William Brennan
William Brennan; drawing by David Levine


The United States stands alone, even among democracies, in the extraordinary degree to which its constitution protects freedom of speech and of the press, and the Supreme Court’s great 1964 decision in New York Times v. Sullivan is a central element in that constitutional scheme of protection.1 The Constitution’s First Amendment provides that government may “make no law…abridging the freedom of speech, or of the press….” In its Sullivan decision, the Court said that it follows that a public official cannot win a libel verdict against the press unless he proves not only that some statement it made about him was false and damaging, but that it made that statement with “actual malice”—that its journalists were not just careless or negligent in researching their story, but published it either knowing that it was false or in “reckless disregard” of whether it was false or not. The decision imposed that strong burden of proof only on public officials; it left private individuals free to recover damages according to state law, which traditionally allows plaintiffs to win who prove only that statements about them are false and damaging.

The Court’s decision freed the press to investigate and report news, without the “chilling” fear that a jury might seize on some factual mistake or some journalistic lapse to award a libel verdict that would bankrupt the publisher. The Sullivan rule has made the American press much less cautious in criticizing public officials than journalists tend to be in Britain, for example, where public figures commonly sue newspapers and often win large verdicts against them.2 It is doubtful whether the Watergate investigation, or similar exposés, would have been possible if the Court had not adopted something like the Sullivan rule. But as Anthony Lewis makes plain in Make No Law, his fascinating book about the case, the decision had even wider importance, because Justice Brennan, in his opinion for the Court, redefined the fundamental premises of the First Amendment in terms that affected not only libel but First Amendment law much more generally. Though I shall argue, later, that this redefinition was not as successful, in retrospect, as it might have been, Brennan’s opinion is the modern foundation of the American law of free speech.

On March 29, 1960, The New York Times published a full-page advertisement titled Heed Their Rising Voices, which described the treatment of protesting black schoolchildren by the Alabama police. The advertisement contained some mistakes of fact. It said that students in Montgomery had been expelled from school after singing “My Country ‘Tis of Thee” on the state capitol steps, though they were actually expelled for a sit-in in the courthouse grill, and that the students had been locked out of their lunchroom to “starve them into submission,” which was apparently not true. L.B. Sullivan, a Montgomery city commissioner in charge of the police, claimed that the advertisement would be understood to be critical of him, though he was not mentioned in it, and would harm his reputation. He sued the Times in an Alabama court. After a trial in which the judge ordered segregated seating in the courtroom, and praised the “white man’s justice” brought to the country by the “Anglo-Saxon race,” an all-white jury, whose names and photographs had been published in the local paper, agreed that Sullivan had indeed been libeled and awarded him $500,000 in compensatory and punitive damages. The Times appealed, finally, to the Supreme Court.

Had the verdict stood, the Times would have been seriously damaged, and few papers of national circulation would have dared to publish anything about race that a southern jury might be persuaded to think false and libelous. The Supreme Court was therefore presumably anxious somehow to overrule the Alabama decision. But the legal background was inauspicious.

Lewis traces the constitutional history of free speech in America from the adoption of the First Amendment in the eighteenth century to the eve of the decision in Sullivan. For most of that period, the amendment was thought to have established only a very limited principle, and therefore to offer citizens only very limited protection. William Blackstone, the eighteenth-century English jurist whom American lawyers treated as the oracle of the common law, had declared that the common-law right of free speech was a right only against what he called “previous restraint.” He said government must not prevent citizens from publishing what they wished, but was free to punish them after publication if what they had published was offensive or dangerous. That was the traditional English view of free speech: even John Milton, who had campaigned ferociously against prior restraint in his famous essay Areopagitica, insisted that speech disrespectful to the Church, once published, could be punished by “fire and the executioner.”

