so, by large numbers ofpeople.9

And a few years later, a district court struck down another university's "discriminatory harassment policy" on similar grounds, and caustically added that it was "not willing to entrust the guardianship of the First Amendment to the tender mercies of [the university's] discriminatory harassment/affirmative action enforcer". I 0 However, many legal scholars have repeatedly demanded that this noninterventionist approach be abandoned, for example by making the delivery of racial insults an actionable tort, similar to the tort of defamation or of the intentional infliction of emotional distress. 11 In what has become probably the most celebrated of these proposals, Marl Matsuda suggested prosecuting racist speech which has the following characteristics: a message of racial inferiority, which is directed against a historically oppressed group, and is prosecutorial, hateful or degrading. 12 In addition, as a consequence of the plague of racial harassment and abuse at campuses, support has been expressed by many legal scholars for university codes allowing for the punishment of hate speech in universities. Legal control of "hate speech" poses a moral and philosophical dilemma of a particular gravity to anyone committed to the broadly understood principles which govern the liberal and democratic legal order. This dilemma arises out of the need to balance and reconcile conflicting principles and concerns, each of which is valuable and important, and yet each of which must be subject to some preferential ordering in the specific setting of legal responses to "hate speech". There are various ways of characterizing this conflict of values: freedom (which includes freedom of speech) versus equality (which implies concern about the possible exclusion of victimized groups from mainstream public life as a result of stereotypes and prejudices promoted by hate speech); formal equality (which calls for an equal right to speak on matters of public concern regardless of the worth of one's views) versus substantive equality (which implies concern about the dignity of those hurt by offensive speech); deference to majority opinions (when the majority is unperturbed by the offensive speech of some extremists) versus respect for minority interests (which demands protection of the sensibilities of the most vulnerable groups), etc. 9. Doe v. University a/Michigan, 721 F. Supp. 852, 863 (E.D. Mich. 1989). 10. Dambrot v. Central Michigan University, 839 F. Supp. 477, 482 n. 7 (E.D.Mich. 1993). 11. See Richard Delgado, "Words that Wound: A Tort Action for Racial Insults, Epithets, and Name-Calling", Harvard Civil Rights- Civil Liberties Law Review 17 (1982): 133-81; Donald A. Downs, "Skokie Revisited: Hate Group Speech and the First Amendment", Notre Dame Law Review 60 (1985): 629-85; Lasson, passim. 12. Mari J. Matsuda, "Public Response to Racist Speech: Considering the Victim's Story", Michigan Law Review 87 (1989): 2320-81, p. 2357.

182

CHAPTER 6

Regardless of the characterization of the conflict of values in question (and I will leave to one side the observation that some of these characterizations may be more appropriate than others), it is clear that the dilemma depicted is real and often dramatic~ The gravity of the problem is amplified by the fact that racially motivated stereotypes, insults and assaults seem to be on the rise in a number of contemporary liberal democracies. It is no wonder that there is a widespread feeling of frustration and perplexity concerning the approach of mainstream liberal theory to the problem, and also a certain angst amongst liberal theorists themselves who feel that it constitutes perhaps the most direct challenge to their established views about the limits of punitive state action. The orthodox civillibertarian view that "mere offensiveness", or even a shock to someone's feelings, is not a sufficient ground to restrict the offensive action, does not seem to capture the reality of the harm produced by hate speech, and hence distorts the nature of moral dilemmas. There is unfortunately a ring of truth to the statement by Toni Massaro that it is precisely with respect to hate speech that liberalism's "vocabulary is least compelling, most paradoxical, and least responsive to real world conditions and actual human experience".13 Two preliminary observations are in order. One is that, with respect to hate speech, the need to equate "speech" with "symbolic conduct", as discussed earlier in Chapter 2, is particularly obvious. The distinction is of no help to those who would argue that hate speech calls for stringent protection whilst racist symbolic conduct should be regulated. Repression of, or impunity for, racist speech and racist communicative conduct must stand or fall together. This is confirmed by the fact that a number of laws in various jurisdictions explicitly include expressive conduct in their anti-hate-speech provisions. For example, the New South Wales Racial Vilification Act of 1989 equally prohibits "any form of communication" and any other "conduct" including "actions and gestures and the wearing or display of clothing, signs, flags, emblems and insignia" that promote or express racial vilification. 14 It is understood that, for example, wearing badges with racist slogans or making racist gestures (such as a "Heil Hitler" salute) would be illegal under the Act. Similarly, Canadian criminal law which penalizes statements that incite hatred against any identifiable group defines "statements" as "words spoken or written . . . and gestures, signs or other visible representations".15 Anti-hate law in British Columbia expressly prohibits "any conduct or communication" which promotes hatred or contempt, etc. 16

13. Toni M. Massaro, "Equality and Freedom of Expression: The Hate Speech Dilemma", William & Mary Law Review 32 (1991): 211-65, p. 230. 14. Art. 20 B of Anti-Discrimination (Racial Vilification) Amendment Act NSW 1989. IS. Canadian Criminal Code (1970) art. 281.1 (7). 16. British Columbia Civil Rights Protection Act (/981) art. 1 (I).

RACIAL VILIFICATION AND FREEDOM OF SPEECH

183

The second observation is that, when discussing the appropriateness of hate speech bans, one should distinguish between general bans which apply to all public communications in society at large, and those which apply in special settings only, when the particular context lends itself to either a more permissive or more restrictive approach than in general discourse. While this chapter applies to legal restrictions which have a universal applicability, it may be useful to note the variety of special settings and special contexts to which the general laws (such as criminal codes or defamation laws) need not apply, or may apply with all the more stringency. These special settings include, in particular, the workplace. For instance, in the United States racist speech in the workplace is indirectly regulated under the hostile work environment provisions of Title VII of the Civil Rights Act of 1964. This is analogous to a judgment, delivered by the Supreme Court, that sexual harassment including verbal harassment in the workplace creates a hostile or abusive work environment and constitutes a violation of the prohibition against sex discrimination in employment. 17 In a more specific context, a union election conducted in the workplace can be set aside if the employer used racially inflammatory campaign materials as it would violate the fair election provisions ofthe National Labor Relations Act. There are at least three reasons why the workplace can be seen as a setting which lends itself to a greater protection against racial vilification than society at large. First, the workplace is an environment in which the hierarchy and subordination are typically more intense than in a societal interaction which provides a context for public discourse. This hierarchy and subordination creates the potential for a particular vulnerability of victims of workplace verbal harassment: in society at large they are free to reply, even sharply, but such a reply in the workplace may result in harsh sanctions, such as the loss of a job (assuming that the supervisor is the source of verbal harassment, or tolerates the abuse by co-workers). Second, people in the workplace constitute the "captive audience": they have very few methods of escaping the message, and the most effective methods may turn out to be excessively costly. Again, this is because, if the employee physically leaves the workplace to avoid further insults, it may result in the loss of the job. Work is a compelled activity for the majority of people, and it is no answer to them to say that if they do not like what they hear in their workplace, they can go elsewhere. Third, the workplace is a context in which instrumental rationality is much more relevant that in society at large, 17. Meritor Savings Bank v. Vinson, 477 U.S. 57 (1986). On the analogy between sexual and racial verbal harassment in the workplace see Charles R. Lawrence III, "If He Hollers Let Him Go: RegUlating Public Speech on Campus", Duke Law Journal (1990): 431-83, p. 466 n. 127. Note that the Supreme Court in Meritor refused to extend the prohibition of sexual harassment to the racial one under Title VII, 447 U.S. at 67. Nevertheless this refusal was not due to the alleged First Amendment protection of racial insults but rather because the legislature has not determined whether racial insults constituted employment discrimination.

184

CHAPTER 6

and a demeaning, humiliating work atmosphere may result in the overall loss of an employee's morale, with negative consequences affecting efficiency in achieving the tasks of the workplace. After all, the main purpose of the workplace is to get work done. Even quite apart from the goal of protecting the employee, the employer may plausibly determine that restrictions on some types of speech are necessary to allow workers to concentrate on the job at himd, without the distraction of verbal provocation. That is why an employer should have a higher discretion with respect to content-based prohibitions on speech within the workplace, than should a legislator when designing a prohibition for a society at large. Universities and colleges constitute a peculiar category of special settings. Their peculiarity consists in their special nature which, at the same time, seems to warrant a more permissive and more restrictive approach to hate speech. The proponents of campus restrictions on hate speech argue that official toleration of racist insults interferes with the educational process. Hate speech hinders learning and class participation, undermines self-confidence and alienates a victimized student from the rest of the student body. As a result, hate speech adversely affects both academic performance and social integration within the university community. IS "Official tolerance of racist speech in [the university] setting is more harmful than generalized tolerance in the community-at-Iarge", suggests Matsuda. 19 More generally, a degree of civility is necessary for the conduct of the university's core activities (that is, learning and research), and offensive speech harms the university as a whole. On these grounds, some commentators argue that hate speech should be prohibited on campuses even in those instances where such speech could not be restricted within the general community: "The university's relationship to the speech of its members is fundamentally different from the state's.... Implicit in the university's core mission is the regulation of expression to enhance its quality".20 But there is an opposite side to the "university as a special setting" argument, and this is that universities, as places primarily devoted to learning and research, should be prepared to tolerate more speech than should the global society. This is because the "search for truth" is more dominant in universities as an aim of a collective enterprise than in society at large; because the universities' mission is to explore a wider range of ideas in a more thorough way than in the public discourse outside, for example in the mass media; because universities are laboratories of collective life where experimentation is more vital than elsewhere, and experimentation demands more tolerance of 18. See Lawrence, p. 465. 19. Matsuda, p. 2371. 20. J. Peter Byrne, "Racial Insults and Free Speech Within the University", Georgetown Law Journal 79 (1991): 399-443, p. 416; see also Beth C. Boswell-Odurn, "Note: The Fighting Words Doctrine and Racial Speech on Campus", South Texas Law Review 33 (1992): 26186, p. 266.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

185

expression which is seen as hannful by many, etc. A district court that struck down the University of Michigan's speech code emphasized that the principle of not prohibiting speech on the basis of its offensiveness acquires "a special significance in the University setting, where the free and unfettered interplay of competing views is essential to the institution's educational mission".21 But even apart from the educational mission, and thus apart from an instrumental argument, universities have traditionally been viewed as fora for uninhibited public discourse, exempt from many restrictions which are imposed upon public speech elsewhere. These special exemptions have turned universities in various countries and at various times into the frrst sites of dissent, before ideas born there have acquired a wider prominence in society. None of these arguments are targeted to the specific issue of legal toleration of racial insults, but rather support a more general thesis that universities' exceptionality, in general, provides an argument for less control of speech rather than more, compared to society at large. So it seems that in the case of universities, their exceptionalism may support arguments which pull in two opposite directions: in the direction of a higher, and also in the direction of a lower, tolerance for expressions of hate speech than in society generally. The purpose of these remarks is not to indicate which of these two sets of arguments should prevail, and therefore whether, on balance, universities should be fora of higher latitude for offensive speech or, to the contrary, they should impose more restrictions than, say, criminal codes or laws of defamation. But it may be useful to indicate a number of factors which can affect one's judgment on this issue. One factor is that there may be a difference between a public and a private university: private universities and colleges may be allowed more discretion in the regulation of communicative interactions within their campuses than public ones, which should more closely reflect the general legal standards of the nation. (This is the case in the United States where private universities are not bound by the First Amendment, though many comply voluntarily). Second, precisely because of the complex and divergent arguments about universities' exceptionalism, there is no need to impose uniformity in this regard within a given country. It seems proper that, since there are clear connections between the degree of freedom of expression and the purpose of the university, universities may decide for themselves whether they feel that, everything considered, the learning and research process is promoted by tolerating racist insults or by restricting them. Such a judgment should be a part of academic freedom, and the law of the nation should accord a high degree of deference to these judgments. Third, and related to the second point, if there is a great diversity of universities in a given country, then there should be a high degree of deference to academic judgment on that matter, because the students themselves may decide whether they will be more 21. Doe v. University ofMichigan, 721 F. Supp. 852, 863 (B.D. Mich. 1989).

186

CHAPTER 6

comfortable in a permissive-libertarian, or a protected, learning environment. But when the choice is meaningless (because, for example, there is only one university in each major city in a given country), the degree of law's deference for the university's judgment should accordingly be lower. Fourth, there is a difference between racially insulting remarks authored by faculty members, especially in the course of teaching, and those authored by members of the student body. The .latter resemble nonnal, "horizontal" communicative interaction in a society at large, and may be judged by similar standards. Consider, as an example, a student remark which sparked a controversy leading to the judicial invalidation of the University of Michigan's rules on discriminatory harassment in 1989. Under these rules, a graduate student was charged by the University for stating in a research class that he believed homosexuality to be a treatable disease. 22 In an interpretive guide prepared by that university. a number of examples of statements violative of the anti-harassment policy were listed, including: "A male student makes remarks in class like 'Women just aren't as good in this field as men"'. As much as both these statements may be hurtful to, respectively, homosexual persons and women, it seems that a classroom discussion is precisely the forum in which these views can be raised and a debate conducted in which these views and counter-views can be canvassed, particularly all the more so if these topics constitute the relevant subject matter of a lecture or seminar. Here a "marketplace of ideas" is certainly superior to a forced silencing of the offensive views. A somewhat different situation arises when the offensive views are authored by the lecturers because then they come from persons in positions of authority and with considerable power over students. The argument from hierarchy and subordination, mentioned earlier with regard to workplace harassment, applies here as well. A lecturer is not an equal participant in the marketplace (though, admittedly, individual teaching styles may range from deliberately egalitarian to strongly authoritarian) but has a high degree of control over the discourse and over the other participants. Students in a classroom are the "captive audience", in the sense that they cannot avoid hearing the speech without sacrificing some of their important interests. The degree of their "captivity" may be lower than that of employees in the workplace, because no-one is compelled to study in a similar way to which we are compelled to work, but the "captivity" of the audience is a matter of degree, and what counts is the personal cost incurred in escaping the unwanted communication. Fifth, racist remarks by faculty members in the course of their teaching may be the basis of a sanction, not because of their offensive impact, but as possible evidence of the low intellectual qualifications of a lecturer. It is simply not true that academic freedom precludes universities from making viewpoint-based 22. Id. at 865.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

187

judgments about the faculty members. To the contrary, such judgments are made all the time and they affect the sanctions and rewards issued by the university. Think of a history faculty member who espouses, in his or her teaching, a ''viewpoint'' that the Holocaust never happened, a lecturer in astronomy who approaches his subject from the ''point of view" of astrology, or a biology professor who teaches, with all seriousness, "creation science" as tlie correct interpretation of the origins of life on Earth. There may be some colleges and universities which will perhaps tolerate proponents of such theories, but it is very doubtful whether any respectable university in the world would feel compelled to tolerate teaching along these lines, purely on the basis of the principle of academic freedom. Rather, the sacking of professors espousing such theories in teaching would be properly viewed not as a matter of intolerance of an eccentric viewpoint but rather as a matter of the instructor's incompetence to teach history, astronomy or biology. By a similar token, some views which are offensive to racial or ethnic groups may be based on pseudoscience, and therefore can be properly targeted as revealing the insufficient qualifications of the faculty member concerned. Sixth, and finally, the purpose and the context of a given remark may be decisive for judging whether it should be tolerated in a university setting, regardless of its acceptability in society at large, precisely because of the university's special ends. For example, even if certain extremely racist texts would be proscribable by the law of a country, they may be useful in the teaching process either as a source of information about the history of ideas, or as material used for the "devil's advocate" argument in class discussion, etc. These are cases where students should actually learn to put up with, and critically analyse, materials which in any other setting they would legitimately refuse to read or see precisely because of the special nature of the university enterprise.

