LEXSEE 32 WM AND MARY L. REV. 211

Copyright (c) 1991 College of William & Mary.

William & Mary Law Review

WINTER, 1991

32 Wm and Mary L. Rev. 211

LENGTH: 25846 words

FREE SPEECH AND RELIGIOUS, RACIAL, AND SEXUAL HARASSMENT: EQUALITY AND FREEDOM OF EXPRESSION: THE HATE SPEECH DILEMMA

NAME: TONI M. MASSARO *

BIO:

* Professor of Law, University of Arizona School of Law. B.S., Northwestern University, 1977; J.D., Marshall-Wythe School of Law,College of William and Mary, 1980.

SUMMARY:
... Harry Kalven, Jr., once wrote: "One is tempted to say that it will be a sign that the Negro problem has basically been solved when the Negro begins to worry about group-liberal protection." ... Finally, there are the formidable problems of defining an epithet or slur and containing the adverse consequences of restricting this speech. ... Analogies to obscenity, intentional infliction of emotional distress, physical assault, sexual harassment at work, and the law of defamation support the claim that even traditional free speech principles, coupled with principles of equality as expressed in Brown v. Board of Education, indicate that hate speech regulation is, or should be, constitutional. ... The strong civil rights critique of the liberal objection to hate speech regulation is binary. ... In summary, the paradigm-accepting step to the argument for hate speech regulation emphasizes two things: the harm to victims of hate speech is serious, and the value to society in allowing this speech is slight. ... The analytical problems with this one-way approach to hate speech regulation go beyond the fact that it relies on a contested interpretation of equality. ... The accommodationists' writings reflect an effort to hear both the liberal and the civil rights arguments regarding hate speech regulation. ... Both as a practical and a theoretical matter, an aggressive education/counterspeech remedy to hate speech could threaten liberal free speech values as fundamentally as a disciplinary proposal. ...

HIGHLIGHT: "The plain fact is that not all free speech is good speech. Which means that freedom of speech is not always a sound or just public policy." n1

"It is central to the idea of a liberal society that, in respect to words as opposed to deeds, persuasion as opposed to force, anything goes. . . . A liberal society is one which is content to call 'true' whatever the upshot of [free and open] encounters turns out to be." n2

TEXT:

[*211] I.

Harry Kalven, Jr., once wrote: "One is tempted to say that it will be a sign that the Negro problem has basically been solved when the Negro begins to worry about group-liberal protection." n3 If Kalven was right, then perhaps one ray of hope relieves the bleak evidence of a lack of progress, if not a downright retrenchment, in the struggle to overcome racism. In recent years, several statistical sources have indicated that bias-motivated physical violence and verbal hostilities are on the rise. n4 College campuses in particular have reported an alarming increase in verbal altercations in which minority and women students have been the targets of demeaning, sometimes vicious expression. n5 There reports [*212] prompted attempts to regulate racial and other forms of verbal abuse on campus, n6 which in turn spawned legal action n7 and critical commentary. n8 The focus of the commentary has been the uneasy tension between robust protection of offensive expression and protection of the dignity and physical integrity of potential victims of such expression.

The issue, of course, is hardly new. In the late 1970's, many people participated in similar discussions when Frank Collin, an avowed Nazi sympathizer, proposed to give a speech in the predominantly Jewish community of Skokie, Illinois. n9 During the late 1940's and early 1950's, commentators confronted related issues when group libel statutes were enacted in response to World War II's sobering lessons about the peculiar seductiveness of hate speech. n10 Indeed, these two eras epitomize the two competing national instincts that make the problem of hate speech so wrenchingly difficult.

In 1952, the Supreme Court sided with those who believe that hate speech is subject to state regulation because of the state's interest in preventing violence and in protecting the esteem and dignity of its citizens. n11 By the 1970's, however, the judicial response to offensive speech shifted toward granting this expression broad fist amendment protection, lest the government engage in impermissible content regulation of speech. n12

[*213] The revival of this debate in the 1990's reveals that neither resolution proved entirely satisfactory. In recent legal scholarship, writers have proposed three approaches to hate speech, each with its own internal complexities and variations. The first approach allows hate speech in order to maximize opportunities for individual expression and cultural regeneration. n13 The second, highly controversial approach represses hate speech through sanctions that range from official and private reprimands to criminal prosecutions in order to promote equality and the non-subordination of potential hate speech targets. n14 Aggressive versions of this approach urge that hate speech should be punishable only when directed at members of a historically subordinated group, not dominant group members. n15 The third, emerging approach attempts to accommodate the "worthy passions" of the first two approaches. n16 The accommodationists endorse tightly worded, cautiously progressive measures that tend to proscribe only targeted vilification of a person on the basis of race, gender, religion, ethnic origin, sexual orientation, or other protected characteristics.

For people who, like the accommodationists, resonate to poles of this debate, the issue can seem irresolvable. The issue is difficult even for people who have very strong ties to potential or actual victims of these slurs. One of my students, the daughter of two Nazi concentration camp survivors, is a striking example of this tendency. She has observed firsthand the psychological aftermath of religious persecution and, as she puts it, "the extreme fragility" of this Jewish subculture. n17 Despite her close familiarity with the personal consequences of this persecution, however, she is unsure whether religious hate speech should be denied constitutional protection:

When I read the arguments for criminalizing this speech, or for disallowing the Nazis to speak in Skokie, I am utterly persuaded. Then, when I read the counterarguments that favor allowing this speech, even in Skokie, I think "That's right too." If I had to decide, I think I would say that this population is [*214] unique and should be protected from the psychological harms of reliving the Holocaust. But I am torn, and would be disinclined to go much further. n18

Many people share this woman's empathy with the potential victims of hate speech, yet are uncertain that suppressing harmful speech is the lesser evil. Even people who themselves have been targets of these assaults have expressed hesitation. n19

The competing concerns that make this controversy so difficult include both abstract, constitutional, and philosophical factors and intensely concrete, practical ones. The central constitutional dilemma is that the Bill of Rights protects both individual autonomy and certain collective goals, like equality. This constitutional paradox in turn reflects a more general, philosophical quandary: we strive to reconcile the competing claims of contingency and solidarity n20 and to protect the special role of free discourse in a society, like ours, that is responsive to both claims. Disagreements regarding the constitutional meaning of equality, the proper role of government in inculcating values, and the constitutional and social significance of group identity versus individual personality further compound these theoretical complexities.

Practical problems add to the theoretical difficulties. First, both of the proposed "solutions" to the problem of hate speech -- suppression and protection -- evoke nonfrivolous charges that they will cause serious social harms. They conjure up the dual spectres of McCarthyism on the one hand and spirit-murdering n21 denials of equality on the other. Second, both proposals may trigger forceful opposition. Protecting hate speech, especially in controlled environments like the workplace or school, fosters an atmosphere of incivility and tension, which can give rise to unrest and even physical disruptions. Yet, suppressing hate speech, especially under the "one-way" proposals of some civil rights theorists, risks charges of censorship or reverse discrimination, which likewise can give rise to intergroup hostilities and potential disruptions. Another student expressed this latter concern during a seminar discussion of this problem. As he put it, "There is no [*215] way, in the high school that I attended, that the students would accept a rule that said the blacks could call the whites racist names, but the whites could not call the blacks racist names. The students would laugh in your face, or worse." n22

Finally, there are the formidable problems of defining an epithet or slur and containing the adverse consequences of restricting this speech. Is it racist, for example, to state that African-Americans are better athletes than whites? n23 Is it homophobic to declare that AIDS is a product of reckless gay male sexual practices? n24 People express concern that the rules that restrict hate speech will be overbroad and could stultify intergroup discourse or chill academic discourse, n25 political satire or social commentary, n26 contemporary rap music, or other forms of artistic expression. n27 Many worry that regulation of hate speech will lead [*216] to regulation of other forms of offensive and confrontational speech, like flag burning or other political expression. Some grouse that the attempt to control hate speech may be manipulated to attempt to enforce "politically correct" attitudes, with a strong tilt toward the left. Others, more sympathetic to the regulation proposals, worry that once hate speech regulation identifies certain terms as unlawful epithets and slurs, inventive racists will coin new ones faster than the regulators can master the new vocabularies. Indeed, the effort to regulate may inspire these inventions or send them underground. n28 Many people argue, therefore, that no workable solution to hate speech is possible. Any regulation would be either too chilling of good speech or so narrow as to be purely symbolic and likely unenforceable.

Proponents of hate speech regulation have offered rejoinders to these practical arguments. First, they argue that until we adopt these measures we cannot know whether they will prove effective or have bad side effects. Other countries that have adopted group libel laws -- including Canada and Great Britain -- have not reported a catastrophic erosion of civil liberties or free [*217] speech. n29 Second, the fact that a statute is largely symbolic or educational hardly renders it a useless measure. The symbolic and educational effects of a public law may be good and worthwhile. n30 Third, the reformers observe that many laws are quite difficult to enforce, n31 but this does not mean we should not have enacted them. Finally, hate speech proponents reply that we can define hate speech in a sensible and just manner. Useful, workable terms that are equally fluid and contextual abound in law, n32 such as "reckless behavior," "obscenity," "invasion of privacy," and "intentional infliction of emotional distress."

These philosophical, constitutional, and practical factors illustrate the extreme difficulty of striking an appropriate balance between the strong claims of civil discourse and the strong claims of untrammeled expression. In the following pages, I explore these complexities in greater detail and defend my endorsement of an accommodationist approach to hate speech regulation.

II.