The American federalists understood the First Amendment in the same way. In 1798, they adopted the Sedition Act, which made it a crime to publish intentionally “false, scandalous and malicious” reports about members of Congress or the president. Though Madison thought the Sedition Act violated the First Amendment, and Jefferson pardoned all those who had been convicted under it when he became president, the view that only “previous restraint” was impermissible remained the dominant interpretation of the First Amendment for over a century. Oliver Wendell Holmes, whose famous dissents later helped to bury that view forever, embraced it in 1907, when he upheld the contempt conviction of an editor who had criticized a judge. Holmes said that the main purpose of the First Amendment was to prohibit prior restraints, and he added that even true statements could be punished if they were harmful to the judicial process.


By World War I, however, some judges and scholars had adopted Madison’s different view, at least partly in reaction to a wave of prosecutions under the 1917 Espionage Act, which made it a crime to “attempt to cause…refusal of duty in the military or naval forces….” Lewis describes that development with reverent excitement. The early heroes of his book are Learned Hand, then a federal district court judge, who wrote a brilliant (though immediately overruled) opinion in the Masses case3 arguing that the First Amendment barred prosecuting a magazine whose cartoons ridiculed the war and the draft, and Zechariah Chafee, a Harvard Law School professor who wrote an influential law review article arguing that the amendment was intended to abolish all political censorship except direct incitement to illegal acts.

Though Holmes, whose skepticism made him reluctant to overturn any legislative decision, was slower to be converted, he was a lion once he was. His great dissent in the Abrams case, which declared that the Constitution commits us to an “experiment” based on the assumption that “the best test of truth is the power of the thought to get itself accepted in the competition of the market,” became one of the two classic endorsements of free speech before Sullivan.4 The other was Louis Brandeis’s careful, moving, and optimistic opinion in the Whitney case, dissenting from the Court’s refusal to overrule the conviction of Anita Whitney for supporting the Wobblies.5

By the 1960s, the great Holmes and Brandeis dissents had become orthodoxy. The old view that the First Amendment condemned only prior restraint had been replaced by the very different view summarized in Holmes’s famous formula: that government could punish political speech only when that speech posed a “clear and present danger” to society. But the Supreme Court had been careful to say, throughout this revolutionary period, that not all speech benefited from that protection. In Chaplinsky v. New Hampshire, for example, in which the Court said that the First Amendment did not protect “fighting words” which provoked immediate violence, it added that the First Amendment also did not apply to obscenity or private suits for libel.6

The latter exception seemed especially secure. The historical purpose of libel law was not to censor or punish the expression of opinion, but to allow offended citizens to vindicate their reputations. Such suits were governed by state law, and state courts, not the Supreme Court, were the final judge of what their state’s law was and how it should be applied. But the Alabama jury verdict showed that private libel suits could be used to restrict freedom of the press on crucial political matters, and Herbert Wechsler, a distinguished professor at the Columbia Law School whom the Times had hired to brief and argue its appeal to the Court, decided to make the revolutionary claim that the First Amendment did apply to state libel law, after all. The Court unanimously accepted that claim,7 Brennan elaborated it in his landmark opinion, and the Sullivan rule was born.


Lewis has himself created the genre in which he writes. Early in his career as a journalist he spent a year at the Harvard Law School as a Nieman Fellow, and then covered Supreme Court decisions, including the Sullivan case, for The New York Times, earning his second Pulitzer prize and raising the standard of Supreme Court reporting to a new level of legal sophistication. In 1964 he wrote Gideon’s Trumpet, an account of Gideon v. Wainwright, the case in which the Supreme Court established the constitutional right of poor people accused of a felony to a court-appointed lawyer.

Make No Law is an even better book. Freedom of speech and press is Lewis’s special constitutional concern. He taught the standard course in those subjects at the Harvard Law School for many years, and teaches it now at Columbia Law School as James Madison Visiting Professor. He has written several important law review articles on the subject. His account of the craftsmanship of a complex judicial opinion, and of the unique process through which one justice gathers others under a collective opinion all can sign, is itself a contribution to constitutional jurisprudence. Lewis has a journalist’s and a historian’s grasp of the subject, moreover, as well as a lawyer’s, and his prose is lucid, confident, and dramatic. Make No Law is exciting history and it is brilliantly told.