2

The Contours of "Racial ViUfication"

"Racist speech", "hate speech", "racial vilification", "racist hate messages", "hate propaganda", "incitement to racial hatred" - these are some of the labels given in legislation and literature to the types of group vilification which this chapter deals with. Note, incidentally, that while I will focus on the laws which prohibit vilification against racial and ethnic groups, a broader aim of this chapter is to discuss the legal and moral issues related to group vilification in general, whether these groups are defined on racial/ethnic grounds or on another basis, such as religion, gender, sexual preference, etc., which may spark social controversy and hatred. That is why the categories of "group vilification" and "racial vilification" will be used interchangeably here. While in societies such as the United States, Australia, and a number of European countries, hatred

188

CHAPTER 6

generated by racism is of particularly dramatic significance, it is not the only form of group hatred that should cause concern. Unless specifically indicated, nothing in the argument of this chapter binges on the specificity of race as opposed to other group denominators. This is not to say that racism does not involve any sui generis problems, but only that this specificity does not affect the general structure of the argument about the legal and moral justifications for anti-hate-speech legislation. Some legal definitions include the following indicia of prohibited speech: "propaganda ... based on ideas or theories of superiority .of one race or group of persons of one colour or ethnic origin, or which attempt to justify or promote racial hatred and discrimination in any form" (Article 4, International Convention on the Elimination of All Forms of Racial Discrimination); "incitement of hatred towards, serious contempt for, or severe ridicule of, a person or group of persons on the ground of the race of the person or members of the group" (Racial Vilification Act of New South Wales, 1989, art 1). A good definition of "racial insult" is provided by J. Peter Byrne: a verbal or symbolic expression by a member of one ethnic group that describes another ethnic group or an individual member of another group in terms conventionally derogatory, that offends members of the target group, and that a reasonable and unbiased observer . . . would conclude was purposefully or recklessly abusive. 23

In what follows, I do not propose to offer a comparative analysis of various definitions of racial vilification, but rather to indicate the broad contours of the category for the purposes of further discussion. A useful way of doing this is by locating racial vilification on a continuum of types of offensive, dangerous or otherwise objectionable speech. At one extreme of this scale are the categories of speech which, even to the most committed libertarian, would undoubtedly warrant restriction. At the other extreme are the cases of speech that, even to those who support the legal prohibition of group vilification, properly escape legal prohibition. By proceeding in a negative rather than a positive way (that is, by saying what group vilification is not, rather than what it is), I hope to minimize the danger of smuggling in substantive conclusions through a definitional fiat. This methodology seems to me to be justified in part by the fact that legal control of racial vilification is such a contentious thing: there are a number of competent, reasonable people who argue in favour of its prohibition and also a number of equally competent and reasonable people who argue against it. So it seems useful to "define" racial vilification by reference to those categories of speech about which there is a nearly universal consensus that they should be restricted (so that the definition of those categories is not affected by any uncertainty about their proscribable status), and by reference to 23. Byrne, p. 400.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

189

those categories about which there is near consensus that they should be left unproscribed, even though the speech may upset some people. I will begin by considering the non-punishable end of the spectrum of offensive or distressing speech, that is, by listing those cases of potentially offensive speech where a consensus may rather easily be found against any legal prohibition or punishment. First, it will probably be widely agreed that statements of verifiable truth should be legally protected even if a particular group may find them offensive. It may be painful or unpleasant, for instance, to a German minority in a particular country if 1 repeatedly remind the public that "the Germans initiated the Second World War" and yet 1 believe that the statement is a legitimate expression of a historically verifiable truth. There are, of course, problems with this: it may be said that it is offensive to the German nation as a whole to attribute to it the sins of its leaders or of its ruling elite. On balance, however, the statement appears to be a combination of historical observation and legitimate interpretation of facts, and few would consider it to be a case of national vilification. Mutatis mutandis, this first indicium of impunity of speech corresponds to a defence of truth in a civil suit of defamation. Second, the context of a statement may sometimes suggest that it was neither prompted by wrongful motives (such as the desire to vilify a group), nor that it would be likely to cause evil effects (for example, to cause offence). Consider, for example, words which in a different context might be extremely insulting, and yet which can be uttered affectionately with no intention to hurt the hearer. As Judge Kaufman said in a decision ofthe Court of Appeals of Michigan: [W]ords that generally refer derogatorily to racial or ethnic classes may, in some instances, not even rise to the level of insult. For example, I may refer to one of my colleagues by way of ethnic slang without any intended or resulting slur, insult or vilification. Rather, that remark would instead express the warmth and admiration I feel for that colleague. 24

In the 1989 debate on the Racial Vilification Bill in the Legislative Council (the upper chamber of state parliament) of New South Wales, one parliamentarian reported that he had been called "an old bastard" by another member of the Council on a number of occasions: "I know that on every occasion he has done so out of deep affection. Of course, on other occasions these same words, used in different context and with another inflexion of voice, could provoke a punch on the nose". Another speaker in the same debate recalled having been called "a pommie bastard" ("porn" is a mildly derogatory Australian term used to refer to migrants from Great Britain) after arriving in Australia, but agreed with an

24. Ledsinger v. Burmeister, 318 N.W.2d 558, 562 n. 2 (Mich. Ct. App. 1982).

190

CHAPTER 6

interjecting member of the Legislative Council that it could be "an endearing pbrase".2S Of course, relying on the context may lead to contentious specific judgments to a greater degree than if we had a clear, bright rule. Consider some hard cases. The context of the expression suggests that satire should enjoy protection from res1rictions, but would it also apply to racist jokes which may be just another way, and at times a very hurtful way, of propagating contempt towards a certain group? Perhaps it should depend on the setting and the intention: a hurtful racist joke included in a public speech addressed against a particular group may be just another fonn of an insult, indistinguishable from straightforward vilification; on the other hand, "[t]hat [racist jokes] are said with a smile, and in a context that is socially understood as a somewhat more appropriate venue for insult, changes their character somewhaf,.26 It also seems that context-related considerations require that academic or scientific inquiry which may be distressing to some groups and yet is not motivated by hatred towards those groups, should be non-proscribable. But does it follow that a scientific or pseudo-scientific theory purporting to prove the genetic or racial inferiority of a particular group be exempt from punishment under racial-vilification laws? To argue that the personal conclusions of an academic writer, drawn on the basis of empirical or statistical analysis, could be subject to legal restrictions if they hurt a group to which they apply, raises the spectre of censorship of unpopular academic ideas; certainly, a better remedy is to subject such conclusions to criticism within broader academic and public debate. Further, a fair report of a public act, even if the public act itself is proscribable, should no doubt be privileged. But consider a newspaper which selectively gives disproportionate coverage to statements made by racist politicians. It is not difficult to smuggle the racist message under the guise of objective reporting. Third, no restriction is warranted when objections to an act of speech result from someone's excessive or unusual sensitivity. Obviously this standard is not easily measurable, and there will be many controversial and borderline cases. But common sense suggests that there may always be instances of unusual sensitivity which should be disregarded. Which group's sensitivity should be ascertained? One answer is that the measure of "unusual sensitivity" must be gauged by the normal sensitivity ofa target group; otherwise it would be all too easy for a dominant social group to dismiss any complaint as a case of undue sensitivity. "Imagine how it would feel to be the female victim of a deliberately demeaning racial or religious epithet, only to be told by a white male judge that you have suffered a harm that the law does not recognize", suggests Jean

25. Hansard, 10 May 1989, NSW Legislative Council p. 7839 (remarks by Minister for Police and Emergency Sevices E.P. Pickering) and id. p. 7835 (remarks by the Hon. R.S.L. Jones). 26. Matsuda, p. 2369.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

191

LoveP And certainly it would feel even worse to be told that you suffered no harm at all, and that it is all in your mind. Fourth, acts of speech which are private rather than public should escape the threat of legal prohibition. Perhaps it would be pedantic to ask "what is public?" More often than not we rely on intuitive notions of what is "public", as with many such categories which belong to the "we know it when we see it" type of phenomena. That is why legal defmitions of what is "public" are often tautological. Consider, for example, New South Wales anti-racial vilification legislation which defines "public acts" as various forms of communication (such as speaking, writing, printing) "to the public" as well as "any [other] conduct ... observable by the public".28 Regardless of how we coin a positive definition of "the public", undoubtedly at least some types of speech are "private", such as inaudibly murmuring words to oneself. And yet, by definition, any communication is "public". So perhaps what matters is not so much the existence of an audience but rather the nature of the communicative interaction: some types of settings are private, by virtue of the social conventions which control access to a given place. For instance, dinner party conversation among a small group of hosts and invited guests will probably be found to be unquestionably "private"; it is no wonder then that, for example, Canadian provisions dealing with "hate propaganda" implicitly exempt "private conversation" from the offence of "wilfully promoting hatred against any identifiable group".29 But how far should we rely on dominant social conventions in defining "publicness"? After all, they are the very conventions that, more often than not, are responsible for racial stereotyping and vilification. In any event, one way of defining "public". speech which should be resisted is by using the content of speech as a criterion of "publicness". This may result in defining many instances of undoubtedly hurtful speech out of the category of proscribable racial vilification because they will be characterized as "private" in their content. This is how David McGowan and Ragesh Tangri reach the conclusion that the speech restricted by a number of university regulations is really "private", but they define as "public", speech on a topic which citizens should be, or are, concerned about: "Under no conception of democratic selfgovernance would we consider abusive epithets, directed solely toward an individual target, to be normatively significant. . . . Setting public figure cases to one side, the content of personal invective is almost by definition private rather than public speech".30 But under this definitional convention, the "private" character of speech is irrelevant to its proscribability: much of the 27. Jean C. Love, "Discriminatory Speech and the Tort of Intentional Infliction of Emotional Distress", Washington & Lee Law Review 47 (1990): 123-59, p. 149. 28. Anti-Discrimination (Racial Vilification) Amendment Act (NSW) 1989, art. I. 29. Canadian Criminal Code (1970) art. 281.2(2). 30. David F. McGowan and Ragesh K. Tangri, "A Libertarian Critique of University Restrictions of Offensive Speech", California Law Review 79 (1991): 825-918, pp. 856-57.

192

CHAPTER 6

speech that is "private" in the sense of its content is restrictable. Therefore, for current purposes, the private character of the speech will denote the method of dissemination rather than the content of the speech. Different commentators have suggested various other categories of protected, though possibly insulting speech: it was suggested, for example, that legal exemptions from anti-hate speech legislation should be granted to expressions of opinions on religious subjects and to speech that is part of artistic or entertaining expression, including racist humour. 31 But it seems that the four categories listed above (with all the reservations and doubts indicated) constitute the most important examples of potentially offensive speech which should remain legally permissible, and they will not be taken into account in this chapter as putatively prohibited instances of group vilification. I will now tum to the other end of the spectrum of speech, that is, to the categories of speech that are uncontroversially proscribable and which, therefore, do not enter into the concept of ''racial vilification" as used in this chapter. This end of the spectrum contains two categories: incitement and defamation. A number of provisions, both in international and national legal instruments, refer to "incitement to violence" alongside racial vilification. For example, Article 20 of the International Covenant on Civil and Political Rights declares that "any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law". Similarly, Article 4(a) of the International Convention on Elimination of All Forms of Racial Discrimination (1965) speaks of "incitement to racial discrimination, as well as all acts of violence or incitement to such acts". The Canadian Criminal Code prohibits the incitement of hatred against any identifiable group "where such incitement is likely to lead to a breach of the peace" (art. 281.2 (1». And yet, incitement to violence, constituting as it does the incitement to commit unlawful acts, is beyond the focus of this chapter since it does not raise similarly problematic issues from the point of view of freedom of speech. This is not to say that the law of incitement does not have the potential to damage freedom of speech. In fact, various legal systems and legal scholars have wrestled with the difficult problem of imposing some clear limits upon the punishable crime of incitement. For example, should exhortations addressed to the world at large be covered by the notion of punishable incitement? In a British case of the last century,

31. See, respectively, remarks by Kathleen Mahoney in ''The James McConnick Mitchell Lecture: Language as Violence v. Freedom of Expression: Canadian and American Perspectives on Group Defamation", Buffalo Law Review 37 (1988): 337-373, p. 351 (opinions on a religious subject); R.A. Smolla, "Rethinking First Amendment Assumptions About Racist and Sexist Speech", Washington & Lee Law Review 47 (1990): 171-211, p. 188 (artistic or entertaining expression); Matsuda, pp. 2368-69 (racist humour).

RACIAL VILIFICATION AND FREEDOM OF SPEECH

193

[t]he encouragement and endeavour to persuade to murder, proved at the trial, was publication and circulation by [the defendant] of an article, written in German in a newspaper published in that language in London, exulting in the recent murder of the Emperor of Russia and commending it as an example to revolutionists throughout the world}2

The link between the publication and the desired criminal effect seems to be very tenuous and yet there is no doubt that what was being advocated was a grave crime. In the United States it has now been accepted that "advocacy" amounts to punishable "incitement" only when it is directed to inciting or producing imminent lawless action and is likely to incite or produce such action. 33 Accordingly, United States law does not currently criminalize "the mere abstract teaching ... of the moral propriety or even moral necessity for a resort to force and violence".34 But these are controversies, as it were, within the doctrine of incitement, and they all assume that the causal connection is between speech and the unlawful action that the speech advocates; the controversies are -about the nature of the connection, not the nature of what is being advocated. Racial vilification poses different problems because it is not directed (not explicitly, anyway) at committing specific violent or illegal actions against its victims. If it is, it enters into the domain of "traditional" incitement and can be properly handled by existing anti-incitement laws. Of course, it may be argued that acts of racial vilification constitute a sort of incitement to violence or other unlawful acts. Indeed, it is frequently argued by proponents of the criminalization of group vilification that it is indistinguishable from incitement, or that it necessarily leads to incitement. But this is a matter of substantive argument rather than a definitional question, and it will be considered later, in Section 3.2 of this chapter. At this stage it suffices to say that if a given act of racial vilification does amount to the incitement of racially motivated violence, then there is no reason to resort to a separate crime of racial vilification. Racial vilification should be treated as a separate and distinct category which need not be intrinsically related to the actual incitement of racial violence. The same applies to defamation, as commonly understood in common law and in the criminal codes of various legal systems. For all the differences between the laws of defamation in different legislations, one thing is common: the prohibition of defamation is unproblematic when applied to the protection of "individual" targets of defamation. One general principle in the English and American laws of defamation is that it will be punishable only when targeting identifiable individuals. This is traditionally expressed in the principle that a 32. R. v. Most (1881)7 QBD 244. 33. Brandenburg v. Ohio, 395 U.S. 444 (1968). 34. Noto v. United States, 367 U.S. 290, 297-98 (1961).