Perhaps a cause, but certainly a reflection, of our ambivalence toward untrammeled speech is the Supreme Court's first amendment jurisprudence. The Court has assigned varying degrees of protection to different types of speech, depending on the purported social value and potential harm of the speech. For example, the Court has concluded that obscenity promotes no worthwhile first amendment end and thus is not protected speech. n33 Commercial speech, n34 defamation, n35 speech in schools n36 and in the [*218] workplace, n37 and speech that inflicts intentional emotional distress n38 all receive less than full first amendment protection.

Yet the Court also has said that "[t]here is no such thing as a false idea," n39 that "[o]ne man's vulgarity is another's lyric," n40 and that speech content is an inappropriate basis for government speech regulation. n41 Most recently, in its controversial flag burning cases, the Court reaffirmed its commitment to the general notion that the offensiveness of expression, by itself, is not grounds for its suppression. n42 Speech that falls outside of the Court's designated categories of "low value" expression thus cannot be censored absent an imminent danger of serious disruption n43 -- a clear and present danger.

Taken together, these holdings reveal an uneven, fitful pattern of protection for speech and an unwillingness to give full weight to the bromide that the best cure for bad speech is always, and only, more speech. The Court has acknowledged that, in some circumstances, with certain types of expression, the state may interrupt the message or even ban it altogether. n44

Most constitutional experts nevertheless interpret the Court's current position to be that racial and other types of epithets and insults are "high value," maximally protected speech, unless they are "fighting words," uttered face-to-face and likely to trigger physical violence. n45 According to some scholars, the justifications for this protection are that we must: defend against a slide down [*219] the slippery slope of constitutional protection into unreasonable repression of "good speech"; n46 remain steadfast in our commitment to government neutrality toward speech content; n47 teach the national value of tolerance through restraint of the strong inclination to silence dissonant voices; n48 secure greater liberty and social stability by permitting this rancorous speech and thereby defusing or exhausting the hostility through speech ventilation, rather than physical violence; n49 and finally, trust the audience to reject bad ideas and embrace good ones.

Thus, if a Nazi sympathizer today wishes to give a public speech extolling the virtues of Hitler, he may do so. Moreover, if the crowd reacts with shouting or threats of physical violence, the appropriate police response is to contain the hecklers, if possible, not to punish the speaker.

Although in 1952 the Court upheld a "group libel" statute that criminalized the speeches of a neo-Nazi or any other speaker who "'exposes the citizens of any race, color, creed or religion to contempt,'" n50 later cases cast serious doubt upon whether that decision remains good law. n51 Consequently, a conservative reading of contemporary constitutional law reveals that hate speech cannot be suppressed unless it satisfies the very narrow conditions of the Court's fighting words doctrine. n52

[*220] This reading renders unconstitutional several of the proposed forms of hate speech regulation, including the University of Michigan's notorious, broadly worded disciplinary rules n53 and other reform efforts that go beyond face-to-face verbal attacks to include group libel as a disciplinable offense. n54 Some people further argue that statutes that specifically name racist, sexist, homophobic, or other forms of harassment as grounds for punishment are also unconstitutional, insofar as these statutes are overbroad, vague, and impermissible forms of viewpoint discrimination. n55

A conservative reading, however, is not the only possible interpretation of the case law. Several writers argue that if we examine the Court's first amendment doctrine as a whole and measure the value of epithets and slurs against the traditional justifications for first amendment protection, then a different picture emerges. n56 A longer, more comprehensive view of the matter shows that the Court already has approved content-specific speech regulation for reasons that resemble those offered in favor of hate speech regulation. Moreover, the case law repeatedly recognizes that, in some settings, privacy and other interests justify speech regulation. Analogies to obscenity, intentional infliction of emotional distress, physical assault, sexual harassment at work, and the law of defamation support the claim that even traditional free speech principles, coupled with principles of equality as expressed in Brown v. Board of Education, n57 indicate that hate speech regulation is, or should be, constitutional. n58

To the extent that the case law does not comfortably support the most radical arguments for criminalizing hate speech, however, [*221] some writers urge a rethinking of the old parameters and an extension beyond the prevailing metaphors and cliches. For example, they describe the "marketplace of ideas" metaphor n59 as misleading and insensitive to the relative market power of different speakers. n60 They critique the counterspeech remedy for bad speech as insufficiently mindful of the disutility and even physical danger of meeting some forms of verbal harassment with verbal repartee. n61 Indeed, the general, traditional free speech justifications appear quite feeble when they are invoked in defense of the specific example of hate speech. Thus, say some scholars, judges should move beyond formalism and contextual theory, lest we "degenerate from an abiding faith in the First Amendment to an obsession with the 'alluring abstractions' of neutral principles." n62

In particular, the advocates of hate speech suppression underscore the concrete psychological and physical harm that hate speech causes its victims. n63 Such speech, they admonish, is a far cry from the type of reasoned, political discourse thought to be core value first amendment expression. n64 Moreover, these private acts of repression -- no less than any official policy of discrimination -- can constitute flagrant denials of the principles of equality and full political participation of all citizens. n65 Interweaving narratives and personal anecdotes into their doctrinal arguments, the civil rights theorists submit that hate speech is speech that government both can, and should, regulate -- far more than the narrow fighting words exception allows. n66

Doctrine therefore yields no clear answer to whether the first amendment protects speech that is as confrontational and potentially destructive of human dignity and social solidarity as is hate speech. The Court's holdings can be construed -- or extended -- to support either suppression or protection of verbal harassment, though not with equal ease. In any event, the dominant interpretation, which suggests this speech is protected, is subject to [*222] substantial criticism, as the critical commentary and the Court's vacillating position on the matter prove.

One thing is clear from the case law and the critical commentary: both the Court and nearly all theorists agree that government can regulate speech when the potential harm of allowing the expression seriously outweighs the potential harm of suppressing it. The disagreements thus spring from different estimations of the degree and nature of the harm in hate speech and the harm in its suppression, which arise from competing visions of the world as it stands and the ideal conditions to which we should aspire. Understanding these underlying assumptions and perfectionist aims is an essential first step toward a sensitive response to the hate speech dilemma.

III.

The approach to hate speech that is most familiar to contemporary Americans is that urged by the civil liberties theorists. Their argument, which urges broad protection of offensive expression, hinges on faith and fear. The faith is in people's capacity to withstand, reject, or fob off insults, or to engage in critical and effective counterexpression. The fear is of people's tendency to define the category of offensive ideas too broadly and idiosyncratically, so as to suppress important criticisms of orthodoxy and to curtail revolutionary possibilities and individual creativity. The analytical premise of the civil liberties approach is that freedom of expression is the baseline condition of democracy. n67

Civil libertarians insist that government cannot censor or hinder speech on the basis of the viewpoint expressed, no matter how offensive. n68 They reject the argument that racial insults and epithets are mere inarticulate grunts undeserving of protection. Slurs are words, and as such have ideational content, however vulgar and confrontational. Government thus cannot cut off this expression unless and until a fight or other serious disruption is imminent. n69 Otherwise, government would be violating the principle of content neutrality by distinguishing among ideas and privileging some viewpoints over others. n70 The civil liberties theorists posit that regulation of racial or other specific types of [*223] epithets and insults is unconstitutional because it implicitly favors one viewpoint -- that discrimination against someone because of her race, gender, ethnicity, sexual preference, or religion is wrong -- over others -- that members of these groups are bad, inferior, or otherwise deserving of contempt.

The argument that hate speech regulation constitutes nonneutral, content-based discrimination is a powerful and rather subtle objection that may become clearer with an example. Assume that male students yelling "faggot" verbally assault a gay man on a college campus. Nearly everyone would agree that this speech is offensive and hurtful. A libertarian, however, would argue that regulation of this insult, but not other kinds of wounding verbal insults (for example, "asshole"), implicitly embraces the normative statements that homosexuality is an acceptable lifestyle and that discrimination on the basis of sexual preference is wrong. One might respond that a content-neutral explanation justifies the selective regulation; that is, the injury is greater with a sexual preference insult than with other slurs because of the history of pervasive legal and social discrimination against gays and lesbians. But members of many groups -- such as people with handicapping conditions -- likewise could claim a history of social abuse and legal injury. If, at the root, the argument for punishing the epithet "faggot" without punishing all demeaning slurs includes the normative claim that discrimination on the basis of sexual preference is wrong, or worse than other forms of insults, then the argument does seem to violate a strict neutrality principle. Moreover, the particular normative claim it implies does not coincide with majority attitudes. n71 If some people believe that homosexuality is wrong, and if some of them express this viewpoint venomously and inelegantly by calling someone a faggot, then how can we demand that they be silenced without making a controversial judgment about the content of their expression? Unless we ban all such insults and epithets, or equally hurtful means of expressing our unease with another's difference, then we seem to be endorsing government regulation of the content, if not the viewpoint, of the speech.

[*224] If this characterization is an apt description of the impact of hate speech regulation, then one can readily appreciate why many first amendment defenders, especially libertarians, feel compelled to condemn such regulation. To them, the underinclusiveness of the regulation is a sign of an underlying governmental preference for the idea of nondiscrimination. However worthy and widespread this idea may be, some theorists insist that government cannot promote nondiscrimination by censoring speech in opposition to it. n72

One might instead propose that racial or gender epithets, rather than homophobic slurs, are a more appropriate subject for suppression, because fourteenth amendment jurisprudence makes clear our public stance toward the idea of race and sex discrimination, whereas discrimination against gays and lesbians is not, at present, unconstitutional. In other words, hate speech regulation is supportable whenever the speech is directed at members of a "suspect (or quasi-suspect) class," as defined by equal protection jurisprudence.