Sullivan became a landmark case not just because it revised the constitutional law of libel, but because Brennan’s language and images came to dominate the whole body of First Amendment law. Yet his opinion, for all its nobility, did not set out a complete intellectual basis for free speech law. In order to explain why it was incomplete, I must describe a controversial issue of constitutional theory.

The First Amendment, like the other great clauses of the Bill of Rights, is very abstract. It cannot be applied to concrete cases except by assigning some overall point or purpose to the Amendment’s abstract guarantee of “freedom of speech or of the press.” That is not just a matter of asking what the statesmen who drafted, debated, and adopted the First Amendment thought their clauses would accomplish. Contemporary lawyers and judges must try to find a political justification of the First Amendment that fits most past constitutional practice, including past decisions of the Supreme Court, but also provides a compelling reason why we should grant freedom of speech such a special and privileged place among our liberties.

The old Blackstonian explanation that appealed to many of the First Amendment’s framers—that the First Amendment was designed only to protect publication from prior restraint—is now obsolete. What new explanation, which accounts for the vastly greater protection the amendment is now understood to provide, should take its place? That is a central question, because their understanding of the point of protecting free speech will guide judges’ decisions in difficult and controversial cases about, for example, how far the right of free speech applies to nonpolitical speech, like art or commercial advertising or pornography, or how far that right is consistent with legal limitations on political campaign expenditures, or whether the First Amendment protects racist or sexist speech.

Constitutional lawyers and scholars have proposed many different justifications for the free speech and press clauses. Most of them fall into one or the other of two main groups; however. The first treats free speech as important instrumentally, that is, not because people have any intrinsic moral right to say what they wish, but because allowing them to do so will produce good effects for the rest of us. Free speech is said to be important, for example, because, as Holmes declared in his Abrams dissent, politics is more likely to discover truth and eliminate error, or to produce good rather than bad policies, if political discussion is free and uninhibited. Or for the reason Madison emphasized: that free speech helps to protect the power of the people to govern themselves. Or for the more common-sense reason that government is less likely to become corrupt if it lacks the power to punish criticism. According to these various instrumental views, America’s special commitment to free speech is based on a national endorsement of a strategy, a collective bet that free speech will do us more good than harm over the long run.

The second kind of justification of free speech supposes that freedom of speech is valuable, not just in virtue of the consequences it has, but because it is an essential and “constitutive” feature of a just political society that government treat all its adult members, except those who are incompetent, as responsible moral agents. That requirement has two dimensions. First, morally responsible people insist on making up their own minds about what is good or bad in life or in politics, or what is true and false in matters of justice or faith. Government insults its citizens, and denies their moral responsibility, when it decrees that they cannot be trusted to hear opinions that might persuade them to dangerous or offensive convictions. We retain our dignity, as individuals, only by insisting that no one—no official and no majority—has the right to withhold opinion from us on the ground that we are not fit to hear and consider it.

For many people moral responsibility has another, more active, aspect as well: a responsibility not only to form convictions of one’s own, but to express these to others, out of respect and concern for them, and out of a compelling desire that truth be known, justice served, and the good secured. Government frustrates and denies that aspect of moral personality when it disqualifies some people from exercising these responsibilities on the ground that their convictions make them unworthy participants. So long as government exercises political dominion over a person, and demands political obedience from him, it may not deny him either of these two attributes of moral responsibility, no matter how hateful the opinions he wishes to consider or propagate, any more than it may deny him an equal vote. If it does, it forfeits a substantial ground of its claim to legitimate power over him. The wrong is just as great when government forbids the expression of some social attitude or taste as when it censors explicitly political speech; citizens have as much right to contribute to the formation of the moral or aesthetic climate as they do to participate in politics.