194

CHAPTER 6

defamatory statement is not actionable unless published "of and concerning" the plaintiff. This does not mean that the plaintiff must be specifically identified so long as the tenor of the statement implicates him or her, rather than applying to "a large or indiscriminate number of persons described by some general name". This is so, in the words of Lord Atkin in a House of Lords decision in 1944, because of "the difficulty of establishing that the plaintiff was in fact included in the defamatory statement, for the habit of making unfounded generalizations is ingrained in ill-educated or vulgar minds, or the words are occasionally intended to be facetious exaggeration".3S There is a difference between a generalization that all members of a large group allegedly possess a certain discrediting characteristic and an accusation against a particular person or even a group of persons, each of which can be individually identified as being targeted by the speaker. Compare, on the one hand, the statement that "All Poles are liars", and the statement that all members of the governing board of the MiniMax Company have taken bribes. In the first case, an individual may feel defamed indirectly because the group to which she belongs has been offended as whole; in the latter case, the group membership is only a method of identifying all the individuals to whom a particular wrongdoing was attributed, either because the group is very small or because the accusation is clearly aimed at anyone of the group's members, rather than being a broad generalization. Whether group vilification should also be punishable is a matter of separate substantive and normative argument. Indeed, one of the frequent arguments in support of the prohibition of group vilification is that it would be inconsistent to prohibit individual defamation and to tolerate defamation that victimizes a person qua a member of a group. As David Riesman wrote in his classic article: "It is only through strengthening the protection of the groups to which an individual belongs that his own values and his own reputation can be adequately safeguarded".36 Sometimes it is even claimed that group offences are more hurtful to individuals than "personalized" ones. These are substantive, not definitional, arguments, and that is why they should be deferred until Section 3 of this chapter; at this stage, it suffices to say that nothing is gained by extending the notion of individual defamation to group vilification, and it is clear that there are many who would find the former, but not the latter, a proper object of legal concern. Since these two forms of injuring people's reputations raise distinct problems, their legal treatment calls for distinct normative arguments. The upshot of this section is that, for the purposes of our further discussion, a statement will not be characterized as "racial vilification" if it is verifiably true, or if it is made in a context which neither suggests a malicious intention nor a 35. KnupJJer v. London Express [1944] AC 116, 122. 36. David Riesman, "Democracy and Defamation: Control of Group Libel", Columbia Law Review 42 (1942): 727-80, p. 731.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

195

likely damaging effect, or it is made in private, or if the resulting offence stems from the unusual sensitivity of a member of a group. Furthermore, the concept of "racial vilification", as used here, does not embrace statements which are defamatory in the sense of individual defamation, or which constitute incitement to unlawful acts against a particular group. Having thus cleared the definitional field we may now turn to a normative discussion about whether certain harms resulting from racial vilification warrant legal proscription of this category of speech.

3

The Harms of Hate Speech

There are three principal types of harm that result from offensive utterances made about a particular group in a society: (1) those which are produced by a violent reaction by the victims of group vilification, (2) those which result from violence committed by other persons incited by "vilifiers" to assault the victims of racial vilification, and (3) those which consist of the very utterance of offensive words, regardless of any further violence committed. This is not to say that there do not occur other, collateral, types of harm which may be identified as ensuing from racial vilification. For one thing, one may argue that there is harm to the society generally, regardless of any more specific harm inflicted upon the direct targets of racial vilification. The broadcasting or publication of material which is offensive to particular racial or ethnic groups is damaging, not only to those directly offended, but also to the community: it degrades the standards of everyday life, of civility, of the quality of public discourse. Further, it undermines social peace and harmony, and the entire society suffers as a result. This may well be true, yet these harms will not be considered here as separate categories distinct from the three principal types of harm listed above. The reason is that these harms are either weaker than, or derivative from, the three principal harms. A case based on harm that consists of the lowering of the standards of civility in a given society is arguably weaker than a case based on harm that consists of provoking violence or inflicting psychic harm upon the target of vilification. One may well imagine that a society may be prepared to abandon uniform standards of civility on the basis of, say, a libertarian view about unrestrained self-expression. But this does not dispose of the argument regarding harm inflicted upon the victims of vilification. At any rate, it is questionable whether a liberal society may enforce, through legal means, its views about the standards of civility. As far as the breach of peace and the threat to social harmony are concerned, these results derive from the more specific harms listed above: namely, possible violence applied by the targets of vilification against the perpetrators or by third

196

CHAPTER 6

parties against the victims of vilification. These hanns only occur if the more direct harmful effects occur in the first place.

3.1 Violent reaction by the victim Consider this scenario: you witness someone giving a public speech in which your nation is vilified and ridiculed. The speech is deeply offensive to you, and you try to shout the speaker down. Someone in the audience, a supporter of the speaker, tries to carry you out of the hall, and you struggle to remain. A brawl follows. Who is to blame? Morally speaking, the blame is probably shared (though unequally) by the speaker, yourself and all those who took part in the violence. The speaker should have avoided the offensive language; you should have refrained from reacting violently; other members of the audience should have used peaceful means of solving the conflict between you and the speaker; and if these were insufficient, they should have called the police. But who should the police have charged as responsible for the violence? This is precisely the dilemma that is at issue; in order to decide the responsibility for the wrong, we first need to solve a normative problem about what should be a proper response to group vilification. In other words, what we want to know is whether making an offensive speech should have been prohibited in the first place, on the basis of its tendency to cause a violent reaction by a victim of the speech. An appeal to the responsibility for violence does not help because the placing of responsibility is the conclusion, not a condition, of a nonnative argument about the comparable costs and benefits of restraining the speaker, as opposed to restraining the targeted audience. Similarly, an emphasis on the speaker's right alone will not help. This is the strategy adopted by Joel Feinberg who implies that it is axiomatically true that if speakers have a right to say what they are saying (or if demonstrators have a right to do what they are doing), then a response by an enraged audience cannot be a ground for prohibition.37 But this proposed principle sounds strangely circular, precisely since what is questioned here is the scope of the speakers' (or demonstrators') right, including whether the scope of this right can be affected by the likely audience response. Feinberg seems to assume that which in fact should be left open to substantive argument. But while an assertion of the speaker's right cannot dispose of the problem here, undoubtedly it should be the focus of our attention before we consider the appropriateness of any particular response to the violence following from the speech. Violence that is likely to follow a vilifying speech is undoubtedly a type of harm that a state is obliged to prevent, but, on the other hand, a respect for speakers' rights demands (in accordance with strict scrutiny, as outlined in 37. Joel Feinberg, The Moral Limits of the Criminal Law. vol. 2: Offense to Others (New York: Oxford University Press, 1985), p. 90.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

197

Chapter 2) that the speech-harm relationship be direct and immediate, and that there be no other practical ways of avoiding the harm. Neither requirement is met in our hypothetical example. Violence directly flows from a listener's reaction to the speech, not from the speech itself. If the angry listener simply left the room, or reacted in a peaceful way, the brawl might have been avoided. Consequently, there is an alternative way of preventing the harm, other than by silencing the speaker: it is to insist upon self-restraint and self-control by the targets of the speech. The question is, of course, whether this is a reasonable thing to do, for by such an insistence, we seem to be shifting the burden from the perpetrator to the victim. To address this question one must introduce some further distinctions among different situations in which offensive words and utterances "assault" their targets. One distinction concerns the specificity of the assault: whether it is aimed at a particular person who happens to be in the audience, or is addressed to a group as a whole. The former comes close (depending on the specific circumstances) to "fighting words". Another distinction concerns the avoidability of the assault. While avoidability does not redeem the offensive speech, it nevertheless significantly weakens the speech-harm link. It is one thing to find oneself a member of a "captive audience" with no possibility of escaping a violent assault on one's reputation or dignity. It is another thing to suffer humiliation and ridicule where such effects could have been expected and avoided. As the United States Supreme Court once said: "Where a single speaker communicates to many listeners, the First Amendment does not permit the government to prohibit speech as intrusive unless the 'captive' audience cannot avoid objectionable speech".38 To express the idea that the hostility of an "uncaptive" audience cannot justify the suppression of the speech, the free speech jurisprudence of the United States coined the concept of a "heckler's veto". The idea is that to allow the hostility of an audience to warrant a restriction on speech would amount to making audiences the ultimate judges of constitutional rights. The result would be that we would "allow the intolerance (and threats) of a vocal minority (or even the majority) to determine who shall and shall not speak".39 The law, as many liberals proudly remind us, is there to protect speakers against intolerant audiences, not to protect audiences against unpopular and even offensive speakers. And yet, for all the persuasiveness of this principle, these words have a ring of unreality to them when applied to our subject matter. After all, the victims of racial vilification are, more often than not, dispossessed, disempowered and hopelessly poor minorities who are being ridiculed and offended by bigots and racists. Somehow, the examples of an insulted Pakistani youth in London, an 38. Consolidated Edison Co. v. Public Servo Comm'n, 447 U.S. 530,541 (1980). 39. NAACP Legal Defense & Educ. Fund V. Devine, 727 F.2d 1247, 1261 (D.C. Cir. 1984).

198

CHAPTER 6

Algerian worker in France or an Aboriginal in Sydney do not easily fl.t the stereotype of a ''heclder'' or a member of an "intolerant minority". On the other hand, a white-supremacist leader does not immediately evoke the idea of an unpopular speaker who should be protected against heclders. Significantly, the doctrine of not allowing "hostile audiences" to cause suppression of unpopular speakers originated from an opposite fact-situation to the one exemplified by typical cases of racial vilification. In the United States, the doctrine originally helped to secure the speech of civil rights demonstrators in the South who protested against the conditions of discrimination in the early 1960's. At that stage, the doctrine was fully compatible with the "fighting words" doctrine which refused protection to words that "by their very utterance inflict injury or tend to incite an immediate breach of the peace".40 The ''hostile audience" doctrine was associated with a positive assessment of the value of the speech in question, and consequently with a negative judgment of the audience's hostility to the speech. It was only subsequently that the dissociation occurred, eventually leading to the use of the "hostile audience" argument in the so-called Skokie cases,41 in which the hostility of Jewish survivors of concentration camps to neo-Nazi ideas was not allowed to affect the right of neo-Nazis to march with swastika signs throughout a Jewish populated suburb of Chicago. Irrespective of all the arguments in favour of viewpoint neutrality, surely we can draw a distinction between the intolerance of white bigots toward civil rights demonstrators, and the intolerance of Jewish citizens of Skokie towards the enthusiasts of gas chambers and concentration camps. This doctrinal evolution, consisting of the expansion of the "hostile audience"-based protection upon worthless and distasteful speech, was logically combined with the gradual erosion of the doctrine of "fighting words". This was because the original refusal to accord protection to "fighting words" could be maintained in its integrity only by confining the "hostile audience" doctrine to irrational, hurtful, illegitimate speech. In time, the thrust of the "fighting words" doctrine has been modified in a number of ways, of which the most relevant for our purposes is the insistence that unprotected "fighting words" designate "captive" situations in which the target has no reasonable means of escape. This seems to be crucial to a strict scrutiny of the speech-harm relationship, under 40. Chaplinsley v. New Hampshire, 315 U.S. 568, 572 (l942). For a good discussion of the evolution of the "fighting words" doctrine, see Downs, pp. at 631-36. 41. The Skokie incidents involved attempts by a neo-Nazi party to hold marches in a suburb of Chicago with a large Jewish population. Skokie took a number of legal measures to prevent the marches; all of these measures were largely unsuccessful. More specifically, Skokie unsuccessfully sought an injunction banning the march, Village 0/ Skokie v. National Socialist Party, 366 N.E.2d 347 (l977). Skokie then passed three ordinances that required permits for which the Nazis would not qualify, but the courts ruled that the ordinances violated the First Amendment, Collin v. Smith,447 F. Supp. 676 {N.D. 111.}, ajJ'd, 578 F.2d 1197 (7th Cir.). Despite this legal victory, the Nazis eventually abandoned their Skokie plans.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

199

which a restriction is validated only if it is narrowly tailored to the avoidance of a specific, severe harm. The question, from this perspective, is not whether group vilification is harmful and hurtful, since it certainly is. Rather, the question is whether there are ways of avoiding or minimizing the harm produced by speech related to public affairs other than by suppressing the speech itself. So far as we consider the volume of harm related to likely violent audience reaction, the answer largely depends on the ability of the target of speech to avoid finding oneself in a situation in which one may lose selfcontrol. Perhaps an analogy with the standards of a defence of provocation in criminal law may be relevant here; as one criminal-law scholar noticed, "racial slurs of an extreme kind might in some circles now be regarded as a more substantial provocation than the sight of marital infidelity".42 In English, United States and Australian criminal law (as in many other legal systems), an eXCUlpatory claim of provocation may, for instance, sometimes reduce murder to manslaughter. The standard of provocation in this case is as follows: "[Tlhe extreme emotional reaction of the accused and his loss of self-possession must be normal under the circumstances. His behavior must be of a kind that is occasionally part of the reaction repertory of most people, even though for most of us it does not often take a homicidal form".43 An English judge gave the following definition of provocation in his direction to the jury in a murder case: "Provocation is some act ... done by the dead man to the accused which would cause in any reasonable person, and actually causes in the accused, a sudden and temporary loss of self-control, rendering the accused so subject to passion as to make him or her for the moment not master of his mind".44 Applying these standards to racial vilification, provocation occurs when offensive remarks assault a victim suddenly and the target's reaction is immediate and spontaneous. But this does not extend to a situation in which a violent reaction could have been expected by the targets themselves, and avoided. That is why the Supreme Court of Il1inois reminded the Jewish residents of Skokie that "it is their burden to avoid the offensive symbol [the swastika] if they can do so without unreasonable inconvenience".45 Finally, the answer regarding restrictions of an act of speech on the basis of a violent reaction also depends on the intentions of the speaker: the context of the incident may suggest that the speaker actually wanted to provoke the victim to an angry reaction. In this case, the speech loses much of its privileged status: it 42. Hyman Gross, A Theory o/Criminal Justice (New York: Oxford University Press, 1979), p. 175. 43. Id., p. 157. 44. Devlin, J., quoted in R. v. Duffy [1949] 1 All ER 932. 45. Vii/age o/Skokie v. National Socialist Party, 373 N.E.2d 21, 26 (1978). See, similarly, Carl Cohen, "Free Speech and Political Extremism: How Nasty Are We Free to Be?", Law & Philosophy 7 (1988) 263-80, pp. 266-7.

200

CHAPTER 6

is part of the violent conduct, and the hostile remarks represent a course of action which should be viewed under the ordinary harm calculus, rather than enjoying the privileged status lent to them by the principle of freedom of speech. 3.2 Inciting others to commit violence Now consider another scenario. A delivers a speech in which he violently accuses members of an ethnic minority of his country of a high tendency to commit crimes, to lie, and to engage in bizarre religious practices. B, having heard the speech, becomes so enraged against the minority vilified by A that, the following day, he violently beats up C, a member of that minority. Is A responsible for this violence, and should the outcome determine the legality of A's speech? Most people would probably say that "it all depends", and that the criteria for legal prohibition should mainly refer to the proximity between the speech in question and the violent action, as well as the speaker's intent concerning the occurrence of these violent actions. People will disagree about whether both of these indicia must be present in order to make a speaker liable for the resultant harm, or whether either element alone constitutes a sufficient condition for liability. But there will probably be a wide consensus about the following two propositions. First, we cannot be held responsible for any harmful action anyone might take under the impact of our words - this would certainly stifle any free expression in society. After all, someone might always misinterpret or wrongly use our words, contrary to our intentions. Second, we cannot be absolved of all responsibility for the consequences of our words; for example, if we use someone else as a mere instrument for attaining our criminal intentions, then we should be rightly held liable for inciting others to violence. In order to chart a middle course between these two unacceptable extremes, one has to determine the requirements of "directness". One deceptively simple approach would be to apply a "but for" test: if it were not for the speech in question, violent actions would not have taken place. But the test of necessity is extremely difficult to apply to the connection between the speech of one person and an action of another. There is simply too much that we (and the court) do not know about the subjective conditions of the agent. Consequently, the only practical way to discern a causal connection between the speech and the action is to inquire into the substance and the circumstances of the speech, and the mental propensities of the agent. But then the ''but for" test virtually loses its meaning, for all we are saying is that a given speech was very likely to cause a particular response in a given person. Quite apart from this evidentiary problem, the "but for" test seems to be both too restrictive and too lenient. It is too restrictive because it would stifle a lot of speech which may be perfectly legitimate in most circumstances but, when

RACIAL VILIFICATION AND FREEDOM OF SPEECH

201

heard by a person with particularly violent tendencies, may lead to aggression. The test is also too lenient because it would leave unpunished an unsuccessful incitement to commit a specific crime a clearly counterintuitive consequence. Another approach is to draw a distinction between different types of expression which motivate others to act. Thomas Scanlon distinguishes between speech that provides others with the means to do what they wanted to do anyway, and expressions which motivate others to act by pointing out what they take to be good reasons for action.46 An example of the first kind of speech would be disclosing the combination of a bank safe to an intending robber; an example of the latter would be speech that provides arguments for why banks should be robbed. It is only in the former case, according to Scanlon, that a speaker may properly be held responsible for the harmful effect of another person's action. In the latter case, the responsibility of an agent who adopts the invidious reasons as his or her own, supersedes the agent's responsibility. "An autonomous person cannot accept without independent consideration the judgment of others as to what he should believe or what he should do",47 and it is this "independent consideration" by the agent which is a direct source of harmful action, thus "superseding" the speaker's arguments. There is a certain initial appeal in this proposal. It is based on a general liberal intuition about individual autonomy: everyone takes responsibility for his or her actions, and it is not the role of the government to protect mature, adult individuals from the influence of other people's pernicious ideas. "The harm of coming to have false beliefs is nat one that an autonomous man could allow the state to protect him against through restrictions on expression", says Scanlon.48 In a liberal state everyone has a moral duty to reject ideas the execution of which would harm other people. But, while this is attractive at first blush, the principle based on Scanlon's distinction would also lead to strongly counter-intuitive results. Suppose I keep suggesting to you, step-by-step and persistently, that there are excellent reasons for you to kill your neighbour who is a police officer. I impress upon you that society would be better off without police, that the police officers are enemies of the common people, that they are dangerous to all of us because they are the armed instrument of a governmental conspiracy against its own people, and that the best way to get rid of this social evil is to kill each and every one of them. You listen carefully and offer counter-arguments, but I destroy them one by one very effectively. After a long and persistent indoctrination of this kind, you 46.