Civil libertarians also would reject this refinement, however, because it draws political consensus justifications into the argument for suppressing speech, a move that collides directly with the conventional wisdom that the first amendment exists primarily to protect the right of political dissidence. In essence, the argument implies that national consensus, as expressed in the emerging interpretations of the Bill of Rights, is a proper limitation on free speech. The obvious problem with this argument is that if we freeze national consensus at any particular historical moment and repress all speech that is seriously inconsistent with, or regresses from, that viewpoint, then we will curtail revolutionary possibilities. In many ways, subordination and equality are contested, contextual, and fluid phenomena. Although we may be willing to bound our interpretations of the fourteenth amendment to current estimations of equality, the free speech clause is more future-directed and focused on generating alternative interpretations.

A second objection to anchoring an argument for repressing racial or other epithets to the Constitution is that this makes possible the awkward argument that these epithets are a form of political counterexpression. For example, a vulgar sexist slur could be characterized as a revolt against the fourteenth amendment [*225] proscription of discrimination against women. This move would locate hate speech in the "core" of first amendment protection and thus would strengthen the claim that it deserves the highest level of speech protection.

A final objection to an argument that invokes majoritarian values in defense of speech suppression is that this argument runs contrary to the basic assumption that the first amendment exists to protect the minority from an oppressive "pall of orthodoxy." n73 The Bill of Rights, after all, is the antimajoritarian rider to the democratic Constitution. It exists to protect the individual from the overwhelming force of the community, at least in some isolated and "fundamental" respects. Thus, although a majority of the community may wish to express its commitment to equality by censoring speech that vilifies people because of their race or other protected status, the first amendment prevents it from doing so. The community can restrict the act of discrimination or engage in activity or speech that fosters equality, but it cannot inculcate this value by suppressing speech of those who disagree with it.

These civil liberties objections to hate speech regulation are based, in part, on an exceptional emphasis on individual freedom from the dominant community. Such freedom is difficult to square with the community's desire and obvious need to inculturate its members into certain preferred communal values and practices. Other political theorists, in contrast, are suspicious of the dominant community and are anxious to preserve ways of individual escape from its coercive influence. To them, the first amendment is the main vehicle of escape.

The civil libertarians' distrust of the dominant community springs in part from a strong belief that many, if not all, alternatives to dominant communal practices may be legitimate and valuable. Freedom of expression maximizes the chances that these alternatives can challenge, supplement, and even supplant the dominant practices. The result is a more vigorous, adaptable, and generative society. Given these beliefs, the civil liberties theorists view any retreat from, or effort to contextualize, the rule that offensive speech is protected speech as a threat not [*226] only to individualism, but also to cultural pluralism and vitality. n74

Another signal apprehension of the libertarians is that approval of hate speech regulation may open the door to other normative judgment calls, which could lead to the suppression of other worthy forms of dissidence, not merely the speech of Frank Collin and his ilk. n75 Fear of this boomerang effect is, in their minds, considerable and should make people wary of any "ought-based" arguments for hate speech suppression. That is, they might concede that some speech is bad -- even worthless -- but believe it is too risky to try to carve out exceptions to fit only such speech.

In support of their claims that the harm in libertarian is real and that the risk that the power will be abused is substantial, civil liberties theorists point out that the work "McCarthyism" is both homegrown and of recent vintage. n76 The McCarthy era shows that first amendment provincialism can be ruinous to human lives and that freedom of speech is always vulnerable to capture by shifting political and moral alliances, even in the United States. Moreover, libertarians are inclined to warn, the victims of free speech repression here and elsewhere often have been civil rights activists. Thus, civil rights theorists, of all people, should recognize that the expectation that government will do good things with expanded power to regulate "uncivil" speech is simply at odds with much American historical and contemporary experience. Indeed, conversation-closing moves of any kind are the modus operandi of those who favor the status quo, not of those who wish to change it. Civil liberties theorists thus regard the decision to abandon aggressive first amendment protection "in this case" as inconsistent with a commitment to rethinking the status quo, and even as a form of civil rights movement heresy. n77

The foregoing points reveal that the civil liberties objection to hate speech regulation takes two forms. The first form is that content regulation is wrong per se, as a matter of neutral first amendment principles. The second form is strategic. This form [*227] maintains that even if some content, including hate speech, deserves to be silenced because it is worthless or evil, governmental content regulation nevertheless is too dangerous because we cannot devise a truly distinctive, easily understood, and cabined rationale for censoring hate speech, but not other forms of offensive speech.

* * *

The libertarian approach to offensive expression often is justified in terms of two, alternative philosophical stances. The first stance is that of post-Enlightenment rationalism, which holds that objective reason and truth exists, that the pursuit of "Truth" is our highest aspiration, and that reason is the compass that points us to Truth. Classical American first amendment theory reflects these rationalist assumptions, insofar as freedom of speech is often justified in terms of this truth-seeking objective. Indeed, the most common description of the purpose of broad freedom of expression is that such freedom is essential to the advancement of knowledge and the discovery of truth. n78

Many modern philosophers, however, have become skeptical of rationalism in ways that are relevant to the traditional liberal first amendment justifications. Some scholars now maintain that our notions of good and evil are contingent, n79 and that reason, trust, and meaning are determined historically and culturally -- not objectively or a historically. n79

The contingency claim is destabilizing and imperils the traditional rationalist justification for freedom of expression. The strongest and least palatable form of this claim suggests that all ideas, and thus all forms of discourse, may be relative because we have no neutral, a historical way of choosing among them. Thus, either we can choose arbitrarily and oppressively from among the competing expressive possibilities, or we can abolish most, if not all, restrictions on conversation, in recognition of the contingent nature of our understandings.

Few theorists, however, embrace this strong version of the contingency insight. Most writers would reject the claim that all conversations are equal. Nevertheless, they are persuaded that rationalism is a far less useful or convincing construct than previously believed. Skepticism thus has become their first, and often controlling, instinct, particularly when confronting matters of moral or political philosophy.

[*228] Skepticism has changed some liberal thinkers. Those who have remained liberals despite the demise of rationalism have had to reevaluate their political assumptions, including their assumptions regarding free speech. Most have concluded, however, that liberalism can survive the alleged death of rationalism, and that liberalism minus rationalism still points toward free and open discourse. n80

The revised liberal argument for free speech is expressed in different, yet familiar, terms -- such as "[o]ne man's vulgarity is another's lyric." n81 The outcome, however, remains the same: an embrace of free and open discourse, no matter how jarring the ideas. If anything, the contingency insight makes the liberal argument for unbridled speech more, not less, compelling because strong rationality claims drop out of the equation and thus do not bound the sphere of protected discourse.

The practical difference between the older "liberalism based on rationalism" and the newer "liberalism minus rationalism" n82 thus may not be profound. Both formulations express the liberal end in Millian terms: it is an attempt to effect an optimal balance between leaving the individual's private life along and preventing suffering. n83 Both formulations also would embrace strong versions of the claim that "in respect to words as opposed to deeds, . . . anything goes." n84

Liberals under either view encounter extreme theoretical difficulties, however, whenever words resemble deeds and cause serious human suffering. Only by invoking the often arbitrary distinction between speech and conduct, n85 by romanticizing the current conditions for free and open discourse, and by exaggerating the usefulness of counterspeech or averting one's eyes can [*229] liberals remain comfortable defending hate speech. To regulate the content of speech is to undermine a liberal first principle: truth is discovered in, or is whatever results from, free and open discourse -- not the marker of its permissible contours. Yet, to permit hurtful discourse is to violate another baseline assumption: individual freedom is subject to restriction when it causes harm to others.

Historically, liberals escaped from this circle by establishing a strong presumption against speech censorship. Liberals demand that the suffering caused by speech be more powerful than that caused by other forms of behavior before the speech can be suppressed. Consequently, liberals tend to favor only those restrictions that seem absolutely necessary to preserve the most basic conditions of civil discourse and hence will endorse, at most, a rule that proscribes speech hurled in a victim's face like a punch and that is likely to inspire immediate physical retaliation. n86 Only when a racial or other epithet satisfies these conditions can it be proscribed.

The liberal justification for restraining these assaults, though, relies solely on the analogy to a physical assault, rather than relying on psycho-emotional concerns. The reasons for this restriction are several. First, much speech contains the potential for causing psychological or emotional wounds. This potential disinclines liberal theorists to endorse a rule that psychologically wounding speech is regulable, because such a rule would be too open-ended. Second, liberals discount the psycho-emotional harms of speech because they tend to presume that all people have or should develop the fortitude for penetrating, destabilizing, and invasive verbal volleys. This corresponds with liberalism's assumption that individuals are powerful and atomistic beings with an extensive and equal capacity for self-definition.

"Suffering" thus assumes a restrictive, rather elusive meaning within liberal philosophy -- one that is narrower than psychic or emotional harm. Defining the proper balance between private freedom and preventing human suffering is an ongoing liberal inquiry, not a matter of ready reference to a canon. In general, however, the liberal assumption is that offensive speech, by itself, is not a sufficiently wounding invasion to merit repression, unless [*230] uttered intolerably close to the target and threatening to cause imminent, extreme, and usually physical consequences.

These philosophical, political, and practical observations offer a richer sense of why a civil liberties theorist would defend the right of the Klan to speak in Chicago's Marquette Park or of Frank Collin to speak in ie. They also indicate why defending that right is such a distasteful liberal task. Liberals tend to assume that speech is not cruelty in most instances, but they know that this assumption is weak as applied to hate speech. They also recognize that intolerance can threaten the baseline social condition of a liberal society: general recognition and respect for our shares vulnerability to the pain of humiliation. n87

Liberal theory thus handles poorly the problem of racist and other vicious slurs and epithets. A political philosophy that emphasizes endless ways of escape from cultural norms offers no good response to the problem of whether to allow escape into oppression of difference. As such, the hate speech issue presses against liberalism's fontanel -- the place where its theoretical vocabulary is least compelling, most paradoxical, and least responsive to real world conditions and actual human experience.