Of course, the instrumental and constitutive justifications of free speech are not mutually exclusive.8 John Stuart Mill endorsed both of them in On Liberty. So did Brandeis in his remarkably insightful and comprehensive dissent in Whitney: he said that “those who won our independence believed that the final end of the state was to make men free to develop their faculties” and that “free speech is valuable both as an end and as a means,” which is a classic endorsement of the constitutive view.9 Brandeis was right that both kinds of justification are needed in order fully to explain First Amendment law; it is hardly surprising that so complex and fundamental a constitutional right as the right of free speech should reflect a variety of overlapping justifications.10

The two kinds of justification are moreover similar in many ways. Neither claims that freedom of speech is absolute; both allow that the values they cite may be overridden in special cases: in deciding, for example, how far military information may be censored. But the two justifications are nevertheless crucially different, because the instrumental justification is both more fragile and more limited. It is more fragile because, as we shall see, there are circumstances in which the strategic goals it appeals to might well be thought to argue for restricting rather than protecting speech. It is more limited because, while the constitutive justification extends, in principle, to all aspects of speech or reflection in which moral responsibility demands independence, the instrumental one, at least in its most popular versions, concentrates mainly on the protection of political speech.

If the point of freedom of speech is only to ensure that democracy works well—that people have the information they need in order to vote properly, or to protect democracy from usurping officials, or to ensure that government is not corrupt or incompetent—then free speech is much less important in matters of art or social or personal decisions. The First Amendment then protects sexually explicit literature, for example, only on the strained and easily resisted assumption that people need to read such literature in order to vote intelligently in national or local elections. Indeed, some scholars who accept the instrumental view as the exclusive justification of free speech have argued, as Robert Bork did, that the First Amendment protects nothing but plainly political speech, and does not extend to art or literature or science at all.11 Even those who reject that view, on the ground that literature and science can sometimes bear on politics, nevertheless insist that the main burden of the First Amendment is the protection of political speech, and that any protection the Amendment offers for other kinds of discourse is derivative from that principal function.

Brennan seemed to rely almost exclusively on the instrumental justification in his opinion for the Court in Sullivan. He limited First Amendment protection to cases involving libel of “public officials” rather than extending protection to all libel defendants. He quoted Madison’s instrumental argument that free speech is necessary in order to make the people rulers of the government rather than the other way around. He quoted passages from earlier Supreme Court decisions emphasizing the different instrumental argument Holmes had made in his Abrams dissent in which he said that truth emerges in a free market of ideas. He quoted, for example, Learned Hand’s endorsement of the same instrumental claim: “[The First Amendment] presupposes that right conclusions are more likely to be gathered out of a multitude of tongues, than through any kind of authoritative selection. To many this is, and always will be folly; but we have staked our all on it.”12 Only at one point did Brennan suggest a constitutive justification for free speech. He spoke of “the citizencritic” of government; he said, “It is as much his duty to criticize as it is the official’s duty to administer,” and cited Brandeis’s opinion in Whitney, which, as I said, recognized that free speech is an end as well as a means. But Brennan limited even this isolated suggestion of a constitutive justification to the political context.13

He was not just following a rhetorical tradition in endorsing the instrumental view. As Lewis demonstrates, Brennan was anxious to make his decision seem as little radical as possible—perhaps he could not otherwise have collected five other justices behind his opinion—and he therefore wanted to overrule the traditional exemption of libel actions from the First Amendment only to the degree absolutely necessary to prevent states from using libel to muzzle political criticism. The instrumental justification was very well suited to that purpose, because it seems to explain why it is particularly important to protect speech that is critical of public officials. The goal of helping the marketplace of ideas generate the wisest public choice of officials and policies is particularly badly served when criticism of officials is driven from that market.

In retrospect, however, Brennan’s near exclusive reliance on the instrumental justification, and his emphasis on the special role of political speech, seems regrettable even if it was necessary to collect a majority, because it may, unwittingly, have reinforced the popular but dangerous assumption that that is all there is to the First Amendment, and that the constitutive justification is either misplaced or unnecessary. In fact, relying exclusively on the instrumental justification is dangerous for free speech in ways that have already begun to be realized, and that may grow more serious now that Brennan and Thurgood Marshall, two of the most devoted defenders of free speech in the Court’s history; have been replaced by David Souter and Clarence Thomas.