47. 48.

Thomas Scanlon, "A Theory of Freedom of Expression", Philosophy & Public Affairs 1 (1972): 204-26, p. 212. Note that more recently Scanlon retracted the part of his theory which hinges upon this distinction: T.M. Scanlon, "Freedom of Expression and Categories of Expression", University ofPittsburgh Law Review 40 (1979): 519-50, p. 532. Scanlon, "A Theory", p. 216, emphasis added. Id., p. 217.

202

CHAPTER 6

murder your neighbour. Am I not, both morally and legally, liable for the crime? I have provided you with reasons only, not with the means, and yet surely my responsibility cannot simply be "superseded" by your guilt in acting on such pernicious opinions. One might say, of course, that a rational and basically moral person will not come to hold beliefs like those, even after continual persuasion. And yet we know that not everyone in a society is equally rational and moral, and we cannot and should not found our principles about moral responsibility for evil committed to others upon an abstract dogma about the perfect rationality and basic morality of all adult individuals in society. Some people are more susceptible to aggressive ideas than others, and this fact must weigh upon our decisions concerning the exercise of our freedom of expression. Much of the judgment about this weight is context-bound. It is one thing to test arguments made during a university seminar in ethics about an alleged moral duty to kill police officers, where we know and trust the good judgment of all the seminar participants; it is another thing to use these arguments in a speech addressed to a large crowd in which many individuals are unknown to us. In other words, we have a responsibility for anticipating the likely results of our acts of expression if these results may entail violence done to other people, even if the connection between our expression and another person's violent action merely consists of providing the agents with reasons to act in this particular way. In our example, moral responsibility for the resultant murder is shared (albeit unequally) by both the inciter and the actual agent, and the fact that the agent's independent judgment came into play to motivate the action, does not fully "supersede" the role played by the speaker's acts of persuasion. Here, my intuitions coincide with those of Robert Amdur: "When a speaker or writer persuades or incites another person or persons to commit a harmful act, our intuitions suggest that the speaker or writer bears some responsibility for the resultant harm".49 This example helps us to recognize not only the existence, but also the limits, of responsibility for inciting others to violence. There is a fme and yet crucial line between me telling you that all police officers are evil (and that the world would be a better place without them), and my advice that you should murder a police officer. In the former case, I may be free of any murderous intentions towards police officers. I may, indeed, be strongly opposed to any violence against them and the moral link between my anti-police views and your criminal action may be quite remote. There is a mediation of your views about the practical action to be taken and, hence, no necessary connection exists between those abstract views and your practical action. The difference here is one between propagating a belief and advocating an action; in the words of an 49. Robert Amdur, "Scanlon on Freedom of Expression", Philosophy & Public Affairs 9 (1980): 287-300, p. 295.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

203

American judge, the "essential distinction is that those to whom the advocacy is addressed must be urged to do something . . . rather than merely to believe in something".50 But this is not enough because an expression of a belief may sometimes be a disguised form of advocacy: the surface structure of the speech may, in itself, be misleading. There may be a very fine line between advocating a belief and advocating an action which consists of acting upon that belief. Ycit is there really any qualitative difference between "a call to violence" and "the teaching of the moral propriety or even moral necessity for a resort to force and violence",51 as the United States Supreme Court once suggested? It would probably be unwise to merely rely on the semantic form of an expression in giving effect to a distinction which marles a dramatic difference between impunity and criminal conviction. Something like a "clear and present danger" test, or an equivalent, must therefore be introduced in order to judge the immediacy of tbQ}Urreat produced by one's words, even if these words sound more like a statement of belief. This test has the advantage of contextualizing the analysis of the speech-act connection, and also of diverting our attention away from a formal, semantic structure of the expression. But again, it is not without its own problems. The test was initially formulated by Justice Holmes: ''The question in every case is whether the words are used in such circumstances and are of such a nature as to create a clear and present danger that they will bring about the substantive evils that Congress has a right to prevent. It is a question of proximity and degree".52 It was subsequently reformulated in a landmark 1969 decision of the United States Supreme Court which declared that the advocacy of law violation was protected by the First Amendment unless such advocacy was directed to inciting or producing imminent lawless action and was likely to incite or produce such action. 53 But in that same case Justice Douglas wrote a powerful dissent in which he claimed that the "clear and present danger" test was irreconcilable with the First Amendment in days of peace, because of its anti-libertarian consequences. According to Douglas, the "clear and present danger" test was discredited during the Cold War, when it was "so twisted and perverted" as to "erod[e] substantial parts of the First Amendment.54 This immediately suggests that the test is a double-edged sword: it may justify confining punishment for speech only to cases of imminent danger but it also may justify the expansion of responsibility for "abstract advocacy" if the danger of other people acting upon this advocacy is very significant. Part of the problem is related to the fact that the test incorporates two distinct measures: imminence and gravity of danger, 50. 51. 52. 53. 54.

Yates v. United States, 354 U.S. 298, 325 (1957) (Harlan, 1., for the Court). Noto v. United States, 367 U.S. 290, 298 (1961). Schenck v. United States, 249 U.S. 47, 52 (1919). Brandenburg v. Ohio, 395 U.S. 444,447 (1969). Id., p. 454 (Douglas, J., dissenting).

204

CHAPTER 6

and these measures can be traded-off against each other. Under some interpretations. the volume of the threat must be discounted by its likelihood: an anti-libertarian implication is that even if the threat is not very likely, it may be deemed to pass the "clear and present danger·' threat if the evil advocated is sufficiently serious. The two criteria: advocacy of unlawful violence and the immediacy of danger, applied independently of each other, seem to describe a plausible category of punishable incitement to crime. This is not a strictly libertarian position: a libertarian would demand a cumulative application of the two principles, so that the absence of either would make punishment for the speech, or restriction of the speech, unjustified. And yet, it seems that the advocacy of assault against other people should be penalized even if the harm is not imminent. This resonates with the traditional English common-law misdemeanour of incitement, where the elements of the offence do not require the actual performance of the indictable crime by the principal offender so long as the anticipated principal offender heard and comprehended the words of incitement. Of course, the "clear and present danger" requirement does not mean that a crime has to actually be committed for the incitement to be punishable but only that it is very likely that a grave unlawful action will be taken as a result of incitement. But the absence of the principal offence may be used as possible evidence that the danger was not clear or present. The prohibition of incitement to violence against members of a racial group derives from the prohibition of the principal crime of assault against others. A development of an argument supporting this proposition is outside the bounds of this book, though it must be admitted that. as stated, the proposition is question-begging. It would be theoretically challenging to explain the idea that the unlawfulness of advocacy of crime derives from the unlawfulness of the crime itself, even if no crime has been committed as a result of the advocacy, and even if a danger of committing the crime is not imminent. On its face, the idea is vulnerable to the criticism that it "will often lead to justification of suppression of illegal advocacy that presents absolutely no danger of harm to anyone".55 But it is irrelevant to our current discussion because the types of racist speech which are caught under racial vilification laws (the actual and the proposed ones) do not amount to "advocacy" of violence. Whether advocacy of violence should be prohibited only when imminent danger is present (as libertarians suggest) or always, does not affect the argument about racial vilification. On the other hand, even the mere spreading of "beliefs" (as contrasted to advocacy of illegal acts) can be restricted under the proposed conception here if there is a clear and present danger to particular people and groups in society. 55.

Martin H. Redish, Freedom of Expression: A Critical Analysis (Charlottesville, Va: Michie, 1984), p. 199.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

205

But the spreading of offensive beliefs about groups should not be prohibited under this category of harm if there is no imminent danger of violence. Ultimately then, the argument for penalizing group vilification on the basis of its tendency to induce others to act violently cannot be based on incitement, that is, on such advocacy of harm where the connection between the speech and the harm is proximate and straightforward. These situations are handled by imposing penalties for inciting others to commit a crime, which is in itself a criminal offence recognized in many, if not all, legal systems. In Australian and English law, for instance, a person who counsels, commands or advises the commission of an indictable crime is guilty of the common law misdemeanour of incitement. The subjective condition (mens rea) of incitement is an intention that the substantive offence be committed. The common law also acknowledges the category of "accessory before the fact", that is, a person who encourages the commission of the crime, but is not present at its commission. In essence it has been held that the accessory is liable, as is the principal offender, if the accessory has counselled or procured the commission of an offence.56 Hence, there is no need for a new criminal category to describe incitement to racially motivated violence as the offence is already handled by general antiincitement rules. Penalizing group vilification therefore applies to situations which are not captured by criminal "incitement", as it has been traditionally interpreted, but to those cases where the connection between speech and another person's criminal behaviour is more remote. But this is a dangerous consequence because where the peril is not clear andlor present, or where the speech advocates hatred (which, in itself, is not illegal) but not violence, it is both unfair to a speaker and stifling to the general exchange of views in the society, to hold the speaker responsible for violence which may occur at a later point. Such responsibility might perhaps be imposed upon the speaker if the principle of freedom of speech did not confer any privileged status upon speech; in other words, if the allocation of legal liability was a simple matter of a calculus of costs and benefits. We might then plausibly say that widespread racist propaganda may, in the long run, contribute to an increased level of racist violence in society, and therefore that the costs of legally tolerating racist speech are higher than the costs of the restriction. But if the principle of freedom of speech demands particularly strong justifications for restrictions which, in the case of incitement, translates into a search for a clear, direct and immediate connection between the prohibition of speech and the prevention of an identifiable harm, group vilification which falls short of incitement to crime cannot be prohibited on this particular ground.

56. See R. v. Creamer [1966] 1 QB 72.

206

CHAPTER 6

3.3 Psychic injury Now consider a third scenario. Imagine that you are a recently naturalized resident and, since you want to be an active and loyal citizen of your new country, you go to a meeting of the town council in your suburb. At the fIrst meeting, quite unexpectedly, a speaker rises and makes a fiery speech in which she proposes a zoning ordinance which will make it more difficult for new immigrants, especially from your home country, to settle in the suburb. This, she says, is to protect the safety of children and the standards of cleanliness. She then discusses in detail the defects and sins of your national group, emphasizing your genetically inferior intelligence as evidenced by a recently published book on the topic, and your lack of a developed moral sense. She ends her speech with a few jokes about your nationality - jokes which you recognize as substantially similar to the jokes that people in your old country used to tell each other about a neighbouring, and widely despised, superpower. You do not react violently to the speech: you are too frightened, too intimidated, or too embarrassed about your distinctly foreign accent, or perhaps it is simply that violence is not your way of reacting to insults. Nothing in particular happens after the incident; no-one assaults you in any perceptible, physical way . Yet, since the day you heard the speech, your life has clearly been transformed for the worse. Whenever you meet your neighbours or co-workers or salespersons in the shops, you search for expressions of dislike or contempt in their eyes. When they are rude, you attribute it to their hatred of your ethnic group. When they are polite, you treat it as a symptom of their patronizing attitude, or their attempts to protect you from further distress. They know that you are a member of your particular ethnic minority, and that your minority has not been getting good press in this country for some time now. You know that they know. And they know that you know that they know. Being an educated person, you recall a book by Erving Goffman who described this phenomenon as "the infinite regress of mutual consideration,,;S7 you also recognize in your own psyche those same sensations that Goffman claimed are produced by an ideology of stigma: self-isolation, anxiety and uncertainty (though the victims in Goffman's book were stigmatized for different reasons to you). From now on, your relationship with others can no longer be normal because you and them are entangled in a web of mutually self-reinforcing attitudes of suspicions, fears and dislikes. Your self-image is inevitably shaped by your belief that others are contemptuous of you or that they feel unwanted pity, which at times feels even worse. You begin discerning a bizarre ambiguity, not to say schizophrenia, in your own attitudes towards your ethnic identity: you are proud of your heritage but, at the same time, you wish you could get rid of the difference that costs you 57. Erving Goffinan, Stigma: Notes on the Management of Spoiled Identity (Hannondsworth: Penguin, 1968, 2nd ed), p. 30.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

207

so much in your new country which you enthusiastically wanted to embrace. You recall another quotation from Goffman which corresponds to this distressing ambiguity of yours: Instead of cowering, the stigmatized individual may attempt to approach mixed contacts [i.e. contacts with "non-stigmatized" individuals] with hostile bravado, but this can induce from others its own set of troublesome reciprocations. It may be added that the stigmatized person sometimes vacillates between cowering and bravado, racing from one to the other, thus demonstrating one central way in which ordinary face-to-face interaction can run wild. 58

In this "vacillation between cowering and bravado" you are confused, upset, distressed and angry. And, above all, you are deeply offended. Liberal political philosophy is replete with arguments about the insufficiency of ''mere offence" in warranting the prohibition of an offensive expression. Liberal jurisprudence proudly declares that the offensiveness of speech is not a good reason for restricting it: ''the public expression of ideas may not be prohibited merely because the ideas are themselves offensive to some of their hearers".59 Indeed, some liberals go a step further and claim that the fact that a given instance of speech is offensive or distressing is all the more reason to protect it against restrictions: "[T]he fact that society may find speech offensive is not a sufficient reason for suppressing it. Indeed, if it is the speaker's opinion that gives offense, that consequence is a reason for according it constitutional protection".60 And yet, for all the attractiveness of the general principle, and for all the edifying talk about protecting unpopular views against tyranny by the majority, there is a strange lack of correspondence between these high liberal proclamations and such personal experience of hurt and humiliation as described above. The victim in my example would be genuinely surprised if he was told that the damage done to him by humiliating racist speech can be summarily described as "mere offence". It is more than just being disgusted or even shocked at an obscene picture or a rude word. In our example, he was harmed right in the very heart of his own identity, in the core area of his own self. Liberals traditionally have had trouble grasping the severity of, and relating sympathetically to, the kind of psychic harm inflicted by group vilification. A possible reason for this is that they have usually applied to such psychic harms, the paradigm of members of the majority offended by unorthodox, shocking minority expressions. Within this paradigm, the power of the state to protect people against harm to their sensibilities would indeed "effectively empower a 58. Id., p. 29. 59. Street v. New York. 394 U.S. 576, 592 (1969). 60. FCC v. Pacifica Foundation, 438 U.S. 726, 745 (1978).