IV.

These vulnerabilities in liberal thought are not lost on the civil rights theorists. In particular, they object to the way in which liberal argument tends to assume away human pain in order to preserve its basic formula that the answer to bad speech is counterspeech.

The civil rights theorists reverse the priorities set by liberal theorists. In their view, equality trumps speech, insofar as freedom of speech is meaningless absent true equality. n88 This reversal leads the civil rights theorists to favor restriction of hate speech in the form of student discipline, civil remedies, and criminal sanctions. The most radical of their proposals is that hate speech directed at a member of a subordinated group should be punishable, [*231] whereas hate speech directed at a member of the dominant group should not be. n89

The strong civil rights critique of the liberal objection to hate speech regulation is binary. The first line of attack is paradigm-accepting, in that the theorists argue from within the traditional first amendment framework in defending hate speech regulation. The second line of attack is paradigm-shifting. This more radical step to the critique challenges the liberal assumptions that hold up the civil liberties conceptual structure. Once these support beams are weakened, the traditional content-neutrality principle itself becomes less persuasive, not merely subject to tight exceptions.

The paradigm-accepting step rejects a narrow construction of precedent and emphasizes that the Supreme Court already has made content-based distinctions in other areas. n90 The general first amendment practice is to suspend judgments about good and bad speech as often as possible and to destabilize claims to authority by allowing confrontational, revolutionary, offensive speech on our streets and in our parks. We do, however, have numerous Court-approved exceptions to this rule. Speech interests often must yield to interests in privacy, n91 curtilage, n92 reputation, n93 and repose. n94 Moreover, when it comes to children, we forego even the pretense of value skepticism and openly inculcate community estimations of decency and democracy through government-run education, despite the first amendment. n95 Thus, cultural, historically bound estimations of "harm to the community" already limit freedom of expression, despite our alleged commitment to value relativism or to the "marketplace of ideas."

The civil liberties theorists believe that the harm produced by hate speech equals or exceeds the harm produced by the body of speech that already receives less than full first amendment protection. Drawing on social science data, psychological studies, [*232] historical practices, and personal experience, the civil rights theorists maintain that racism is a distinctively abusive phenomenon, so that the government may prevent injuries based on race in ways it cannot prevent other kinds of assaults to a person's dignity. n96 They cite Brown v. Board of Education, n97 as constitutional recognition that racism is a distinctive social harm and that the fourteenth amendment protects against stigmatic injuries, as well as economic, political, and physical ones. n98 Regulation of discriminatory speech, they argue, is a logical and constitutionally defensible extension of Brown.

The civil rights theorists also believe that the liberal refusal to view hate speech as low value expression betrays insensitivity to the nature and seriousness of racial insults. n99 Only people who have not experienced this abuse, they remark, are inclined to assume that a racial epithet is no worse than a nonracial invective, such as "Asshole!," and that a racial epithet thus should be subject to no more restriction than this other sort of verbal attack. The civil rights scholarship tries to bridge this empathy gap with narratives and other empirical evidence that make vivid the nature and the severity of the wound racism inflicts. n100

Some of the civil rights theorists -- at least thus far in development of their proposals -- claim that the wound of racism, versus other forms of discrimination, truly is distinctive. It is sui generis, according to Mari Matsuda. n101 Yet several of the new hate speech proposals extend beyond racial slurs and include slurs against women, religious groups, gays and lesbians, and members of other outgroups. The apparent basis for including these other groups is that slurs and epithets based on gender, religion, sexual preference, and certain other characteristics likewise are subordinating, degrading, and implicitly connected to physical violence. n102 The peculiar power of these slurs is tied to historical patterns of abuse against the protected class members, which render the members of all of these marginalized groups more vulnerable to group-referent slurs than members of dominant groups. n103

[*233] The civil rights theorists invoke the fighting words doctrine, n104 but with a twist. Fighting words, in constitutional parlance, mean words that "by their very utterance inflict injury or tend to incite an immediate breach of the peace." n105 More specifically, the phrase refers to words that are "likely to provoke the average person to retaliation, and thereby cause a breach of the peace." n106 Some civil rights theorists argue that the Court should extend the fighting words doctrine to include epithets that are directed more generally at all members of the group, on the theory that such speech likewise inflicts injury and threatens the social peace and the mental peace of the members of the target group. n107 Moreover, they deny that likely physical retaliation is the proper standard for measuring the acceptability of insults. Instead, a "fight or flight" response should be the measure, n108 in recognition that physical retaliation is not a prudent or natural response for many members of marginalized groups. "Flight" refers both to physical withdrawal and to the internal adverse emotional reaction to verbal attack. n109 Thus, civil rights theorists favor, at a minimum, a "fighting words plus" version of the Chaplinsky formulation.

The civil rights theorists also observe that extensive, relatively noncontroversial case law supports the government's right to regulate hate speech in certain restricted venues. The "captive audience" doctrine, n110 the "public forum" doctrine, n111 and the case law that permits government-as-employer, n112 government-as-educator, n113 or government acting in other capacities with significant control over speech all provide stable authority for the claim that highly offensive speech, including hate speech, can be controlled in certain contexts, even if it cannot be controlled on the street corner.

[*234] This paradigm-accepting step to the civil rights theorists' argument draws heavily on Richard Delgado's 1982 article, in which he outlined a tort remedy for racial insults. n114 Delgado argued that the available tort remedies of intentional infliction of emotional distress, invasion of privacy, and libel, by themselves, inadequately redress the special problem of hate speech. n115 He therefore recommended development of a new tort that is tailored to the particular and distinctive psychological and physiological injury occasioned by racist speech, but that resembles closely the tort of intentional infliction of emotional distress. n116 In essence, his proposal takes what we would ordinarily regard as evidence of the outrageousness of an insult -- that it is racial, draws on a history of subordination, and implies physical violence -- and uses that evidence as the basis for forming a subcategory of "outrageous per se" remarks, that is, racial insults. The plaintiff still must prove damages, n117 intent to demean, and that a reasonable person would understand the particular language used to be a demeaning racial insult. n118 What she need not do, under Delgado's proposal, is establish that racial insults constitute a serious civil wrong. Thus, his proposal, like the proposed extensions of the fighting words doctrine, relies heavily on traditional legal concepts, but expands those concepts to take particular, customized account of the problem of racism and other forms of group-status bias against marginalized people.

In summary, the paradigm-accepting step to the argument for hate speech regulation emphasizes two things: the harm to victims of hate speech is serious, and the value to society in allowing this speech is slight. Thus, according to the civil rights account, if the first amendment is about balancing harms and values in speech, then the balance tips in favor of suppressing hate speech.

One could stop at this point and join issue on the problem. Indeed, lay discussions of the matter tend to hover at this level of abstraction, and policy decisions often are made without a deeper inquiry. But, as was true of the civil liberties approach to hate speech, one can understand fully the civil rights approach only if one dips below this surface into the underlying philosophical and political assumptions that animate these writers. And, [*235] as was true of the liberal approach to hate speech, these background assumptions have a subliminal, if not always articulated or apprehended, impact on the frontline lay discussions. The two approaches represent, in many important respects, quite different views of the world -- differences that must be understood if they are to be transcended.

The differences become manifest in the paradigm-shifting step to the more radical civil rights proposals. These proposals would make hate speech unlawful in all settings, not merely in face-to-face individual confrontations. Here, the civil rights theorists shed the cramping influence of traditional discourse and attempt to blaze new paths. Their arguments become more threatening to liberal thought and more resistant to accommodation efforts. They also become more arresting, in part because the writers use narrative in their arguments and in part because the analytical arrows strike at such undeniably crucial liberal targets.

The first move of the radical civil rights theorists is to challenge liberals' devotion to individualism. They argue that an overemphasis on the individual and a disregard for the " group(s) behind the man" infects first amendment theory, as it does other aspects of constitutional law. n119 To discount the significance of group affiliation, however, is to miss the true harm of group vilification. If one is defined in part by one's group affiliations, then statements that demean one's group are a matter of individual concern and injury. Thus, racist comments do wound self-esteem, even if they are not targeted at a particular member of the group.

The salience of group identity also is relevant to our theory of rights. If these subgroups are integral to the personal identity of subgroup members, then discrimination against a subgroup, as such, will disable its members. Some scholars therefore urge that wrongs to the group deserve group-sensitive remedies like affirmative action. n120

The argument that group identity is important to human personality, and that to recognize its importance points toward banning group libel, is not new. In 1942, David Riesman authored a thoughtful defense of group libel statutes in which he relied [*236] on the significance of groups within the social process. n121 His observations sound prescient and likely have influenced many contemporary writers who advocate a similar approach to hate speech. Riesman wrote:

In the political as in the economic struggle, modern democracy operates through the interplay of group activities, and it is through participation in groups that persons contribute to the social welfare and develop their individual capacities. Hence, defamatory attacks on groups are attacks both on the pluralistic forces which make up a democratic society and derivatively on the individual members whose own status derives from their group affiliations. Yet the very importance of groups in the democratic process means, if stratification is to be avoided and a dynamic social life retained, that each group must be subject to the scrutiny and criticism of opposing groups -- and of its own membership. Here again, policy must be discriminating in judging what sorts of criticism -- though mistaken in fact -- further the democratic cause and what sorts of defamatory falsehoods hinder it. n122

As Riesman observed, the most significant obstacle to group libel statutes in the United States is not definitional problems or other technical difficulties; it is "the American heritage of middle-class individualistic liberalism." n123 He rejected this strong embrace of individualistic liberalism in favor of democratic principles. This democratic bias led Riesman to endorse a civil action for group libel and to deny that the Constitution bars such a remedy. n124 The regulation he endorsed, however, anticipated that truth would be a defense to a charge of group libel. n125 This method is a very clumsy way of dealing with the harm of most epithets, in that few slurs are phrased as statements of fact. What is admirable about Riesman's piece, therefore, is not that he managed to work out an ideal response to hate speech, but that he offered such an illuminating unpacking of Americans' resistance to hate speech regulation and laid some of the intellectual foundation for modern civil rights proposals.