We should start, in considering that danger, by noticing that the Madisonian version of the instrumental justification, on which Brennan particularly relied, cannot provide an intellectually acceptable justification even for the First Amendment’s political core. Madison’s argument that free speech is necessary if the people are to be in charge of their own government does explain why government must not be allowed to practice clandestine censorship which the people would reject if they were aware of it. But that argument does not explain why the majority of people should not be allowed to impose censorship that it approves and wants. A referendum might well reveal, for example, that a majority of Americans would prefer government to have the power to censor what it deems to be politically and diplomatically sensitive material, such as the Pentagon Papers. If so, then the Court’s obviously correct decision that government does not have that power14 can hardly be justified by Madison’s instrumental argument, except on the most implausibly paternalistic grounds. The great expansion of First Amendment protection in the decades after World War I plainly contracted, rather than expanded, the majority’s power to have the form of government it itself wants.

Some of that expanded protection can of course be justified on the different instrumental argument made by Hand and Holmes: that the truth about political issues is more likely to emerge if no idea is excluded from the discussion. It is certainly plausible that the public will make more intelligent decisions about race and civil rights if newspapers are free to write about these matters without fear of libel suits, and better decisions about war and peace if newspapers cannot be stopped from publishing documents like the Pentagon Papers.

But even this form of the instrumental argument cannot justify some of the most important of the federal court decisions expanding First Amendment protection in recent decades, including the Supreme Court’s decision, in Brandenburg v. Ohio, that states may not punish someone who says, wearing a hood at a Ku Klux Klan rally, that “the nigger should be returned to Africa, the Jew returned to Israel,”15 and the Seventh Circuit’s decision that a small band of neo-Nazis could not be prevented from marching with swastikas in Skokie, Illinois, where many Holocaust survivors lived. 16 Is our electorate really in a better position to choose its leaders or its policies because it permits speech of that kind? Would we be in a worse position to sift truth from falsity—would the marketplace of ideas be less efficient—if klansmen or Nazis or sexist bigots were silent?

It might be said that we cannot trust legislators or judges to draw distinctions between valuable and worthless political comment, so that in order to protect serious newspapers discussing serious issues we must also protect klansmen and Nazis spreading hate and causing pain. But that slipperyslope argument ignores the ability of lawyers to draw difficult distinctions here as they do in every other part of the law. If the Supreme Court can distinguish political speech from commercial speech, which it has decided enjoys much weaker constitutional protection, then it could also distinguish racist or sexist speech from other forms of political comment. It could uphold a statute carefully drafted to outlaw only speech that insults people on grounds of race, religion, or gender, in the manner of the British Race Relations Act, for example.

I emphasize this point not, of course, to recommend such a course, but to show that the instrumental justification does not offer much genuine protection against a statute of that character. In fact, the Supreme Court will soon rule on just such a statute. Last December, it heard oral argument in R.A.V. v. St. Paul and will presumably announce its decision sometime this spring. The City of St. Paul adopted an ordinance prohibiting display of a symbol that can be expected to cause “anger, alarm or resentment in others” on the basis of their race, religion, or sex, and providing a ninety-day jail sentence for that offense. Robert Viktora was prosecuted, under that ordinance, for burning a cross on a black family’s lawn. Of course, burning a cross on someone else’s lawn is forbidden by ordinary criminal law, and Viktora will be tried for that ordinary crime even if the Supreme Court decides that he cannot be punished under the special ordinance. The Viktora case raises the question whether a state may constitutionally make an assault a special crime, carrying a larger sentence, because it is intended to express a conviction the community disapproves of. The Court’s decision will undoubtedly have repercussions for the constitutionality of the regulations that many state universities, which are subject to the First Amendment, have recently adopted forbidding speech that expresses racial or sexual hatred or bias.17