208

CHAPTER 6

majority to silence dissidents simply as a matter of personal predilections".61 Consequently, some liberals point out that "if 'offensiveness' were the test, majority rule would replace the ftrst amendment".62 And liberals have generally been strengthened in their resolve by the attitude of conservatives who keep proclaiming, unabashedly, their commitment to the protection of majority sensitivities. Consider this, rather incredible, extension of the offence principle by the Chief Justice of the United States: "Surely one of the high purposes of a democratic society is to legislate against conduct that is regarded as evil and profoundly offensive to the majority of people - whether it be murder, embezzlement, pollution, or flag burning".63 Against such ideological opponents, no wonder the liberals are so adamant about sticking to their principle that the majority's offended sensitivity cannot, in itself, warrant a prohibition of offensive speech. But the pattern of the majority versus minority clash in racial viliftcation laws is quite different: it is not a case of the majority trying to silence the minority, but rather of a minority trying to silence the intolerant majority, or trying to enlist the support of the majority in silencing a vicious minority. The silencing involved in enforcing anti-racial-viliftcation law is not the kind of silencing normally associated with majoritarian tyranny. A minority group that seeks help through anti-racial-viliftcation laws is precisely the sort of group which has traditionally been seen by liberals as deserving special legal protection against possible majoritarian oppression: a powerless, subordinated and disadvantaged minority. The whole pattern of minority protection versus majoritarian oppression is, in the case of racial viliftcation, opposite to the sort which prompts many liberals to raise the alarm against· granting legal recognition to one's aversion towards others' speech. In contrast to the orthodox liberal view, I will take here as obvious something that for many other liberals is question-begging: offensive remarks about one's group do infliCt harm upon members of the vilifted group. The central question is whether, and why, this harm is different from the sort of harm that unobjectionably leads to the legal prohibition of other harmful conduct. To clear the fteld for the argument, let us ftrst quickly dispense with the old liberal chestnut that prohibition of speech on the basis of its offensiveness would annihilate any freedom of expression because (as the argument goes) it would give anybody an effective veto over another person's exercise of his or her right to speak. The argument is similar to the one made a long time ago by H.L.A. Hart against punishment for distress which is incidental to the belief that others are acting in a wrong way. Hart claimed that 61. Cohen v. California,403 U.S. 15,21 (1971). 62. Steven Shiffrin, "Defamatory Non-Media Speech and First Amendment Methodology", UCLA Law Review 25 (1978): 915-963, p. 951. 63. Texas v. Johnson, 491 U.S. 397,435 (1989) (Rehnquist, C.J., dissenting).

RACIAL VILIFICATION AND FREEDOM OF SPEECH

209

"[t]o punish people for causing this form of distress would be tantamount to punishing them simply because others object to what they do; and the only liberty that could coexist with this extension of the utilitarian principle is liberty to do those things to which no one seriously objects".64 This was reiterated by David Richards: [T]he equal liberty principle absolutely forbids the prohibition of ... communications on the ground of . . . detestation and outrage alone. Otherwise, the liberty of expression . . . would be a pitifully meager pennission allowing people to communicate only in ways to which no one has any serious objection.65

This is a considerable overstatement. From the suggestion that people's objections based on the offensiveness of speech may count in a decision about restrictions of speech, it does not follow that only speech to which no one objects will be allowed. What does follow is that the harm of speech, consisting of offending some people, will be taken into account along with other factors in the overall calculus of harms and benefits before any decision about the restriction is made. This is, after all, what we routinely do with regard to all other kinds of conduct. Some offensive speech acts will pass the test successfully (if the benefits outweigh the costs related to the offensiveness); others will not. But this is a far cry from Richards' proposition that unless we absolutely reject any offence-based restrictions of speech, any single person will have a veto over what can be said. The issue at hand is not just any offensiveness of a speech-act. Whether some shocking words and images that hurt individual sensibilities should be legally prohibited is not the question here. We are concerned with a particular kind of harm to one's sensibility; that is, the harm resulting from the vilification of a victim's group. This is a different matter to someone simply claiming protection from unwanted exposure to words and images that he or she fmds disgusting. The difference is mainly in the degree of implication of one's own identity. If someone protests against unwanted exposure to pornography or indecent language, then one is more remotely and indirectly implicated in the subject matter of the offensive material than when one protests against contemptuous or hateful statements concerning one's own racial or ethnic group. It is a matter, so to speak, of having the moral standing to protest. This is important to stress because some liberals have a tendency to subsume groupvilification actions under a broader category of offensiveness, and then they can easily dispense with the problem by appealing to the principle that

64. H.L.A. Hart, Law, Liberty, and Morality (Stanford: Stanford University Press, 1963), p. 47.

65. D.AJ. Richards, "Free Speech and Obscenity Law: Toward a Moral Theor'y of the First Amendment", University ofPennsylvania Law Review 123 (1974): 45-91, p. 68.

CHAPTER 6

210

offensiveness alone is an insufficient ground for restriction. This is a non sequitur because group vilification is not a matter of "mere offensiveness". But the matter is more complicated than that. Forget racial vilification for a moment and think of some other types of group vilification. Blasphemy is an example of speech that can be characterized either as causing ''mere offence" or as "group vilification", depending on the hearer's perspective. If I identify with my religion strongly (which is, after all, what all religious adherents are supposed to do), then it means that offensive remarks about my religion do not merely "offend" me; they implicate my own identity in a way which expresses contempt, humiliation and hate towards me. To say that a Muslim is offended by "Satanic Verses" probably does not fully capture the effect of Rushdie's novel upon the psyche of many Muslims: it hurts them personally because they see themselves implicated in a story which (they may think) expresses contempt for their group. And since their group moulds their personal identity to a high degree, they feel hurt in themselves. The case of pornography is similar: to some viewers it may be merely offensive, but to those women who see it as an expression of contempt for women, or as part of an ideology that treats women as objects of sexual exploitation, it is more than merely offensive: it is an insult to themselves in a way that implicates their own identities. This is not merely a problem of where to draw the line between "mere offence" and offensiveness which implicates the identity of a viewer or a hearer. The problem is more fundamental: whether we classify a given offence as falling on one side or the other of the distinction ultimately depends upon our substantive assessment of the merits of the vilifying speech. And this has some disturbing consequences. To reflect upon this, consider the issue of symbolic conduct. According to some commentators, acts such as wearing a jacket bearing profanities in order to voice an objection to conscription cannot be prohibited because, in contrast to group vilification, such acts do not harm anyone. In the words of a writer who is otherwise in favour of criminal liability for group vilification, symbolic offensive speech would escape punishment because "an essential characteristic of symbolic speech ... is that by definition it poses no serious harms to substantial public interests".66 This is supposed to contrast with harm caused by "fighting words" and other categories of nonprotected speech. But the distinction is ultimately in the eyes of the beholder: one may well imagine a situation in which harm to one's psychological wellbeing and self-respect, caused by unexpectedly sighting a "Fuck the Draft" slogan, may be serious and may implicate one's own identity. In particular, consider a disabled war veteran, or the parents of a soldier who is currently participating in a war, etc. Hence, the judgment that no serious harm to one's psyche is caused by a particular offensive sign hinges upon a judgment about 66. Note, "Group Vilification Reconsidered", Yale Law Journal 89 (1979): 308-332, p. 322 n.59, emphasis in original.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

211

the substantive value of this sign, and more importantly, about whether the viewer is justified in finding his own identity implicated by the sign. Another example is provided by an argument in a Harvard Law Review note which advocated penalizing group vilification. The author(s) suggested that "the law should recognize the sensibility harm experienced by the targets of group vilification",67 which is distinct from'''interest harms" (the sole form of harm so far recognized by courts in the United States). The difference is that in the case of "sensibility harms" the harm cannot be defmed independently of a hearer's attitude towards the speaker's point of view, while in the case of "interest harms" it can. The note's recommendation for such an extension appeals directly to a communitarian vision of the individual and society: "the communitarian argument holds that the law should recognize the sensibility harm experienced by the targets of group vilification, because to disregard it is to compromise the shared commitments that make freedom possible".68 So far so good. The problem, though, is that if we allow all "sensibility harms" to be protected by law, we would end up restricting even the most valuable and justified speech if the sensibility of a criticized person was harmed. What about banning civil rights demonstrations which may offend the sensibility of a racist? To escape such counter-intuitive conclusions, the authors of the note draw a further distinction between "instrumental" and "constitutive" communities; it is only in the latter communities that members' identities are directly affected by statements relating to the community as a whole. Most communities, we are told, are instrumental, and so do not warrant sensibility-harm based protection against group vilification. For instance, we are assured that a white bigot will not be able to claim protection for his "sensibility harm" against civil-rights activists, because the communities affected by anti-discrimination speech, such as the Ku Klux Klan, are not constitutive. 69 The aim of this argument is clear: it is to reconcile the prohibition of group vilification with the denial of such protection to "wrong" groups. It is also perhaps to anticipate and disarm an argument that, while protecting the grouprelated sensibilities of the listeners by banning racially vilifying speech, we are trampling upon the group-related sense of identity of the speaker; after all, as Byrne noted, "if society accepts the primacy of group identity, by what shared values can it justify to the speaker restrictions of his outspoken preference for his group to the listener's?,,7o But the note's argument merely purports to draw a neutral and clear line between constitutive and instrumental communities; in fact, the operative line is between those communities which we approve of and 67. Note, "A Communitarian Defense of Group Libel Laws", Harvard Law Review 101 (1988): 682-70 I, p. 692. 68. Id., p. 692. 69. Id., p. 694. 70. Byrne, p. 409.

212

CHAPTER 6

those which we do not. Consequently, the distinction has nothing to do with the extent to which one's identity is defmed by one's group membership. A Klan member may well be psychologically and morally affiliated with his organization to a higher degree than many other people affiliated with their nations or their religions. His identity may be shaped by his Klan membership to a very high degree. And yet we (rightly) deny him protection from even the most severe and offensive criticism of th,: Klan. We do it, not because we doubt the seriousness of his identification with the Klan, but because we believe that the Klan's ideas are not worthy of protection. Hence, the distinction between constitutive and instrumental groups is just a proxy for a substantive judgment regarding which group identities deserve protection and which do not. One other way of defining the "constitutive" characteristics would be by identifying them with the "immutable" properties of individuals, as opposed to those properties which individuals can alter through their own action. 71 There is a certain immediate attractiveness in this proposal: it would easily allow us to draw a line between offending a racial minority and offending Klan members, or members of any other voluntary association for that matter. It would also resonate with an intuitive feeling that there is something particularly wrong with inSUlting people who are selected on the basis of characteristics which are beyond their control. But once we articulate a reason (or, the reason) for according immutable characteristics a special protection from offence, the argument immediately collapses and the initial intuition loses much of its allure. One such reason would be to say that immutable characteristics are, by their very nature, much more tightly linked to the identity of an individual than are the alterable characteristics which are more defining of a person's changeable roles in society. Under this argument, "immutability" is just a proxy for identity-defining characteristics. But, unless this equivalence is a matter of definition, so that anything that is immutable is defined as identity-constituting, in which case the argument is circular, this is a very imperfect proxy. There are some characteristics which are immutable but which do not define anything particularly significant about an individual identity (for example, freckles on one's back), and there are a~so characteristics which define an individual to a great degree but which are alterable (for example, membership of a political party by a fanatical activist who devotes all of his or her time and effort to that particular party). A second reason why one might choose "immutability" as a criterion of constitutive characteristics for the purposes of a racial vilification discussion, is on the basis of an argument that offence along lines which do not give a victim any opportunity to escape an offended group is more severe, or unfair, than offence which gives one a chance of escaping the offended group. But again, the 71. See McGowan and Tangri. pp. 865-66. It has to be pointed out that McGowan and Tangri canvass the immutability criterion, but do not endorse it.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

213

very articulation of this reason is sufficient to discredit it. It is analogous to an argument that discriminating against blacks would be considered less harmful if blacks could easily change their skin colour. A heightened protection against offence Gust as a heightened protection against discrimination) should not be contingent upon the inescapability of a protected category. Consider the case of offence against (or, in a parallel argument, discrimination against) gay men and lesbians. Would it make any difference to the wrongness of the offence if we found a strong confIrmation that sexual orientation is an inborn characteristic rather than a freely chosen "life-style"? A corollary to the argument that "alterability" of sexual orientation lessens the protection against homophobic insults is an argument that gay men and lesbians may easily avoid vilifIcation by changing their sexual orientation to a heterosexual one, and so no heightened protection - reserved for truly immutable characteristics - is allegedly warranted. But this argument collapses, because the central question is whether it is fair for society to impose a penalty (in the form of less protection against group vilifIcation) upon members of a group who do not renounce their particular conduct, values, or set of beliefs. The weakness of the immutability criterion is particularly obvious if we consider the case of religious groups. Religion is usually listed, alongside with race, as one of those characteristics which should constitute a special ground for protection against vilifIcation. SuperfIcially, it could itself be an argument against the "immutability" theory because, while religion is certainly viewed as a matter which constitutes a person's identity to a very high degree (at least in the case of a great number of believers), there is nothing "immutable" about it. But perhaps the immutability argument should be split into one of "entry" and "exit" points: it might be claimed that, while we all have the opportunity to abandon a particular religion (the "exit"), we do not "enter" religions voluntarily but are born into them. But this is obviously untrue for a great number of people: for non-believers who at a certain stage of their life discover a religious truth for themselves, or for those who convert from one faith to another. Why should the level of protection against vilifIcation be higher for those people whose religious affiliation was not a personal matter of conscious choice? The only reason for such a suggestion would be by virtue of an identifIcation of the involuntariness of a given characteristic (lifestyle, beliefs, preferences, etc.) with the depth of commitment. But this is patently implausible. Just to the contrary, it would seem that we are often more committed to those values, relationships and lifestyles which we have chosen voluntarily, or at least those which we can abandon but decide to keep, notwithstanding the exit option. So the immutability argument seems to be at odds with the "constitution of identity" argument if the depth, or strength, of the commitments are seen (as they should be) as indicative of those characteristics which defIne a person's identity to a very high degree.