This is not to say, however, that Riesman's democratic principles necessarily solve the riddle of how to handle hate speech [*237] any better than liberalism does. In the above-cited passage, Riesman acknowledges that intergroup defamation both imperils and furthers democracy. To limit speech that is an outgrowth of tensions between and among groups, including racial groups, thus is to risk limiting "core value" political speech, even under Riesman's democratic approach. That is, we may acknowledge the importance of group affiliations to human personality and reject more aggressive versions of individualism, yet still be uncertain whether a particular form of confrontational speech between groups is a tolerable feature of democratic or communal life. Nevertheless, an awareness of the significance of group affiliation to human personality clearly enhances one's appreciation of the wound of racism, misogyny, or homophobia. This awareness, which the civil rights theorists stress, makes hate speech appear less worthy of protection. When coupled with a rejection of strong individualism, this awareness may tip the harm/benefit balance in favor of hate speech suppression.

In recent years, many constitutional scholars have joined in the Riesman critique of the liberal emphasis on individualism. Several of these scholars have attempted to define a more republican version of constitutionalism, one more mindful of the benign aspects of community and the ways in which we are all shaped, defined, and realized by these connections. n126 These communitarian theories may prove useful to the civil rights theorists, at least to the extent that they lend support to arguments for suppressing hate speech. Communitarianism is responsive to appeals to solidarity, shared values, and the authority of the community to impose standards of civility on those who threaten the integrity of the whole or of its subgroups. Communitarians thus may be more inclined than are liberals to suppress hate speech in the interest of group solidarity. n127 As such, communitarians and civil rights theorists may become allies on the issue of hate speech regulation.

[*238] This alliance does not mean, however, that civil rights theorists necessarily embrace communitarianism, or that all communitarians necessarily would endorse group libel laws. Civil rights theorists are outcome- rather than process-oriented. Until the outcome of equality is achieved, process-oriented models are of secondary concern.

Moreover, as many writers have cautioned, pure communitarianism is only as attractive as the community in question. n128 In an oppressive, nonegalitarian community, to defer to the communal will would be unbearable. Civil rights theorists are unlikely to embrace unmodified communitarianism or any other pure political model and follow it wherever it leads; instead, they begin with a view of where they wish to go and adopt political theories, or parts thereof, most likely to get them there. Their destination is a society in which human difference is not the occasion of subordination. n129

This firm committment to a specific substantive outcome is remarkable and further separates the civil rights theorists from the liberal theorists. The civil rights theorists make unapologetically ought-based arguments. For example, Mari Matsuda has said that "[w]e can attack racist speech -- not because it isn't really speech, not because it falls within a hoped-for neutral exception, but because it is wrong." n130 The contingency insight makes this absolute moralism sound "unmodern" as well as illiberal, [*239] especially when invoked as a reason to suppress speech, rather than conduct.

This moralism nevertheless appeals to some peoples' very strong sense that relativism is bad social policy, if not bad metaphysics. Very few people, including intellectuals, truly believe that all ideas are of equal social value. Most people feel, though not with equal passion, that we both can and must give less approbation to the ideas of Hitler than to those of Dr. King. That is, even if we presently lack a convincing, ahistorical matrix with which to justify our choices, we still believe we know evil when we see it. n131 The civil rights theorists invoke our particular "knowledge" that racism is wrong and suggest that we retreat from the "infinite expansion" concept of public discourse to one that better accounts for this particular, shared knowledge.

A third departure from liberalism which appears in the civil rights theorists' scholarship is the renunciation of the public/private distinction. n132 Here again, the intellectual seeds were sown some years ago, but are now beginning to bear more impressive and abundant fruit. As applied to the hate speech issue, the argument is as follows:

[T]he choice in the group libel problem is not between the restraint of free expression and the absence of restraint. It is rather a choice between two forms of restraint: one carried out by private groups operating outside the law, and another, of a more limited nature, carried out by legal authorities under the constraints of a formal statute. n133

That is, coercion of individual expression may occur regardless of whether we regulate hate speech. Government is not the only instrument of individual oppression; indeed, government may often be the only way to prevent this oppression. To erect an arbitrary barrier between the public and private spheres and to argue that constitutional violations arise only when government acts -- not when it fails to act -- is to let stand circumstances that may seriously compromise the constitutional aspirations of equality and of free and open dialogue.

[*240] Applying this reasoning, the civil rights theorists argue that when government does not regulate hate speech, it lends its imprimatur to assaults on values espoused in Brown v. Board of Education. n134 Omission, in this case, is commission. The civil rights violation thus is linked to the government and is a matter of constitutional moment. n135

More generally, these scholars note that significant Supreme Court cases already have outflanked the Maginot Line between public and private discrimination. n136 Moreover, federal statutes impose extensive nondiscrimination mandates on private employers, such that vigorous insistence that the Line cannot be crossed comes too late. As a compelling example, civil rights theorists note that sexual harassment is speech and a violation of federal law. n137 Hostile environment claims of workplace discrimination likewise represent occasions in which speaker autonomy collides with victim equality, but the equality interests prevail. n138 Thus, argue some theorists, states may pass laws that regulate hate speech in other, less restricted settings, including public fora, in order to promote the compelling state interest in equality.

A final distinguishing characteristic of the civil rights approach to hate speech is methodological. Many of these scholars employ narrative in order to stir imagination and to inspire recognition of the harm in racism and in other subordinating instincts. n139 They abandon the detachment that is characteristic of traditional legal scholarship and engage in storytelling as a means of sparking empathic understanding. But they do so with a common mission: to further the cause of equality. In this way, they engage in the activity that Richard Rorty recently described as the only one that can bind humans together and eliminate cruelty. n140 He [*241] believes that "novels and ethnographies which sensitize one to the pain of those who do not speak our language must do the job which demonstrations of a common human nature were supposed to do." n141 If his is an accurate account, then the deepest significance of the civil rights scholarship may lie not in the specific proposals for legal reform, but in the way in which the literature sensitizes us to the pain of others and fills the void left by philosophy's failure to point us on a sure path to truth and virtue.

* * *

Few people oppose the civil rights substantive agenda of eliminating cruelty and combatting discrimination. Some nevertheless take issue with the civil rights approach to hate speech. One objection is that, like the liberalism that underlies the civil liberties approach, the equality theory that underpins the civil rights proposals is vulnerable to analytical critiques. n142 A second objection is that the specific proposals fail to respond fully to several of the practical objections mentioned earlier. n143

The first, analytical problem is that the equality theory on which some theorists rely runs contrary to some constitutional theory and popular sentiment. Mari Matsuda proposed a one-way version of hate speech regulation, under which slurs against subordinated people are punishable, whereas slurs against dominant group members are not. n144 The advantage of this approach is that it takes into account our shared sense that the impact of the epithets used against marginalized groups is far greater than that of slurs against dominant groups. This approach also rescues from punishment the speech of African-American rap groups like 2 Live Crew n145 -- at least under a hate speech statute -- as well as other confrontational and racist speech used by outgroups to attack the dominant group. The underlying assumption, with which I agree, is that equality is not merely a matter of identical social and legal treatment of individuals, but it also must take into account inequalities in legal and social outcomes and the historical maldistributions of social goods and political power.

Many Americans, however, have difficulty accepting an asymmetrical account of equality. For example, a very thoughtful university president, who presided over the changeover of a [*242] formerly all-male liberal arts college to a coed institution, once expressed utter amazement that some people who opposed the university's all-male admissions policy did not also object to a neighboring school's all-female admissions policy. n146 He saw this position as fatally contradictory and as a denial of the principle of gender equality. In his view, equality means that discrimination on the basis of gender works both ways. A college should deny access to neither men nor women because of their gender. Asymmetrical hate speech proposals thus might well jar his and other people's notion of equal treatment because whites could be punished for racist talk, whereas people of color could not.

In effect, the Matsuda approach makes the criminality of speech hinge on the race of the speaker and the victim. As such, the approach collides with the constitutional principle expressed by Justice Stewart in Loving v. Virginia, n147 that "'it is simply not possible for a state law to be valid under our Constitution which makes the criminality of an act depend upon the race of the actor.'" n148

Even a civil statute could violate popular and judicial notions of equality when viewed as a remedial measure designed to correct past imbalances of power. As Richmond v. J. A. Croson Co. n149 and the Court's more recent FCC decision n150 demonstrate, however, the status of the argument that past, general discrimination against a subgroup constitutes a good, present reason for state or local government to adopt wide-ranging, race-conscious remedial measures remains constitutionally complex and politically charged. A group libel statute that protected some, but not all, subgroups from racial slurs based on whether the subgroup members were victims of past discrimination may well encounter the same judicial and social resistance as have government contract set-aside programs and affirmative action plans.

Contemporary constructions of equality thus seem to range from straightforward, procedural equality -- everyone treated the [*243] same under an individual rights model -- to more aggressive, substantive equality -- everyone entitled to equal outcomes, including group rights, as a component of the analysis. These disagreements about the meaning of equality, of course, are highly relevant to the hate speech controversy. If one disputes the underlying equality-of-outcome-across-groups theory, one likely will reject the group-sensitive remedy that grows out of this theory. Moreover, to impose speech limitations based on a culturally contested theory seems especially wrong.