It is very important that the Supreme Court confirm that the First Amendment protects even such speech; that it protects, as Holmes said, even speech we loathe. That is crucial for the reason that the constitutive justification of free speech emphasizes; because we are a liberal society committed to individual moral responsibility, and any censorship on grounds of content is inconsistent with that commitment. The instrumental arguments Brennan relied on in Sullivan are now being widely used, however, not to support but to undermine that view of liberal society. In a recent defense of campus constraints on “politically incorrect” speech, for example, Stanley Fish insisted, “Speech, in short, is never and could not be an independent value, but is always asserted against a background of some assumed conception of the good to which it must yield in the event of conflict.” Fish rejects the very possibility of what I called a constitutive defense of free speech; he insists that any defense must be instrumental and that censoring politically incorrect speech will serve the instrumental purpose better than freedom will.18

Catherine MacKinnon, Frank Michelman, and others have offered a similar argument for censoring pornography and other material offensive to women. They say that since women are more effective participants in the political process when they are not insulted by offensive material, the instrumental goal of effective democracy is actually better served by invading than protecting freedom of speech. They suggest, for example, that the ordinance Indianapolis adopted in response to a feminist campaign, which prohibited, among other kinds of literature, materials that “present women as enjoying pain or humiliation or rape,” would have improved rather than compromised democracy, because such literature “silences” women and so decreases their voice and role in democratic politics. The Seventh Circuit Court of Appeals, in an opinion by Judge Frank Easterbrook which I have discussed in these pages,19 rejected that argument, and held the statute unconstitutional because it outlawed not obscene publications generally, but just those promoting a particular idea or attitude. Easterbrook tacitly relied on the constitutive rather than the instrumental justification for free speech, and the Supreme Court can honestly declare the St. Paul ordinance invalid, as it should, only if it too recognizes that justification as well as repeating the old instrumental rhetoric.

An instrumental argument is also used to support bills pending in several state legislatures, and a federal bill introduced by Republican Senator Mitch McConnell of Kentucky now being considered by the Senate Judiciary Committee, which would allow a victim of sexual assault to sue a pornographic film or video’s producer or distributor for damages if she claimed the cause of the crime was that her attackers had watched that material.

The parallel with the Alabama libel law that Sullivan held unconstitutional is striking. If such antipornography laws are adopted, either federally or in particular states, juries in civil cases will be able to destroy a publisher or distributor by deciding that a rapist had watched a particular video, that that video is covered by the statute in question, and that it had incited his crime. Jurors who understandably despise violent pornography might well accept such a claim in spite of the fact that no respectable study or evidence has shown any causal link between pornography and actual violence.20 Since the trial would be a civil action for damages, the ordinary protections of the criminal law process would not apply. A rapist might even cooperate by testifying that pornography did cause his crime: criminals have often claimed, as if it were a kind of excuse, that their acts were caused by something they read or saw.21 Video shops would become extremely cautious about the films they stocked; they would be loath, for example, to stock The Accused, a much-praised film about gang rape, which would fall under the descriptions of pornography in some of the proposed laws.22 As Leanne Katz, who has formed a feminist group to fight the legislation, has said, the idea “that there should be a legal cause of action for what is ’caused’ by exposure to ideas, is a truly frightening prospect.”23

In a recent decision, the Supreme Court of Canada accepted a different instrumental argument for upholding a statute censoring certain forms of pornography.24 The Canadian Charter of Rights and Freedoms protects freedom of expression, though with qualifications the First Amendment does not recognize. The Canadian Court conceded that the effect of its ruling was to narrow that constitutional protection, but said that “the proliferation of materials which seriously offend the values fundamental to our society is a substantial concern which justifies restricting the otherwise full exercise of the freedom of expression.” That is an amazing statement. It is the central, defining, premise of freedom of speech that the offensiveness of ideas, or the challenge they offer to traditional ideas, cannot be a valid reason for censorship; once that premise is abandoned it is difficult to see what free speech means. The Court added that some sexually explicit material harms women because “materials portraying women as a class as objects for sexual exploitation and abuse have a negative impact on the individual’s sense of self-worth and acceptance.” But that kind of harm is so close to mere offensiveness that it cannot count, by itself, as a valid reason for censorship either. Every powerful and controversial idea has a potential negative impact on someone’s self-esteem. The Canadian Court, presumably, would not uphold a ban on nonpornographic literature whose purpose was explicitly to deny the equal worth of women, no matter how persuasive or effective that objectionable literature might be.25