214

CHAPTER 6

The religion example also indicates that the very act of drawing a line between immutable and mutable characteristics, and then characterizing a given property as falling on one or other side of the line, is not neutral towards the members of a group defined by this property. In the case of religion, whether you believe that religion is an immutable or mutable characteristic depends. to a large extent on whether you are religious or not: secular liberals tend to describe religion as an "alterable" characteristic which can be affected by a person's choice, while those with a religious outlook complain that such an approach trivializes and distorts the nature of religious commitments. According to Michael Sandel, a liberal approach "depreciates the claims of those for whom religion is not an expression of autonomy but a matter of conviction unrelated to choice" and fails "to respect persons bound by duties derived from sources other than themselves".72 Perhaps there is a tendency (though by no means universal) for adherents of a given set of beliefs, values and commitments, to perceive them as not having been freely chosen, not being "mere preferences", and not being discardable through an act of free will. Indeed, Sandel further complains that a liberal approach to religion (which links religion with the exercise of free choice) "may miss the role that religion plays in the lives of those for whom the observance of religious duties is a constitutive end, essential to their good and indispensable to their identity".73 If any group's self-awareness is positively built upon a sense of originating from something other than free choice, then any attempt by the law to classify some characteristics as "immutable" and others as "alterable" is non-neutral towards those groups which are characterized as defined by immutable characteristics and those which are not. If, as the case of religion shows, there is a link in people's minds between recognition of the strength of commitment and recognition of commitment as unrelated to a choice, then a legal characterization of some properties as immutable is parasitic upon a judgment about the depth, strength, sincerity and significance of the given belief or commitment. As a result, not much is really left to "immutability" as such - it merely becomes a disguised vehicle for discriminating between those beliefs and commitments which the government respects more, and those which it respects less and thus accords lesser legal protection. But if this is the case, then it should prompt us to rethink the distinction between "mere" offensiveness and offensiveness that directly implicates one's own identity. This distinction becomes extremely tenuous unless we are prepared to say openly that it hinges upon our views regarding which group identities are worth protecting and which are not. If we are prepared to make 72. Michael 1. Sandel, "Freedom of Conscience or Freedom of Choice?", in Articles of Faith, Articles of Peace: The Religion Liberty Clauses and the American Public Philosophy, ed. by J. D. Hunter & O. Guinness (Washington, D.C.: Brookings Institution, 1990), p. 89. 73. Id.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

2lS

such value judgments about the worthiness of the sensibilities that deserve protection, then we must face the unfortunate consequences of unrestrained majoritarianism. For example, it may occur that we would disallow Muslim claims for special protection against blasphemy because we value their religion less than other religions, and because we think that their religious claims, even if genuine, are somehow less worthy in our society. It is very doubtful if any liberal legislator or liberal legal theorist would be happy to accept this consequence. The only alternative to drawing the distinction in this way is to abandon the distinction altogether, and to lower the level of legal protection for sensibility harm across the board. This is a more candid and honest solution. The theory behind it may be summarized as follows: The gravity of sensibility harm is in the eyes of beholders. Law does not and should not draw the line between those sensibility harms which are worthy of protection and those less deserving of legal concern. The degree of one's personal identification with his or her own wider group cannot be ascertained by the law: some people may have a stronger identification with their church than with their nation, while the ranking may be converse for others. The law should not pass judgment on the quality of this identification. There is, therefore, no such thing as an agent-neutral and rigid distinction between mere offence and offence that implicates a person directly when his or her group is insulted. However, the price to be paid for this legal abstention from judgments about offensiveness is that, in order to win legal protection, the claims for prohibition of group vilification will be viewed with extreme and uniform suspicion. This is the trade-off: all complaints about group vilification are considered genuine (that is, vilification is seen to attack victims' identities) but since the danger of a majoritarian bias in assessing the identityimplicating harm is very high, and a discourse about groups is close to a public discussion about political matters, a conclusion about legal protection against psychic harm has to be insulated from an ordinary calculus of harm. This is not to say that racial vilification laws will necessarily fail the test, but that the test of harm must be much more stringent than the one applied, for example, to defamation laws. The stringency lies, not merely in requiring a higher-than-usual demonstrated seriousness of harm, but also in requiring proof that this kind of prohibition is the only way to avoid the harm. In this context, a state education policy and the strict enforcement of anti-discrimination laws immediately come to mind. If we consider the availability of such alternative ways of reducing emotional harm, and also that harm is unlikely to follow from every speech-act which, under racial vilification laws, would qualify for restriction, the harm of psychic and emotional injury is unlikely to overcome the hurdles of a strict scrutiny of racial vilification laws - a test which demands both a very high gravity of harm and the non-availability of other, less speechrestricting alternatives.

116

CHAPTER 6

It should be emphasized that the argument just presented is different from the standard liberal argument which attempts to show that group vilification is somehow less severe, less serious, or more diluted, than an individual insult or defamation. It has also been said that, in the case of group defamation, the nexus between the defamation and the complainant is less certain than ~ the case of libel of the individua1.74 In traditional legal argument it is established that "to be actionable the defamatory words must be understood to be published of and concerning the plaintiff",75 and this condition is easier to meet in individual rather than group defamation cases. And yet, regardless of the evidentiary problem of demonstrating the nexus, it is doubtful whether group vilification is generally less severe than individual insult. Indeed. one may claim that group libels are, by their very nature, even more defamatory than individual ones because they "sweep with a broader brush" and therefore "they can be that much harder to avoid".76 The aggregate harm is said to be higher because a greater number of people are defamed,77 but also the harm to an individual victim may be more severe if the victim's group is attacked, rather than the victim personally. To be sure, it may well be that some racial slurs against my race will be less upsetting to me than insults attacking me personally, but one can well conceive the reverse situation: I may be hurt more by an insult against me qua a member of my group than qua an individual. There is nothing unreasonable in the belief that, as Justice Frankfurter said in Beauharnais, "a man's job and his educational opportunities and the dignity accorded him may depend as much on the reputation of the racial and religious group to which he willy-nilly belongs, as on his own merits".78 These words have an obvious ring of truth, especially if one considers the specific subject-matter of the Beauharnais case, namely the punishment of Joseph Beauhamais for the distribution of particularly odious racist leaflets addressed against African-American residents of Chicago. The title of the leaflet said it all: "Preserve and Protect White Neighborhoods From the Constant and Continuous Invasion, Harassment and Encroachment by the Negroes". From the point of view of the severity of injury, Beauhamais' leaflet must have been at least as hurtful to an African-American reader as insulting words shouted directly at him or her. And yet the difference lies in the rationale for establishing the standard of scrutiny oflegal restriction: Beauharnais' appeal to the authorities of Chicago to segregate African-Americans, odious though it

74.

See Note, "Group Defamation: Five Guiding Principles", Texas Law Review 64 (1985): 591624, p. 596 n.26. 75. Knupffer v. London Express Newspaper, Ltd, 1944 A.C. 116, 121 (H.L.), emphasis added. 76. Hadley Arkes, "Civility and the Restriction of Speech: Rediscovering the Defamation of Groups", Supreme Court Review (1974) 281-335, p. 292. 77. Natan Lerner, "Group Libel Revisited",lsrael Yearbook on Human Rights 17 (1987): 18496, p. 189. 78. Beauharnais v. IllinOis, 343 U.S. 250, 263 (1952).

RACIAL VILIFICATION AND FREEDOM OF SPEECH

217

is, is his contribution to the public debate about racial relations in his country. Insulting, fighting words addressed to an individual African-American are not. This is a good reason for suspicion of group libel or group defamation laws. In many legal systems, including that of the United States, it has long been established that criminal libel has a scope which is limited to defamation of individuals, not groups. To be sure, there are also other features which distinguish group vilification from defamation in the strict sense. For example, defamation typically applies to false statements of fact which disparage the reputation of another person, while group vilification does not usually lend itself as easily to assessments of ascertainable truth or falsity. However, the weight of this distinction is sometimes overstated, both by the opponents and the proponents of restricting group vilification. 79 For one thing, racial slurs are often a mixture of abuse and statements of alleged facts which can be considered in terms of truth and falsity. For another, even with regard to "traditional" defamation, there is considerable disagreement over the boundary between fact and opinion. so Important difficulties may arise with respect to statements, such as parodies, which on the surface have the form of statements of fact but which "cannot be reasonably interpreted as stating actual facts". As the United States Supreme Court declared in the landmark case of Hustler v. Falwell, an advertising parody which portrayed the leader of Moral Majority as having engaged in a drunken incestuous rendezvous with his mother, could not reasonably have been interpreted as stating actual facts about him. S1 However, the crucial difference is related to the role of a given expression in public discourse about matters of common concern. Individual defamation or "fighting words" addressed to a particular person have very little, if any, claim to a political role, as a contribution to the debate about public issues. But the situation is different with regard to statements, however odious, about large groups: about their characteristics, role in the society, and relationships with other groups.

4

Liberalism and Prohibitions of Bate Speech

"Even though members of the self-appointed liberal elite would never dream of stooping to racist speech, neither, it seems, would they ever dream of taking legal steps to stop it".S2 These harsh words by an English legal scholar, no radical himself, indicate the severity of the dilemma for liberals that is raised by the proposed and actual laws that restrict racist speech. This is perhaps the most 79. See David AJ. Richards, Toleration and the Constitution (New York: Oxford University Press, 1986), pp. 190-191; Delgado, pp. 157-59. 80. See e.g. W. Page Keeton, "Defamation and Freedom of the Press", Texas Law Review 54 (1976) 1221-59, p. 1249. 81. Hustler Magazine v. Falwell, 485 U.S. 46, 57 (1988). 82. Simon Lee, The Cost ofFree Speech (London: Faber & Faber, 1990), p. 42.

218

CHAPTER 6

serious challenge to liberal orthodoxy which has prevailed in the West with respect to the philosophy of freedom of expression. This liberal orthodoxy has been based on a radical severance of the moral assessment of the worth of a conduct from the determination of one's freedom to engage in it. To use the distinction favoured by John Rawls and his followers, the former is guided by an argument of "the good", the latter by considerations of ''the right". The cherished regime of rights precludes the operation of moral judgments of the good in the area of the right; in this latter. area, the state must be neutral between competing conceptions of the good. Hence, in an apt phrase of Sandel, "[l]iberals often take pride in defending what they oppose pornography, for example, or unpopular views".83 But there are limits to one's capacity to genuinely "take pride in defending what one opposes", and beyond these limits, a moral dilemma flowing from supporting a right to do wrong reemerges. One can persist in defending individuals' rights to reprehensible conduct up to a point, but as the gravity of the wrongdoing grows, the intellectual and psychological resources necessary to protect a right to this conduct may be exhausted. In other words, the conflict between "the good" and "the right" is reasonably manageable as long as the violation of ''the good" is not truly dramatic in proportions. That is why liberals have more problems in defending a right to racist speech than, say, a right to flag burning: the evil of the latter seems to be more symbolic than substantive in terms of deeply hurting identifiable individuals. But even with regard to these rather symbolic provocations and offences, patience towards the liberal argument of the type "I hate what you are doing but I respect your right to do it" does not seem to be unlimited. In her article, Robin West expresses such an impatience with the attitude of many United States liberal lawyers to controversy about flag burning cases and the proposed constitutional amendment. In testimonies before the Congress and in various publications, prominent liberals "made clear that their defense of the rights of flag burners did not rest on agreement with the ideas or the individual expressing the ideas. If anything, liberals went to excessive lengths to dissociate themselves from the content of the flag burners' protest".84 This pattern of "defending what one opposes" is consistent with traditional liberal orthodoxy and yet, according to West, it is ultimately formalistic and possibly counterproductive. It defends rights for rights' sake (rather than for the sake of the values which they promote), and it denigrates the social dissent expressed in such actions as flag burning. An alternative path to defend the right of flag

83. Michael J. Sandel, "Introduction", in Liberalism and Its Critics, ed. by M. Sandel (Oxford: Basil Blackwell, 1984), p. 1. 84. Robin West, ''The Supreme Court, 1989 Term - Foreword: Taking Freedom Seriously", Harvard Law Review 104 (1990) 43, 94-95, footnote omitted, emphasis in original.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

219

burners, West argues, is by appealing to their functions in a socially valuable process of communication and self criticism: The measure of our liberality becomes not the extent to which we can tolerate the offensive, hateful, or simply unpopular ideas of others, but the degree to which we individually take responsibility for the truth of our utterances and collectively value and nurture the communicative and truth-promoting realms of political culture and political dissent. 85

The path of discerning positive values in actions which initially seem merely offensive and provocative, may be available to us with regard to flag burning but hardly with regard to racial vilification. And yet West's general observation is instructive: it may simply be impossible to defend a right to a practice in which we can discern absolutely no redeeming positive features. That is why, at the end of this chapter, it may perhaps be useful to reflect upon whether the practice of allowing group vilification to remain unpunished has some benefits of its own (even if group vilification itself is undoubtedly odious and harmful, as it no doubt is). This has not been the general argument of this chapter. Rather, the argument was based on a conviction that the speech-harm link, in the case of racially vilifying speech, has to be subjected to a particularly exacting scrutiny before a conclusion about legal restriction can be drawn because racially vilifying speech can be part of a public debate on political issues. Further, it was suggested that laws prohibiting racial vilification do not easily pass this test in any of the three main areas of harm: harm reSUlting from violence by a target of the speech; harm resulting from violence by other people against the target; and harm inherent in the offensiveness of the speech itself. But the assumption underlying the argument was that group-vilifying speech is harmful, and that the harm it inflicts is such that it is proper for the state to treat it as a morally relevant reason in support of the proposed legal restrictions of liberty. It is on the judgment about how this harm fares in the test of a link between speechrestriction and harm-avoidance, that the burden of one's conclusion against or in favour of anti-group vilification laws should rest. But now that the principal argument has been submitted, it may be useful to lift an underlying assumption and, in accordance with West's remarks, consider whether there may be any good in allowing racial vilification to go unpunished. That is to say, is there any social good to be derived from allowing racist hate speech, other than the underlying value of providing strong protection of free speech that is related to self-government in a democratic state? First, one good of the legality of such speech is that it might prevent and combat complacency towards the existence of racism and racist attitudes. Racists are there, and it is better to be reminded about them by their own speech 85. Id. at 96, footnote omitted.

220

CHAPTER 6

than to allow an illusion to develop that the problem has been solved simply because racist statements have been made illegal. Group vilification is a symptom, not a source, of deeper problems that give birth to hate, contempt and prejudice shared by some people towards other groups in society. By prohibiting public statements that vilify those groups we may reduce the hurt to the feelings of their members but at the same time we risk removing the issue of racism from the public agenda. The good of allowing group vilification is that it helps maintain the visibility of a dramatic problem which is there anyway, regardless of the prohibition. Second, the legality of group vilification may, ironically perhaps, be valuable to the subordinated groups themselves. In a recent article, Kenneth Karst points out that unrestricted freedom of expression is a mixed blessing for the members of subordinated groups: on the one hand they are victims of speech by members of the powerful groups in their society, but on the other hand ''precisely because an important part of a group's subordination consists in silencing, their emancipation requires a generously defined freedom of expression, a freedom that overflows the shallow capacity of the model of civic deliberation".86 It is hard to say in abstract terms whether the benefit resulting from broad freedom of expression, or the benefit produced by the restriction, prevails in the calculus of costs and benefits for the subordinated and powerless groups. And yet it seems that a broad regime of freedom of expression which allows speech considered insulting, offensive and shocking to remain unpunished, may at times be useful to those groups who are alienated from the cultural and political mainstream. Conversely, it is likely that racial vilification laws may be invoked (even despite the best intentions of their drafters) to silence anti-establishment speech by disadvantaged groups who might have to resort to shocking and "uncivil" expressions to overcome the systemic bias of official channels of communications, and to get their message across to the community at large. After all, "civility" is defined by a dominant culture which has all the conventional means of mass communication at its disposal. Laws against profanity and insults are likely to be disproportionately enforced against blacks and other minorities. This is reminiscent of a memorable warning by Justice Black when he dissented from the Beauharnais decision which upheld an anti-hate-speech statute: "If there be minority groups who hail this holding as their victory, they might consider the possible relevancy of this ancient remark: 'Another such victory and I am undone'''.87 In the same decision, a similar warning was issued by another dissenting judge, Justice Douglas: "Today a white man stands convicted for protesting in unseemly language against our

86. Kenneth L. Karst, "Boundaries and Reasons: Freedom of Expression and the Subordination of Groups", University ofIllinois Law Review (1990): 95-149, p. 109. 87. Beauharnais v.Illinois, 343 U.S. 250, 275 (1952) (Black J, dissenting).