The analytical problems with this one-way approach to hate speech regulation go beyond the fact that it relies on a contested interpretation of equality. It arguably asks the government to take a side in intergroup hostilities, n151 if only to even the score. This is not, under traditional theory, a proper government role. Rather, government is expected to remain neutral when policing intergroup conflicts. To intervene in a manner that protects 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.

The one-way proposal also sets forth an amorphous criterion. Who, exactly, are the historically subordinated, marginalized groups in American society? Economically marginalized people include whites, especially white women. Moreover, census reports indicate that by the turn of the century one out of every three American students will be a minority. n152 As cultural pluralism [*244] increases, intercultural conflicts are inevitable. The downturn in the economy almost certainly will hone these frictions, particularly if recent studies of the roots of racism prove accurate. n153 If the stress and frustration of eroding economic turf cause some white Americans on the economic fringe to resort to hateful racial invectives, it may be poor social policy to subject their taunts to criminal prosecution or civil damages, but not those of their Mexican-American, African-American, or Asian-American neighbors. Likewise, in communities in which African-American, Hispanic, or Asian representation in positions of political authority is equal to, or superior to, that of whites, there may be no contemporary reason why a mayor or council member of color should be excused from a law that makes racial slurs a crime. For all of these reasons, a one-way approach to hate speech may prove to be bad social policy.

A second, conceptual and practical difficulty that arises with all of the civil rights theorists' proposals, whether one-way or two-way, is that their most compelling evidence of the harm in hate speech involves only racial discrimination. In order to meet the objection that too many kinds of speech are emotionally wounding to permit psychic pain alone to justify suppression, the civil rights theorists have responded that racist speech inflicts a distinctive harm. n154 Yet some of their proposals nevertheless cover misogynist speech, homophobic speech, and, in some cases, speech that attacks people on the basis of religion, age, or handicap. n155 This opening up of the definition of hate speech threatens to undermine the justification that hate speech causes a distinctive stigmatic wound. The argument that all of these groups have comparable histories of subordination and vulnerability to physical violence may not convince some critics, particularly those who fear elastic exceptions to freedom of expression.

[*245] Expansion of the category of protected characteristics beyond race also undermines the appeal to communal knowledge as a basis for controlling speech. Perhaps we all know that racism is wrong, but we surely do not all know that homophobia is wrong. n156 Also, we clearly disagree about which specific manifestations of racism, homophobia, or misogyny are wrong. For example, some people believe that racist speech by outgroups is not wrong, whereas others disagree. Indeed, if everyone agreed about all of these things in the way some theorists insist, then the traditional remedy of counterspeech would be a highly effective, even unnecessary, means of drowning out the hate message, at least for group libel directed at a general audience. That is, the "wrongness" of racist or other forms of hate speech remains contextual and contested. But if it is contested, then the knowledge justification for suppressing most hate speech disappears.

On the other hand, if the civil rights theorists limit the definition of hate speech to include only racial slurs, which some of them have, then other problems emerge. They may alienate loyal civil rights allies and cause hurtful, factitious disputes. Moreover, they may appear to rank human suffering in ways that may not be justifiable or likely to produce positive social change. The better move thus may be to broaden the category of protected characteristics and to cultivate our vocabulary of the shared features of racism, misogyny, homophobia, and religious persecution, rather than to stress the distinctive nature of racism. But, as indicated previously, this move makes the hate speech exception fairly open-ended.

The reflexive nature of prejudice gives rise to another potential objection to some of the hate speech proposals. Charles Lawrence argues convincingly and eloquently that racism is pervasive and [*246] often unconscious. n157 As such, racist slurs often may be made with no specific intent to wound or insult others. A recent incident that brought this problem to popular attention involved the 60 Minutes television program commentator, Andy Rooney. Rooney made several statements on television and to a reporter that sparked charges that he was biased against gays and African-Americans. n158 Rooney defended his remarks in part on the ground that he intended no slur, and he insisted that he was unaware that the statements would offend others. n159 If, however, "Rooneyisms" are the product of ignorance and bone-deep, subconscious racism, then to punish people, especially with criminal sanctions, for these unintentional, reflexive acts of cruelty seems harsh. Some of the hate speech proposals nevertheless do not list specific intent as a condition of punishment.

On the other hand, if hate speech regulation covers only intentional slurs, then a significant body of equally hurtful speech will remain unregulated. Indeed, this may leave the heart of the problem unregulated. A panel discussion of hate speech on campus, held at the University of Arizona, raised this issue. Both a representative from WINGSPAN, a campus gay and lesbian association, and an African-American university administrator panel member remarked that skinheads and other members of hate groups were not their main concern. Rather, these panelists most feared "the boy next door." n160 That is, they regarded the unintentional acts of racism and homophobia as more threatening and destructive than the occasional, clearly outrageous outburst of slurs and epithets commonly understood as hate speech. Yet, to write statutes, especially criminal statutes, that proscribe these unwitting forms of racism struck these two panelists, and most members of the audience, as unreasonable.

This leads to another objection to the extant civil rights proposals: the constitutional proposals may be exercises in futility. n161 The real problem with hate speech, some argue, is the underlying attitude. n162 Only very intrusive regulation, such as mandatory [*247] consciousness raising, might actually dislodge the prejudice that animates hate speech. The proposed hate speech statutes thus will not begin to solve the problem. These commentators point out that the people in the nations that have adopted hate speech laws like the ones now being proposed in the United States have not become less racist. They also remark that silencing bigots only sends bigotry underground and may give bigots martyr status to boot. n163

The civil rights rejoinder to the argument that bigotry will go underground is, "Good." Because the speech is intrinsically harmful, stopping the speech is just as worthwhile as stopping a slap, or a rape, even though the aggressor may still want to hurt you. This justification, however, only supports stopping the speech if and when it is like a slap or other physical harm, which points toward a tighter causation requirement than the more aggressive civil rights proposals anticipate. It points toward proscribing only fighting words, not the less immediate harm of group libel.

If one instead says that the harm lies in its effect on society, not only on a targeted victim, then one comes close to saying that the audience cannot be trusted to weigh the value of the speech on its own. First amendment doctrine counts such a view of people as a bad reason to restrict speech. Unless we have specific proof that this speech is peculiarly seductive and appeals predominantly or exclusively to noncognitive instincts, perhaps like hardcore pornography, then the argument is fatally paternalistic, in many people's minds.

The final obstacle to the civil rights proposals is the quite powerful continued resistance to the total collapse of the public/private distinction. As some admonish, this distinction has been a principal bulwark against governmental invasion of individual privacy, which many civil rights advocates favor. Thus, despite significant erosion of the distinction in certain areas, n164 such as [*248] the workplace, many people would strongly object to a rule that imposed on private, individual actors -- versus private, institutional ones -- identical restraints as are placed on public actors.

Moreover, these critics would offer the doctrinal objection that the Court recently rejected the claim that governmental omission equals commission, even when the omission resulted in the severe physical and mental impairment of a young child. n165 The Court's reluctance to embrace an omission/commission equation, even under such compelling circumstances, likely stems from its fear that this would expose the government to limitless liability under 42 U.S.C. § 1983. n166 The post-Brennan Supreme Court, in particular, is unlikely to rule that governmental failure to regulate hate speech by private individual actors constitutes a government violation of the victim's constitutional rights, except, perhaps, in restricted environments like the workplace or public schools, in which federal antidiscrimination laws are particularly pervasive and well-accepted limits on private conduct.

These practical and conceptual objections to the civil rights proposals demonstrate that even people who are disenchanted with the liberal approach to hate speech may be reluctant to adopt the full civil rights agenda. The conflicting accounts of equality, the practical and political consequences of endorsing group-conscious regulation of speech, and the complexity of defining hate speech in a way that addresses the real injury without colliding violently with even minimalist notions of individualism, all make total allegiance to the more aggressive proposals hard to secure. Thus, although the civil rights discourse makes lucid the gravity and the nature of the harm in hate speech, it does not provide an obvious response to how best to prevent or reduce that harm.

V.

The weaknesses of pure liberalism, coupled with the shortcomings in strong versions of the civil rights approach, have led some scholars to seek another approach. These reformers have authored or endorsed proposals that defer to many of the libertarian [*249] claims and objections, but that likewise acknowledge the strength of the central civil rights proposition that the harm in some hate speech outweighs the potential harm of suppressing it. n167 In several cases, the proposals deal with the specific problem of hate speech on college or university campuses. n168

Unlike the civil liberties theorists, the sponsors and endorsers of these accommodationist proposals conclude that hate speech is an appropriate subject for censorship. Unlike the advocates of the strong civil liberties proposals, however, accommodationists do not, for the most part, favor regulation of group libel. n169 The accommodationist proposals share the common characteristics of being tightly worded, context-specific, and closely tied to the fighting words doctrine and/or the tort of intentional infliction of emotional distress. In essence, the proposals seek to regulate targeted, intentional vilification of a person or small group of persons in a face-to-face encounter on the basis of a protected characteristic. n170

The protected characteristics vary among the proposals. Some list only racial insults. n171 One covers racial, ethnic, or religious group insults. n172 Another proscribes slurs based on sex, race, color, handicap, religion, sexual orientation, or national or ethnic origin. n173 Still another discusses insults based on race, religion, ethnic origin, gender, or sexual preference. n174 These specific differences [*250] among the proposals are important and are likely to prove quite relevant to their success in any court challenge. The more narrow the provision, and the closer it conforms to the fighting words doctrine or traditional tort law, the more likely the regulation will be upheld.