These trends are ominous for liberty and for democracy. If Brennan had given a more prominent place to the constitutive justification in his restatement of the First Amendment premises in Sullivan, it would now be easier for American courts to reject the arguments that appealed to the Canadian Supreme Court, and to hold statutes like the St. Paul ordinance and laws providing tort actions against video distributors unconstitutional. The Supreme Court’s more general treatment of sexually explicit literature might also have been different if he had done so. The Court has several times declared that obscenity falls outside the protection of the First Amendment altogether, on the instrumental ground that obscenity has no “redeeming social value.” As the great First Amendment scholar Harry Kalven pointed out long ago, it strains common sense to think that a society in which hard-core pornography is freely available is better placed to discover the truth about anything for that reason.26 But it has proved enormously difficult for the Court to distinguish obscenity from sexually explicit material that does have at least some redeeming value. The Court has changed its mind about the ground of distinction so often and produced so many unworkable tests that Justice Stewart’s frank declaration that he couldn’t define obscenity but knew it when he saw it became the most-quoted judicial pronouncement on the issue.27

Though Brennan had himself declared, in 1957, that the First Amendment did not protect obscene speech or publication,28 he changed his mind in 1973, and declared in a dissenting opinion that, “in the absence of distribution to minors or obtrusive exposure to unconsenting adults,” the Constitution prohibits government from suppressing any materials as obscene.29 He tried to reconcile this view with the instrumental justification of free speech he had earlier endorsed by emphasizing the “institutional” difficulties courts face in distinguishing obscene from valuable work. But this, once again, was strained, and failed to persuade the Court. If, on the other hand, past decisions, and particularly his own Sullivan decision, had more clearly recognized the constitutive justification of free speech, he would have had a more natural and persuasive argument for his change of view. It is obviously inconsistent with respecting citizens as responsible moral agents to dictate what they can read on the basis of some official judgment about what will improve or destroy their characters, or what would cause them to have incorrect views about social matters.30

The worst and most threatening consequence of ignoring the constitutive justification is undoubtedly, however, the Supreme Court’s recent appalling decision in Rust v. Hill.31 The Court upheld the Reagan administration’s reinterpretation of a 1970 congressional statute, which provides funds to a variety of hospitals for “family planning” services but forbids that any of the funds be used for abortions, to mean that personnel working in such services may not even discuss that procedure.32 The administration’s order forbade doctors, nurses, or counselors even to respond to patients’ questions about abortion; it prohibited them from answering, for example, if a patient asked them where she could get abortion information, or whether abortions were legal.

In March of this year the Bush administration amended the regulations to allow doctors in federally financed facilities, though not nurses or other personnel, to discuss abortion. (The director of National Right to Life, an anti-abortion group, said he did not object to the amendment because “very little of the abortion counseling has been done by physicians.”)33 But the Court’s decision, which sustained the gag rule even when it did apply to doctors, is a dangerous precedent. The Court rejected the argument that the rule violated the First Amendment; it said that since doctors who wish to advise their patients about abortion are free to resign their positions in federally financed family planning services and seek jobs elsewhere, the government was not censoring anyone but simply dictating how money it supplied must be spent. The decision was very widely condemned as illogical and irresponsible, and we must hope that one day it will become as infamous an example of bad constitutional adjudication as the Lochner case, and other agreed Supreme Court mistakes. But the decision would not have been possible, I believe, except for the influence of the idea that the exclusive or cardinal purpose of the First Amendment is the instrumental purpose of ensuring a free flow of political expression.