RACIAL VILIFICATION AND FREEDOM OF SPEECH

221

decisions invalidating restrictive covenants. Tomorrow a Negro will be haled before a court for denouncing lynch law in heated terms".88 Third, legal tolerance of group vilification has an expressive value contained in the message it conveys to people about the nature of their society. To describe it in deliberately extreme and exaggerated terms, the message is as follows: in a liberal society, claims to protection against insult and offence viewed with the utmost suspicion; racial vilification approaches the outer limits of the rationale for tolerance but this is precisely why the educational effect of this tolerance may be so powerful; liberal society is opposed to a communitarian vision of legally protecting an individual's sense of identity with a wider group; under that vision, an attack on the group is viewed as an attack on an individual. In contrast, in a liberal society, people are encouraged to distance themselves from their collective identities, to treat them as social roles rather than as ingredients of their own selves, and to put up with many dignitary injuries that other societies would have treated much more seriously. To use a traditional distinction, a liberal society is much more of a Gesellschaft rather than a Gemeinschaft, or in a useful term coined by Meir Dan-Cohen, "a union of detached roles".89 As contrasted to the ideal of "community", in such a union people are never encouraged to fully identify themselves with the roles they occupy, and their identity is never fully defined by their membership of groups. People are urged to be able to stand back from the roles they are playing, to adopt a critical attitude to them, and to perceive their own identity as transcending the sum of roles stemming from their involvement in various groups. This is one practical sense of'the controversy about "thick" (that is, context-bound) versus "thin" (or "unencumbered") conceptions of the self in a liberal society. The controversy has been sparked by "communitarian" criticisms of liberalism and by positing, in contrast to liberal individualism, a "thick" conception of the self, or the "situated self', where one's aims in life, social achievements and interpersonal ties are built into the very concept of self. Common to those communitarian critiques was the claim that we "cannot conceive our personhood without reference to our role as citizens, and as participants in .a common life" .90 In opposition to a liberal "thin" concept of the self, defined as it is by common rationality and a capacity for autonomous choice, communitarians claim that certain roles which implicate us within a society are constitutive of our very identities, that "we are partly defined by the communities we inhabit",91 and that the idea (which they attribute to liberals) of

are

88. Id. at 286 (Douglas J., dissenting). Meir Dan-Cohen, "Law, Community, and Communication", Duke Law Journal (1989): 1654-76, p. 1669. 90. Sandel, "Introduction", p. 5. 91. Id. p. 6.

89.

222

CHAPTER 6

an ''unencumbered self', prior to, and independent from, the competing purposes and ends in life, fails to appreciate the degree to which our identities are constituted by the communities in which we live. What meaning could be given to a metaphysical proposition that the self is constituted (partly) by the community? How can we meaningfully represent the contrast between two opposing metaphysical concepts of the individual identity: one which builds community attachments, ends and relations into the very notion of a human self, and the other which draws the boundary between the self (understood in a "thin" way) and the social universe? The choice of one or the other conception is determined, I suggest, by the purpose of the theory; neither is advanced, in theories of justice, on purely metaphysical grounds. To say that I am defmed by a number of my social roles (as a member ofa family, of a church, of a nation) may sound obvious, but then it is equally obvious that (at least in some circumstances) I may cease, through my conscious decision, to occupy these roles: we can divorce our spouses, convert to other faiths, emigrate from our countries. My identity is at least equally defmed by my current attachments as by my capacity to abandon them. If I do revise or abandon these commitments, has my identity changed? To say that I am now a different person, in the sense that I am someone else, is odd and redundant in this context. 92 It is not as if I ceased to exist and then reemerged as someone different. If I am "defined" by a particular social role, but at the same time I am free to revise and ultimately abandon this role, the independent weight of this role in constituting my identity seems to be called into question. To represent my current and past attributes and roles as ingredients of my identity, seems to be a slightly pompous and quite unnecessary way of describing those features of my social characteristics which are considered significant. What is the purpose of introducing the notion of social ingredients of a person's identity? Consider these charges by Sandel against the "deontological ethic" which postulates that we view ourselves as independent selves, independent in the sense that "our identity is never tied to our aims and attachments": we cannot regard ourselves as independent in this way without great cost to those loyalties and convictions whose moral force consists partly in the fact that living by them is inseparable from understanding ourselves as the particular persons we are ... Allegiances such as these . . . allow that to some lowe more than justice requires or even permits, not by reason of agreements I have made but instead in virtue

92. See, similarly, Joel Feinberg, The Moral Limits of the Criminal Law. vol. 3: Harm to Self (New York: Oxford University Press, 1986), pp. 81-87.

RACIAL VILIFICATION AND FREEDOM OF SPEECH

223

of those more or less enduring attachments and commitments which taken together partly define the person I am. 93

This passage reveals the grounds for a communitarian insistence on the social ingredients of individual identity. This is talk of commitments, loyalties and obligations which go beyond the ones based on characteristically liberal grounds, that is, the ones based on fairness, reciprocity or consent. Whether one agrees with Sandel on the merits of his propositions or not, the very beginning of the passage indicates that the social identity of individuals is parasitic upon normative views about the stringency of individual commitments and loyalties to their communities: "[W]e cannot regard ourselves as independent in this way without great cost to those loyalties and convictions" - this suggests that the moral price for adopting an unsituated concept of the self is unacceptable to communitarians. The duties we owe to our families, communities or nations, come first; the situated self is merely an expository device for conveying this message. It is therefore just another way of expressing the idea that our commitments go far beyond those voluntarily made by ourselves as free and independent agents. And, in turn, the individualistic concept of the unsituated self is instrumental to the political philosophy which celebrates the supreme value of free choice and the relative independence of individuals from communal ties. The experience of the modem, pluralistic societies celebrated by liberals, expresses constant revision of communal ties, and frequent transfers from one community to the other. Further, we belong to a plurality of communities at the same time. As a result, liberals conceive the person as defined not by communal attachments but by autonomous choice. But in reality, of course, we are both free agents and situated in our communities; we choose our attachments and they are immersed in our history and traditions; we are implicated in social practices and we often challenge them - at times we even reject them entirely. The predicament of life in modem societies is a mix of commitments and of the revisability of these commitments. Sometimes we face the dilemmas of choice between loyalty to the community and the requirements of our independent critical morality which may lead to our questioning the value of this community. The language of "situated selves" suggests the presumptive priority of loyalty over the questioning; the language of a free, unencumbered self emphasizes the importance of independence and criticism. This is a choice between two moralities, not between two ontologies of the self. And if the notion of the "self' is but a proxy for a moral proposition about the stringency of individual commitments and loyalties, it cannot lend support to these propositions.

93. Michael J. Sandel. Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982). p. 179.

224

CHAPTER 6

Legal toleration of racial vilification expresses the priority of an individualistic morality of an "unsituated self' over a communitarian morality of encumbered selves. If, as a rule, we acted in accordance with the encumbered-self morality, it would not be easy to have enough self-control to refrain from violent responses to insulting words - these words would be seen as destroying central aspects of one's self. But it can be more easily done if these insults are not allowed to reach your own identity: when they attack only some of your social roles, which are seen as extrinsic to your real self. If the language of autonomous choice of one's roles and of the revisability of commitments prevails over the language of situated self, you may easily "tum on [your] heels and leave the provocation behind".94 A determination of the degree to which one's identity is constituted by one's community involvement is, from this perspective, just another manner of speaking about how to react to unpleasant or even odious utterances by others. A heterogeneous liberal society must lift a number of traditional protections of one's psychic well-being so as to maintain its pluralistic and cosmopolitan character. In a sense, a liberal society rejects the principle of honour as a good which one may protect through law; the idea of honour related to the community-defined individual is replaced by the central conception of an autonomous individual who may freely shape his or her social identity and attachments. 95 Individual dignity is perceived more in the power of the autonomous shaping of one's own social world, rather than in the existence of a protected sphere of communal attachments which mould an individual self. This liberal insensitivity to many psychic harms is the price of a broadened scope for individual autonomy.

94. Joel Feinberg, The MoraL Limits of the Criminal Law. vol. 2: Offense to Others (New York: Oxford University Press, 1985), p. 91. 95. For a contrast of autonomy and reputation-as-dignity, see Robert C. Post, "The Social Foundations of Defamation Law: Reputation and the Constitution", California Law Review 74 (1986): 691-742, pp. 731-9.

INDEX Aborigines, Australian, 116 Abortion, protests against, 85, 94 Abrams v. United States, 8n, 39n Academic freedom, 185, 187, 190 Access to means of communication, 73-74, 75,81,82,84-85,86,101 Ackennan, Bruce A., 27n Advocacy, 19-20,70,104,110-111,193, 202-205; See also Incitement Afftrmative action, 3, 99-100, 103, 111 Afiican Charter on Human and People's Rights, 179 Alexander, Larry, 41n, 43n Altman, Andrew, 121 Amar, Akhil Reed, lOOn, Illn Amdur, Robert, 202 American Booksellers Association v. Hudnut, 9n, 31n, 60n, 103-104, 1800 American Convention on Human Rights, 179 Anonymity, protection of, 172 Anti-Defamation League ofB 'Nai B 'Rith " FCC, 180n Anti-Discrimination (Racial Vilification) Amendment Act, NSW (1989), 191n Anti-discrimination law, 105-106, 119-120, 191; See also AffJJlllative action Anti-noise law, 80, 90, 91,96 Anti-Semitism, 32, 107 Arkes, Hadley, 216n Associated Press v. United States, 82n Atkin, Lord James Richard, 194 Audience, interests of, 31, 83, 86-90, See also Captive audience, Freedom not to listen, Hostile audience Austin, J.L., 120-122, 126, 127-128, 130, 131 Australian Capital Television v. Commonwealth, 25-26 Authority, 124-126; two concepts of, 131132, see Illocutionary acts, authority in Autonomy, personal, 16-20,71,75,86, 109,221 Baker, C.Edwin, 109-110 Balkin, J.M., 22, 27n Barak, A., 132n

Barendt, Eric, 81 n, 84n Barnum, David G., 88n, 113n Barron, Jerome, 81, 82 Bates v. State Bar ofArizona, 600, 62 Beauharnais v. Rlinois, 180,216,220 Bell, John, 84n BeVier, Lilian R., 21n Black, Charles, 86 Black, Hugo L., 87, 220 Blackmail,163 Blackmun, Harry A., 60,67,137,149 BoeretrJjn, I., 1790 Bollinger, Lee, 8-9,15,22,31-33, 82n Boos v. Barry, 88 Bork, Robert H., 21n Bose Corp. v. Consumers Union of United States,I7n Boswell-Odum, Beth C., 114n, 184n Brandenburg v. Ohio, 193n, 203n Breach of the peace, 112, See also Fighting words, Violent response .Breathing space theory, 11, 88 Brennan, William J., Jr., 54, 93, 159-160, 163-164,171, 173 Broadcasting Bill and Act, France (1986), 84 Broadcasting,73,83-85,90n,96 Burdens ofjudgment, 79 Buttonholer problem, 87, 97 Byme, J. Peter, 184n, 188,211 Cable TV, 84-85 Campbell, Tom D., 2n, 14,25,26,78, 81n, 118n,121n Canada, freedom of speech in, 39, 73, 99, 107,182,191,192 Captive audience, 92-95,113, 183, 186, 197, 198 Chaplins1cy v. New Hampshire, 41, 69-70, 96, 112n, 198n Chapman J.W., 24n Child pornography, 41, 42,60,70-71,136 Chilling effect, II, 113 Civil Rights Act, U.S. (1964), 183 Civility, 49-51,184,220 Clark v. Community for Creative NonViolence, 164n

226 Clear and present danger, 104, 203-204 Coercive speech, 93-95 Cohen v. California, 49-50, 161, 162n, 208n Cohen, Carl, 199n Coliver, Sandra, 82n, 179n Collective-action problem, 9, 21, 24; See also Prisoner's Dilemma Collin v. Smith, 198n Collini, Stefan, 14n, 50n Commercial speech, 9, 70-72,173; See also Gambling, Tobacco advertising Conseil constitutionnel in France, 84 Consolidated Edison Co. v. Public Servo Comm'n., 39n, 138n, 157n, 171-172, 174-175,197n Constitution making, 29 Content neutrality, see Subject-matter neutrality Content-based restrictions, see Subjectmatter neutrality Cornelius v. NAACP Legal Defense & Ed. Fund, 147n, 148n, 149, 150n Cox v. Louisiana, 44n Cox, Archibald, 37, 41 Creation science, 187 Criminal Code, Canada (1970), 191 n, 192 Cronauer, Adrian, 75-76 Cross burning, 115, 157 Cunneen, Chris, 116n

Dambrot v. Central Michigan University, 181n Dan-Cohen, Meir, 221 Day, J.P., 87n Defamation, 9, 65, 70, 181, 189, 192-195, of a group, 194, 219, 220, 221, of politicians, 12, of public figures, 17 Delgado, Richard, 181 n, 217n Democracy, 20-31, 86, 175 Demonstration, political, 48 Discrimination, 3-4, 105-106, 117-118, 119-133, 180, private, 123; See also Anti-discrimination law Doe v. University ofMichigan, 181 n, 185n Door-to-door solicitation, 90, 91, 96-97 Douglas, William 0., 44-45, 203, 220 Downs, Donald A., 181 n Draft cards, burning of, 52-55 Duty to listen, 87-89 Dworkin, Andrea, 104, 110, 119n Dworkin, Gerald, 159n, 177n

INDEX Dworkin, Ronald, 7-8, 118n, 120n Easterbrook, Frank, 9, 103-104, 180 Elster, Jon, 157n Ely, John Hart, 27, 55n, 118n Emerson, Thomas I., 44, 86 Emotional distress, intentional infliction of, 181 Employer's speech, 93, See also Workplace, speech in Epithets, see Fighting words, Racial vilification Epperson v. Arluznsas, 135n Equality of opportunity, 73-81, 99, applied to speech, 74-81, 83, 85, 98-99, 103, 119, meaning of, 76, 77-78, to be heard, 79-80, 86, 98, to convince, 81 Equality, and speech, 3, 73-118, 181, conflict of liberty and, 73-75, principle of, 73 Erznoznik v. City ofJacksonville, 161 Establishment Clause; see Religion, establishment of European Convention on Human Rights, 179 Expressive opportunities, equality of; see Equality of opportunity Extremist speech, 32, Fact/opinion distinction, 9-10,70 Fairness doctrine, 75, 81, 82-83, False statements, protection of, 10-12, FCC v. National Citizens Committee for Broadcasting, 84n FCC v. Pacifica Foundation, 38n, 54n, 9On, 96n,207n Feinberg, Joel, 33, 94n, 158n, 159n, 196, 222n,224n Fighting words, 3, 32, 41, 62-69, 101, 109, 111-118,180,197,198; asymmetrical regulation of, 111-118; concept of, 6769, Il2, First Amendment, 28-31, 37,41, 43, 45, 46, 56,65,66-67,73,81,92,137,139,160, 172, 174,180, 181, 185; conflicts between various clauses of, 145 First National Bank ofBoston v. Bellotti, 175n Fish, Stanley, 7, 8 Fisher, W.A., 132n Flag buming, 47, 48-49,52-53,55,218, 219

INDEX Fourteenth Amendment, 54, 73 Frankfurter, Felix, 216 Fraser, Andrew, 26 Free Exercise Clause; see Religion, free exercise of Free Speech Clause, 147, 148, 151, See also First Amendment Freedom not to listen, 86-98 Freedom of religion, see Religion, free exercise of Freedom of speech; justifications of, I, 2, 735; value of, 7-8 Freedom of the Press Act, Sweden (1949), 179 Fried, Charles, 140n Frisby v. Schultz, 92n, 94n, 98n Gag rules, 157-158, 176 Gale, Mary Ellen, 74-75, 76 Gambling at casinos, ban of advertising for, 61,62 Gard, Stephen W., 62n, 113 Garrison v. Louisiana, 180n Gender, fighting words based on, 62-63, 66 Gertz v. Robert Welch, Inc., 13n Gillette v. United States, 135n Goffman, Erving, 206-207 Greenawalt, K., 13n,66, 114 Greer v. Spack, 98n, 138n, 147n, 150n, 175 Gregory v. Chicago, 87n Grey, Thomas C., 105n, 117 Gross, Hyman, 199n Grossbaum v. Inianapolis-Marion County Building Authority, 146, 166 Guinness, 0., 214n Gunther, Gerald, 145n Haggerty v. Associated Farmers of California, Inc., 80n Haiman, Franklyn S., 89 Harassment, 85, 94,107,181,1"83, See also Nuisance phone calls Hardin, Russell, 24n Harlan, John M., 49 Harm calculus, 3, 28, 32, 38, 43, 47, 6 9, 88,200,209 Harm Principle, 18-19,34,38,42,135-136, 201 Harm, moral, 33,159 Harm, psychic, see Psychic injuryHarmful speech, 37-38,40,42-43,60-62,66-72,