Some of the proposals that deal with campus speech regulation rely heavily on contextual justifications. Specifically, they propound that the university is a special community, within which at least minimal regulation of civility is necessary. n175 Racial and other epithets may so disturb students that they cannot study or interact fully with their teachers or classmates, and thus may become alienated from the life of the college or university. Moreover, Brown v. Board of Education n176 makes clear that full educational equality goes beyond provision of physical resources and includes the intangible elements that produce an environment of full and equal participation in the life of the college or university. n177 As such, some accommodationists conclude that hate speech regulation is a logical and necessary extension of the public university's commitment to equal access and nondiscrimination within higher education. n178 An environment of hostility and intergroup tension, they remark, is hardly conducive to genuine realization of these goals. The relatively restricted campus environment, coupled with the way in which racism and other biases can undermine the educational process, thus provide the government-as-educator greater speech regulation authority than it has as regulator of the general public safety and welfare. n179

Rodney Smolla expressed this sentiment in the following terms, terms that most accommodationists likely would endorse:

A state university is different from a public elementary or high school because by tradition a university is a place of uninhibited public discourse and should remain so. A university, however, is also a unique community in which the state should be permitted to require of its members higher levels of rationality and civility than the state may impose on the general population. It should be permissible for the state to require that members refrain from racist attacks at certain [*251] places and times as a condition for entry into this special community. n180

In addition to substantial agreement on these campus-specific concerns, nearly all accommodationists would agree with the following general observations about targeted hate speech:

If racial and ethnic epithets and slurs are to be made illegal by separate legal standards, the focus should be on face-to-face encounters, targeted vilification aimed at members of the audience. As to these, expressive value is slight, because the aim is to wound and humiliate, or to start a fight. Since fighting words are already punishable and the tort of intentional infliction of emotional distress is available, what would be the significance of separate provisions for the language of group vilification? They could stand as symbolic statements that such language is peculiarly at odds with our constitutional values; and they could relieve prosecutors, or plaintiffs, from having to establish all the requisites of a more general offense or tort. n181

In essence, accommodationists approve of the basic Delgado formulation, though some would extend his proposal beyond racial slurs to include other protected characteristics. They acknowledge that this formulation is a departure from the usual requirement of content-neutrality, but regard it as a warranted departure for the reasons advanced by the paradigm-accepting strand of the civil rights argument. n182 Indeed, the accommodationist position is essentially a civil rights position that works within the existing constitutional framework.

The principal objections to this "fighting words plus" approach to hate speech are those listed in the foregoing discussion of objections to the civil rights proposals. n183 In particular, some object that this moderate civil rights approach is merely symbolic n184 and may be the worst of both worlds, not a delicate synthesis that rescues the most attractive arguments of both sides. Civil libertarians also point out that the perversion of first amendment [*252] goals is more, not less, outrageous when speech regulation occurs on a public university campus. n185

The following, final sections address these objections and discuss briefly the reasons why I agree with the accommodationists, despite these objections.

VI.

Henry David Thoreau once wrote, "It takes two to speak the truth, -- one to speak, and another to hear." n186 The accommodationists' writings reflect an effort to hear both the liberal and the civil rights arguments regarding hate speech regulation. Their midground position may be unsatisfactory to both sides, in the sense that neither side receives all that it desires. But it also may be satisfactory in the best first amendment sense, in that it demonstrates that discourse can influence thinking.

The Stanford policy, n187 drafted by Tom Grey, strikes me as the most convincing of the accommodationist proposals, in terms of both its specific provisions and Grey's explanation of the policy. This policy defines verbal harassment as speech or other expression that

(a) is intended to insult or stigmatize individuals on the basis of protected characteristics;

(b) is "addressed directly" to those insulted or stigmatized; and

(c) makes use of insulting or "fighting" words, n188 defined as words that "by their very utterance inflict injury or tend to incite an immediate breach of the peace." n189

The protected characteristics include "sex, race, color, handicap, religion, sexual orientation, or national and ethnic origin." n190 Punishable words are further defined as those "commonly understood to convey direct and visceral hatred or contempt for human beings on the basis of" protected characteristics. n191 The aim of the policy is to cover only gutter epithets of bigotry -- and even [*253] then, only when targeted at an individual or small number of people. n192

The attractive characteristics of this policy are several. First, it limits speech regulation to the circumstance in which we are most confident that regulation may be warranted: targeted, shocking, verbal affronts. In essence, the policy simply means that a male college student should not be allowed to approach a woman and call her a "cunt" to her face. Nor can a student call a lesbian a "dyke," a gay man a "faggot," and African-American a "Nigger," or a Jew a "Kike" -- to their faces.

The limited reach of the Stanford policy keeps it within the boundaries of sound first amendment practice and philosophy. The social interest in protecting these attacks is negligible, whereas the interest in preventing them is, as the civil rights theorists have demonstrated, compelling. That is, the harm/benefit balance tips in favor of speech restriction in this context. Few people likely would object to the legal restraint of a white man who followed an African-American woman down the street, yelling racial and sexual epithets. Indeed, for all the sound and fury of the civil liberties theorists, the likelihood is that an arrest for harassment made under these circumstances would be upheld. n193

As such, the compelling narratives of the civil rights theorists lead me to worry less about whether this part of the Stanford policy goes too far than whether it goes far enough. For example, the proposal does not cover group libel. Thus, a student or other speaker could use these same words, or express equally destructive and prejudicial viewpoints without using epithets, in a speech to a general campus audience. Why should a college tolerate such discourse, knowing it may wound or enrage some of its students, and compromise further its already fractured efforts to create an environment that celebrates diversity and that welcomes equally all students? The most convincing answer is that group libel laws control ideas, rather than their harmful effects. Only by abandoning altogether our confidence in the audience and our fundamental notion that counterspeech is reasonably effective can we declare that hate speech should be per se regulable in all situations. The civil liberties arguments against hate speech regulation convince me that more expansive regulation than the Stanford policy anticipates is unwise.

[*254] Recent absurd instances of speech regulation, like Florida's attempt to ban bumper stickers that read "Shit Happens," n194 should make us apprehensive indeed about delivering to any state officials, including educators, more speech control than the Stanford policy offers. Moreover, anyone who has participated in counterspeech against the neo-Nazis or the Klan likely recognizes that -- at least when given time to prepare such counterdemonstrations -- the counterspeech can be powerful and highly effective. In addition, in some cases, those people who are likely to be enraged or scored by the bigot's rantings may avoid confrontation by staying out of the area.

I do not mean to imply, by any means, that avoiding the confrontation is a costless or entirely convincing response to verbal bigotry, insofar as the targets of hate speech may be forced out of public settings and into their homes in order to feel safe. As these safety zones get narrower, the lives of potential targets of bigotry become more stunted and unnatural. Women know this and have fought against it on college campuses and elsewhere, in efforts to "take back the night." Fear of verbal harassment, no less than fear of physical assault, may change one's work patterns, jogging paths, choice of evening entertainment, and social patterns.

Despite these serious potential liberty losses, however, I am prepared to avert my eyes in order to promote the free and open discourse ends of the first amendment -- at least when the harassment is a general, purely verbal attack on all women. For example, I was willing to look away when confronted recently by a young man wearing a T-shirt that read "Women are property." (I would not have dreamed, by the way, of engaging in counterexpression; he was an athletic person and substantially larger than I am.) Likewise, I support the University of Arizona's decision that a speaker who often stands in our college mall and declares that women are "whores" as women pass by cannot be silenced. If, however, he confronted a particular woman or a small group of women and continued his verbal assault, he would, in my view, become subject to restraint, depending upon the nature of the epithets and the confrontation. The Stanford policy would treat these situations as I have, and it strikes me as the appropriate [*255] balance between strong individual expressive freedom and the prevention of human suffering.

The Stanford policy also captures best what I regard as the most salient distinction between slurs based on race, gender, or other protected characteristics and other types of words that wound. The key factor is the implicit link to force, indeed to physical violence. For example, a racial epithet invokes the history of physical violence against, and legal subjugation of, African-Americans. This aspect of the epithet, even more than the facts that group affiliation partially defines the self and that race is involuntary, n195 is what makes the racial epithet distinctive. Thus, the target of a racial epithet reasonably may interpret it as an overture to, or reminder of, violence. Likewise, a sexist epithet is heard as an act of aggression and an allusion to rape or other form of physical violence. The fear these epithets evoke is a fear of force, of power. This fear is most powerful, and the remark is most invasive, when the epithet is uttered close enough to carry out the implicit threat, face-to-face, and when the target of the remark is alone or in a small, relatively defenseless group.

I approve, therefore, of continued invocation of the phrase "fighting words" because I regard it as a better reminder of the subordinating and violent character of hate speech than the phrase "intentional infliction of emotional distress." In the case of group libel, which is directed at a general audience, the immediate fear of physical violence is more attenuated. This speech, absent aggravating circumstances, is objectionable primarily because it perpetuates venomous stereotypes. Stereotypes are often reductive and degrading. But they also are part of an ideological framework, however flawed. In essence, stereotypes are shorthand versions of a more elaborate analysis. Wrong-headed ones are bad empirical claims. Rooting out the most pernicious stereotypes should remain primarily an educational endeavor, not a punitive one.

The second attractive feature of the Stanford policy is that is is not limited to racial slurs. Despite the argument that expanding the protected characteristics may undermine the "distinctive injury" defense of hate speech regulation, I favor the broader definition for two reasons. First, it avoids the wrong turn of attempting to rank these related forms of human suffering. I doubt, for example, that the pain of the epithet "faggot" is [*256] measurably less than that of a racial slur. In any event, it strikes me as bad social policy to even try to distinguish among these cognate types of pain. Second, as discussion of the various hate speech proposals continues, other outgroup commentators likely will be able to produce sufficient evidence that epithets based on gender, sexual orientation, handicap, and religion wound in ways that are comparable to the wound of racial epithets.

One may argue that this list is still underinclusive, in that it fails to cover all types of stigmatizing remarks. This criticism is, in some ways, fair. Yet, one can distinguish, as Cass Sunstein has, n196 between underinclusiveness that is a function of "capture" of the regulator by factions or of self-interested decisionmaking versus underinclusiveness that is not a product of either. The underinclusiveness of the policy cannot fairly be characterized as self-interested or as a product of factional tyranny. It has not excluded for political reasons any obviously deserving candidates for protection. The list of protected characteristics is broad enough to cover the entire population, as each of us belongs to an ethnic group, has a gender and race, a sexual preference, and an opportunity to effect ties to a religion. If the policy neglects some obvious group -- though none comes to mind -- it is not because personal reasons or narrow partisan politics unduly influenced the drafter.

Furthermore, the remaining underinclusiveness of the regulation in no way diminishes the claim that the verbal attacks that are included are serious assaults on human dignity. Simply because life offers up many forms of injury does not, by itself, disable the government from preventing some but not all of them. In any event, a principled distinction can be made between the kind of dignity assaults that the Stanford policy includes and those that it omits, in much the same way that we already make legal distinctions between the self-esteem and economic injuries caused by the statement, "You're fired!" and those caused by the statement, "You're fired because you're African-American!"

A third, commendable feature of the Stanford policy is that it is not "one-way," though it most certainly would have a one-way impact. The proposal is framed in terms that anticipate the theoretical possibility that epithets against any racial group might [*257] be regulable. Our practical, contemporary reality is such that no racial epithet against a white person, as such, would satisfy the Stanford policy's standard of "outrageousness." n197 Nevertheless, the proposal is not explicitly wedded to a one-way theory of discrimination. Thus, the criminality of the act does not hinge on race per sex, though the outrageousness of an assault may be a product of actual race relations. This is not, I believe, a figleaf. With the tort of intentional infliction of emotional distress, no reason precludes a jury's taking into account the history of discrimination against African-Americans when assessing the outrageousness of a verbal assault against an African-American. Similarly, the college disciplinary body can take into account this same history when weighing the offensiveness of a racial insult under the Stanford hate speech regulation. In addition, a police officer or judge can distinguish between the breach of the peace occasioned by a white man taunting an African-American woman with sexual and racial epithets, and an African-American man calling a white man a "honky."

That slurs against whites would not satisfy the hate speech standard at present reveals the limited reach of the Stanford policy. It covers only outrageous discourse under very circumscribed circumstances. This is, to my way of thinking, a positive attribute. Again, the policy strives to leave maximal room for protected discourse, while taking into account the most serious harm of speech that barely resembles genuine conversation -- a worthy goal under either liberal or democratic visions of a good society.

The objection that the limited reach of the policy means it serves only symbolic, rather than concrete, ends is unconvincing. Civil liberties people, who in the same breath express deep fear that such policies will chill speech and set dangerous precedent, tend to raise this objection. A purely symbolic, hortatory rule would not, one would think, also pose a great threat to free expression. In any event, our experience with the deterrence value of sexual harassment regulation suggests that sanctioning discriminatory speech may well influence conduct and enhance [*258] the lives of protected class members. Women who have been working for the past fifteen years likely would report that, although harassment remains a serious workplace problem, things have changed for the better, in that much offensive verbal conduct now is widely regarded as improper work conduct and grounds for reprimand by management. Regulation of face-to-face verbal assaults likewise may offer nonnegligible protection against a properly narrow category of harmful speech.

The fourth, extremely important feature of the Stanford policy is that it underplays the significance of the educational "mission" as a justification for hate speech regulation. In fact, the policy, although crafted for a campus, offers a defensible approach to hate speech regulation in any setting. In this way, it avoids the mistake committed in other campus speech regulation proposals of exaggerating the inculcation authority of universities and colleges. Instead, the policy is based on an "equal access" justification that applies in most, if not all, public contexts. n198

The underlying and highly complex constitutional issue is this: Should government-as-educator be granted greater "value inculcation" authority than government-as-regulator? Put another way, the question is whether first amendment principles, or their application, should change within the university setting. The Stanford policy implies that the answer is usually "no." It does not refashion the first amendment for campus life. This approach does not preclude university officials from maintaining that, in some areas of the campus such as dormitories or classrooms, there is a greater need to regulate conduct, including speech, than on the campus mall. This conclusion would hold true, however, regardless of whether the same functions were conducted in an off-campus location. Moreover, the analysis has nothing to do with the university's mission, in either a content- or a viewpoint-specific sense. Rather, the analysis addresses order, safety, and the extent to which speech can be so disruptive of a legitimate governmental activity that the activity cannot be performed. The Stanford policy thus corresponds with a Tinker-type n199 approach to speech on campus, under which only speech that threatens to cause substantial and material disruption of school functioning can be suppressed. This risk will be greater, of course, in the more controlled and function-specific setting of the classroom or the dormitory. But this basis for suppression is [*259] quite different than one of "inculcating fundamental values necessary to the maintenance of a democratic political system" n200 or inculcating "'the habits and manners of civility.'" n201

The significance of avoiding "inculcation" or "educational mission" talk in defending hate speech regulation on campus is greater than some commentators may appreciate. These are open-ended words, which educational administrators at all levels of education often invoke as reasons to chill student, faculty, or other staff expression. Indeed, despite the widespread assumption that freedom of speech and academic freedom are protected vigorously on college campuses, the actual pattern of freedom of speech enforcement for public employees reveals that they enjoy quite limited protection -- even in educational settings. n202 Moreover, in a number of recent decisions, the Supreme Court has retreated from the broad protection of student expression and relied heavily on the inculcation role and the importance of assimilating students into cultural norms of civility and decency. n203 Although these cases dealt with high school students, they betray the seductive nature of words like "inculcation," "citizenship," "civility," and "decency." Even if inculcation is an indispensable function of elementary and high school education, it should be rejected as an inappropriate function of colleges and universities. n204

When strong civil rights theorists, and some accommodationists, discuss the question of the general, value-positing authority of universities, they tend to be more microscopic than telescopic. They invoke only Brown v. Board of Education, n205 rather than the general case law that deals with the right of government-as-educator to inculcate values. n206 As such, they likely mean to encourage inculcation only of the specific value of equality, not the more general values of civility, decency, or even, necessarily, democracy. That is, they favor hate speech regulation as an [*260] extension of antidiscrimination law, not as a revived, expanded form of in loco parentis authority for educators. Their antidiscrimination purpose, however, should be expressed more narrowly than some of these theorists have stated it, lest their argument for campus hate speech regulation be misread as an argument for expansive power of school administrators to establish civility rules for the special community of the university.

In my view, any wide-ranging claim that public colleges and universities should be able to inculcate values through disciplinary measures misstates the proper role of these public institutions. I believe that the more compelling argument is that public schools at this level should not seek to inculcate values, at least not by regulating student expression. n207

My reasons for rejecting any effort to impose the general value of civility on campus through speech regulation are based on traditional first amendment assumptions and the features of college campuses that apply to those assumptions. The traditional first amendment remedy for bad speech -- counterspeech -- likely stands a better chance of succeeding on a college campus than in almost any other public or private setting. Campuses are relatively bounded communities. n208 They are geographically contained, with various institutional means of implementing and shaping dialogue. College students are among the more verbally adept members of the population and are poised between youth and maturity. They thus should be relatively responsive to dialogic appeals. Moreover, school authorities and other college community members have ample opportunities to influence the attitudes of this population other than by punishment. School officials can organize symposia, sponsor speakers, or otherwise condemn racist attitudes without expelling the students who hold such attitudes. Consequently, the conditions for critical dialogue are, or should be, particularly well met on a college campus. If counterspeech is no remedy here, then it likely is no remedy anywhere, a conclusion I am unwilling to embrace.

I therefore side with those who would rely on the counterspeech remedy, except in the egregious situations that fall within [*261] the Stanford policy. That is, hate speech on campuses should be subject to essentially the same limitations as it is beyond campus borders. This includes, however, the accommodations that make sense in various campus locations, based on the need for order and security -- but not on a desire for ideological conformity or for the preservation of a distinct university culture. In other words, I endorse the Stanford policy not because it constrains seriously harmful speech on a campus, but because it constrains seriously harmful speech.

A fifth, related advantage of the Stanford policy is that it also eschews reliance on the Court's often incoherent and analytically dubious public forum doctrine, in which the Court carves up public space and assigns degrees of expressive freedom on the basis of whether it declared the space a "public forum," "limited public forum," or "nonpublic forum." n209 This case law is subject to serious criticism, though the Court seems determined not to rethink it. n210 A full discussion of the Court's public forum doctrine is beyond the scope of this Essay. Nevertheless, the most worrisome aspect of the doctrine is that the Court has used it principally to constrict, rather than to expand, expressive freedom. In effect, the Court's zoning analysis affords government authority over so-called nonpublic forums with no meaningful judicial oversight. The result is an overbroad, often senseless, cordoning off of public space from the first amendment. Although sound, geographically specific reasons exist for speech rights to rise or fall, the Court's public forum doctrine is an inadequate tool for expressing those reasons. Until the Court adopts a more sensitive analysis of the relevance of location to speech, reformers do well to avoid reliance on this analysis.

The drafters of the Stanford policy