227

104; See also Persuasion-related harm, Speech-harm connection Harris, Bede, 139 Harrison, Geoffrey, 16n Hart, H.L.A., 208-209 Hate crimes, 106 Hate speech, 39, 98-111, 115, 119-123, 129, 132, 133, 179-187; harms of, 183184,195-217, in universities, 180-187, See also Racial vilification Heckler's veto, 88, 113, 161, 197-198 Heffron v. International Soc y for Krishna Consc.,89n Hobbes, Thomas, 16 Holmes, Oliver Wendell, 8, 39, 203 Holmes, Stephen, 157-158, 176 Holocaust denial, 20, 187 Hook, Sidney, 135 Hostile audience, 85, 88,160-161,198 Hunter, J.D., 214n Hunter, Nan D., liOn Hustler Magazine v. Falwell, lOn, II, 31n, 217 Hynes v. Oradell, 91 n

1Il0cutionary acts, 4, 56, 119-133, authority in, 123-133 Incitement, 50, 56,110-111,136,180,193, 195, to racial hatred, 19,179,187,192, 195,200-205 Infallibility, 13-14 Institutional facts, 127-129 Insults, 18, II 7, 189-190, See also Offensive speech International Convention on the Elimination of All Forms of Racial Discrimination, 180, 188, 192 Intemational Covenant on Civil and Political Rights, 179, 192 International Soc y for Krishna Consciousness v. Lee, 98n Intimidation, 90-91, 101, 112, See also Harassment Intolerance, 172, 173-177, racial, 64,197198, towards an idea expressed, 67, 156-157, 161-162, 169-170, 174-177. 187 Jarman v. Williams, 46 Jehovah's Witnesses. 90 Jones v. Wolf, 135n Jordan v. Burgoyne, 113n

228

INDEX

Judicial review, 26, 29, 38 Kagan, Elena, 106n Kalven, Harry, 21 Karst, Kenneth L., 41n, 73, 220 Kaufinan, Nathan J., 189 Keeton, W. Page, 2170 Kennedy, Anthony M., 139, 140, 141-146, 153,155,163-165,168,170 Knupjfer v. London Express, 194n, 216n Kovacs v. Cooper, 89n, 98n Krausz. M., 15n

Lamb 's Chapel v. Centre Moriches Union Free School District, 138n, 148-156 Langton,RBe, 121-122, 123, 124, 125, 129, 130-131,132-133 Lasson, Kenneth, 18On, 181 n Law, Sylvia A., 11 0 Lawrence, Charles R., III, 105n, 115-116, 1830, 184n Leaflets, distribution of, 44, 97, 98, 143144, 172 Leased access, 83 Leclercq, Claude, 84n Ledsinger v. Burmeister, 189n Lee, Simon, 217n Legislative motives for restricting freedom of speech, 7-8, 44, 51-52, 58, 59, 70, 158,161,164-165,171,184 Lehman v. City ofShaker Heights, 93, 138n, 147n, 173, 175 Lemon v. Kurtzman, 176n Lerner, Natan, 216n Libel, 32,41,70 See also Defamation Love, Jean C., 191 Low-value speech, contrasted to high-value speech, 41-43, 58 MacKinnon, Catharine, 73-74, 79, 99,103104,106-108,110, 119n, 120, 121, 130, 133 Madsen v. Women's Health Center,lnc., 85n,90 Mahoney, Kathleen, 192n Majority rule, 27, 29 Manner of expression, 50-51 Market failure, 9-10 Marketplace of ideas, 8, 10, 95, 174, 186, Marshall, Thurgood, 136-138, 167 Martin v. City ofStruthers, 91n Mason, Anthony, 25-26

Massaro, Toni M., 5n, 115n, 182 Matsuda, Mari J., 101, 115-116, 119, 120, 121, 181, 184, 19On, 192n McGowan, Devid R., 191, 212n McIntyre v. Ohio Elections Commission, 172n Media ownership, 84 Meiklejohn, Alexander, 20-21, 143n Meiland, J.W., 15n Mendu~Susan, 16n Mental intermediation, 69, 103-105,108 Meritor Savings Bank v. Vinson, 183n Message, distinguished from conduct, 105107, 119-120 Miami Herald Pub. Co. v. Tornillo, 81 n, 82 Military bases, speech in, 98, 138, 147, 175 Mill, John Stuart, 11, 12, 14,34,50-51,162 Miller v. City ofSouth Bend, 45n Ministry ofInformation v. Crlciret Association ofBengal, 88n Moral relativism, 15-16 Motive-based analysis, 28-30, 44, 51-52, 58-59,138, 147-148, 156, 163-165, 184; See also Legislative motives for restricting freedom of speech Motives, impermissible, 28-30, 33-35, 5860,67,71,109,158

NAACP Legal Defense & Educ. Fund v Devine, 197n New York Times v. Sullivan, 11, 31 n, 73n, 82n, 136n, 137 New York v. Ferber, 70n New Zealand, 139 Niemotko v. Maryland, 114n NLRB v. Gissel Packing Co., 93n Noise, 90, 91, 96, 114, 136; See also Antinoise law Non-public forum, see Public forum Nato v. United States, 193n, 203n Nozick, Robert, 75 Nudity, 45-46 Nuisance phone call~ 90-91, 92, 95, 96

o 'Brien test, 54-55 Obscenity, 32,41,65,99, See also Pornography Offensive speech, 37, 47-48, 66-67, 90, 96, 102-103,114,116-117,120,162,184, 186,188,195,195,206-210 Osborne v. Ohio, 42 Oyediran, 1., 179n

INDEX Paternalism, 32-33, 139, 143, 156-162, 169, 173-177 Performative utterances, 120, See also IIIocutionary acts Perlocutionary acts, 71, 119-122, 124-131 Perry Educ. Ass 'n v. Perry Local Educators Ass'n, 138n, 149n, 150, 170-172, 173 Persuasion principle, 108-109; See also Persuasion-related harm Persuasion-related harm, 69-72, 104 Picketing, 44-45, 90, 92, 94, 136, 137-138, 167-168 Police Department of Chicago v. Mosley, 80n, 136-138,167, 174n Political advertising, bans on, 24-26, 29, 173, in Australia, 24-25, in public transport, 93, 138, 147, 173 Political speech, 21-22, 25, 31, 136, 146147,174-175 Pornography, 73,99-111,119-126,129133; authority of, 123-125, 129, 131, 132,209, viewpoint in, 131, See also Child pornography Posadas de Puerto Rico Assoc. v. Tourism Co., 61, 62 Post, Robert C., 10-11, 13, 98n, 102, 224n Powell, Lewis P., Jr., 13, 174-175 Precommitment, 23 Preference formation, 30 President, assassination of, 52-53, 55, threats against, 66 Prisoner's Dilemma, 23-24, 177 Privacy, 70, 86,90, 91,96 Proselytizing, 97, 146, 151, 154-156 Provocation, see Violent response Psychic injury, 65, 206-224 Psychological injury, see Psychic injury Public discourse, 25, 101-102, 157, 158, 165,169-170,174,183,185,195 Public forum, 97-98,138,148-150,162, 171-172,174-175,185,186, designated, limited-access, traditional, 148-149 Public issues, speech on, 21, 27-28, 30-31, 138 Public Order Act UK (1936), 113n Public order, see Breach of the peace Public transport, advertisements on, 93, 147, 173 Pure opinion argument, 13 Pure speech, 44-45, 55, 58 R. v. Butler, 107

229

R. v. Creamer, 205n R. v. Duffy, 199n R. v. Keegstra, 6n, 39n, 86n, 107 R. v. Most, 193n R. v. Oakes, 39n R.A.V. v. City ofSt. Paul, 58-72, 114n, 115, 137n,157 Race Relations Act, UK (1965), IBn, 179 Race, fighting words based on, see Racial vilification Racial Hatred Act, Australia (1995), 179 Racial Vilification Bill and Act, New South Wales (1989),182,188,189 Racial vilification, 4, 37,41,61-64,73, 101, III, 116-118, 119-120, 123, 179224; prohibitions of, Ill, 119-120, 179180, 182, See also Hate speech Racism, see Intolerance, racial; Racial vilification Racist speech, see Racial vilification Rawls, John, 79,218 Raz, Joseph, 16n, 125 Reasonableness of speech regulations, 148156, 148, 149-150, 162-163 Red Lion Broadcasting Co. v. FCC, 83n Redish, Martin H., 71n, 143-144, 204n Rehnquist, William H., 53, 61, 135 Religion, as a viewpoint, 140-148, 151156,165, 166-167,168,170, meaning of, 56 Religion, establishment of, 57, 135, 140, 142,144-145,151,165-167,176 Religion, fighting words based on, 63-64, 66 Religion, free exercise of, 39n, 57, 145, 169-170,176,192 Reply, right of, 81-82 Richards, David AJ., 21 n, 60n, 209, 217n Riesman, David, 194 Right to be heard, 86-88, 98; See also Equal opportunity to be heard Rockwell v. Morris, 161 Roemer v. Board ofPublic Works, 135n Rosenberger v. Rector & Visitors of the University of Virginia, 139, 140-147, 148,149, lSI, 153, 154, 155, 165, 166, 167,168,169-170,176 Rosenbloom v. Metromedia, Inc., 31 n Roth v. United States, 31 n Rowan v. United States Post Office Dep 't, 91n Rules, constitutive and regulative, 127-128

230

INDEX

Rushdie, Salman, 210 Ryan, Alan, 16n Sadurski, Wojciecb, 2n, 14n, 27n, 57n, 81n, lOOn, IISn, 121n, 135n, 136n, 145n Sandel, Micbael J., 214, 21S, 221n, 222223 Sartorius, Rolf, 159n Scalia, Antonin, 61-66, 67,114, 115 Scanlon, Thomas M., 19-20,34,201 Scbauer, Frederick, 2, 13, 23, 40, 46, 55-56, 57-58,74,127-128,174 Schenck v. United States, 203n Schiro v. Clark, liOn Schlag, Pierre, 33 Scrutiny, levels of, 37-40,42-43,47,48, 58,59,60,70, 137, ; intennediate,42, 59; See also Strict scrutiny of speech restrictions Search for truth, 8-16, 184 Searle, John R., 71 n, 127, 131 Self, concepts of, 220-224 Self-censorship, 11, 157 Self-fulfillment, 16-18, 17-17,20 Self-restraint, 22-24, 32, 197 Sexual orientation, 112, 117, 168-169, 170 Shiffrin, Steven H., 22n, 60n, 62, 66, 67n, 208n Shils, Edward, 50-51 Shur, Luba L., 146n Sidewalk counselling, S5-86, 92-93 Silencing, 9S-111, 119,151,170,186,197, 208,220, as demeaning, 101, 102-103, 108, Ill, 151, as intimidation, 101-102 Skelly Wright, J., ISO Skepticism, 13-16 Skokie, Nazi march in, 37, 56,161, 199 Siagstad, Rune, 157n Sloan, Christine OUe, 90n Smith, Steven D., 95 Smolla, R.A., 192 n Solicitation of money, 159-160, 162-164 Souter, David H., 139, 147, 148, 149, 151156, 165 South Florida Free Beaches. Inc. v. City of Miami,46n Speakers, interests of, 31, 83, 88-90,98,99, 196, Speakers, status of, 170-172 Speech-acttheor-r, 121-122,203

Speech-harm connection, 40, 56, 60, 66, 68-72, 198-199,219; See also Persuasion-related harm Speech-plus, 43-44, 47-48, 52, 55, 58 Speech/conduct distinction, 43-58 Spence v. Washington, 45n Stalking, laws against, 90-91, 92 Stephan, Paul B., ru, 168n Stephen, James Fitzjames, 14 Stereotypes, 117,181-182, sexist, 110 Stevens, John Paul, 37, 64, 65n, 157, 171, 172 Stigma, 117-118,206 Stone, Geoffrey, 34, 41n, 59, 600, 142-143 Strauss, David, 108-109 Strauss, Marcy, 92n, 93n, 96n Street v. New York, 207n Strict scrutiny ofspeeeb restrictions, 28, 35, 38-39,42-43,59,137,172,196,198199 Strossen, Nadine, 156, 157 Subject-matter neutrality, 58-60, 62-63, 7374,76,138,141,142,153,156-170 Subject-matter restrictions, see Subjectmatter neutrality Sullivan, Kathleen M., 97, 99-100 Summary Offences Act, NSW (1988),116 Sunset Amusement Co. v. Board ofPolice Comrs. ofCity ofLos Angeles, 46n Sunstein, Cass R., 21n, 34-35, 42n, 43, 52, 101, 109n, 117n, 137n, 177n Symbolic action, 43-58, 182,218, as speech, 47, 55 Tangri, Ragesh K., 191, 212n Taylor, G. Michael, 86n Teamsters Local 695 v. Vogt, 44 Ten, C.L., 34 Terminiello v. Chicago, 160, 162 Texas v. Johnson, 53, 54n, 208n Thomas v. Review Bd., 39n Thomas, Jeffrey E., 13n Threats, 70; See also President, threats against Time Warner Entertainment Co. v. FCC, 83n Time, place and manner regulations, 139, 163, 164 Tjaden, P.G., 132n Tobacco advertising, 71

INDEX Tolerance, 14-16,31-35,63-64,95,96, 106,139,143,157,184,185,187,221, See also Intolerance Tranquillity, interest in, 96 Tribe, Laurence H., 2n, 22n, 34n, 41n, 44n, 68,97 Truth, as justification for freedom of speech, 8-9,12, as a defence, 189, 194,; See also False statements of fact, Search for truth rationale Tucker, D.F.B., 9n, 21n Turchick v. United States, 28n Turner Broadcasting System, Inc. v. FCC,

83n,85n United States Postal Service v. Council 0/ Greenburgh Civic Ass 'ns., 150n United States v. Carolene Products, 27, 30 United States v. Kokinda, 98n, 159-160,

162, 163n, 164 United States v. O'Brien, 53-55

Universities, speech in, 148, 151,153-154, 170,176, lSI, 184-187, 191,202, See also Hate speech, in universities Unwanted communication, 88-96, 113,186 Unwi11ing listeners; see Captive audience, Freedom not to listen, Hostile audience, Unwanted communication Van Nuys Publishing Co. v. City 0/ Thousand Oaks, 89n

Viewpoint discrimination, indirect, 167-173 Viewpoint neutrality, 3, 4, 34-35, 58-60, 62-72, 135-177; and harm scrutiny, 3435,58-62,67-72, and intolerance, 139, 143, distinguished from subject-matter neutrality, 58-59, 60,139,150-156, 162-167 Viewpoint, compared to subject matter, 5859,64-68, 138, 140-148 Viewpoint-based restrictions, see Viewpoint neutrality Vilification, see Racial vilification Vii/age o/Schaumburg v. Citizens/or a Better Environment, 91n Village o/Skokie v. National Socialist Party o/America, 161n, 198n, 199n

Violent response, 62, 65-66, 113-114, 195, 196-200 Virginia Pharmacy Board v. Virginia Consumer Council, 175n

231

Waldron, Jeremy, 162 Wallace v. JajJree, 135n Wardv. Rock Against RaCism, 147n, 164 Warner, Stuart D., 14n Warren, Earl, 54-55 Weiss, Jonathan, 135n Wertheimer, A., 24n West, Robin, 218-219 White, Byron R., 67, lIS, 150, 171 Williams, Bernard, IS, 120n Wollam v. City o/Palm Springs, 81n Women, subordination of, 119-120 Workplace, speech in, 93, 95, 147, IS3184, 186 Wright, R. George, 22n, 49-50 Yates v. United States, 203n Young v. American Mini Theatres, Inc., 37n Zdrahal v. Wellington City Council, 139 Zorach v. Clauson, 135n