No one thinks that the government could use its central role in helping finance health care in order to dictate the political opinions doctors working in hospitals supported by federal funds can express. Government could not allow such doctors to praise the government’s health care policies but forbid them to criticize those policies for example. From the standpoint of any competent constitutive justification of free speech, however, any distinction between such a plainly improper gag rule and the one the Court upheld is illegitimate. From that perspective, a doctor must be as free to give information he believes necessary for his patient’s health or well-being as he is to express political opinions. Requiring doctors, as a condition of their employment, not to give their patients medical information the patients request denigrates the moral responsibility of doctor and patient alike.


Suppose Brennan had set out to base his decision in the Sullivan case not on the relatively narrow ground he chose, which emphasized the instrumental justification of free speech, but on the broader ground I have been recommending, which emphasizes the constitutive justification as well. He might nevertheless have restricted the scope of a broader ruling to libel suits by public officials, on the ground that no more general scope is necessary to protect the moral independence of journalists and their audience. But he would at least have considered a much wider rule, which would have applied the current requirement that plaintiffs who are public officials or public figures must show the defendant was guilty of “actual malice” to all libel suits for damages by all plaintiffs against all defendants. That wider rule would protect a newspaper that had honestly but mistakenly published damaging information about a professor or a dancer or a businessman as well as about a sheriff or a general. Anyone who claimed that he had been libeled would be required to prove that the defendant was malicious, and not just careless or unlucky, in publishing the information the plaintiff claimed was false.

In fact Brennan himself recommended a rule very nearly that broad in a later case, in which he said that even a private individual should be required to satisfy the Sullivan test if he sued over a statement discussing a matter “of public or general concern.”34 As Marshall, who disagreed, pointed out, “all human events are arguably within the area of ‘public or general concern,’ ” so very few plaintiffs would not be covered by Brennan’s suggestion. In that same case, Justice Harlan recommended a rule extending the Sullivan limitations to private individuals who sought presumed or punitive damage for libel. Since plaintiffs seek such damages in almost all important cases, Harlan’s suggestion would have had much the same effect as Brennan’s. But neither suggestion was accepted by a majority of the Court, and there is no reason to think that a majority would ever have adopted the simpler and more straightforward rule I mentioned, which would apply the Sullivan test automatically to all plaintiffs. Nevertheless it is worth considering the merits of such a rule.

It would hardly be unfair to require a libel plaintiff to show at least that the press was in some way at fault in publishing what it did. That is the normal standard in almost all other civil actions for damages. I cannot make you pay on every occasion when you do something that injures me in some way—damage my property, for example. I must show that the injury was your fault, that it was the result of your not having acted, as lawyers say, reasonably in the circumstances. The law of libel has historically been an exception to this general principle: according to common law the plaintiff needed to prove only that what the defendant said was damaging to him, not that the defendant behaved unreasonably in saying it; indeed the burden fell on the defendant to prove that what he had said was true, not on the plaintiff to prove that it was false. That odd (and unjustifiable) exception for libel actions remains part of British law even now.

Plainly, a proper regard for freedom of speech requires at least that this exception be ended, for the benefit of all speakers or writers on any subject. The Sullivan rule, of course, goes further, because it demands that the plaintiffs it covers prove that the defendant was not just careless but published in bad faith or recklessly. But if the argument that justifies a more strenuous burden of proof in libel cases for public figures—the argument that such a rule is necessary if the press is to function fearlessly—is not unfair to them, it is difficult to see why extending that argument to ordinary plaintiffs would be unfair to them either. It is sometimes said that public figures have chosen to enter the kitchen and must not complain of the heat. But that argument has grown progressively weaker as the Court has expanded the range of the plaintiffs who fall under the Sullivan rule, from public officials to public figures who are defined in some other way. In any case, the argument begs the question: public officials have consented to a greater risk of being libeled only if there is some other reason why people in that position should have less protection than more private people.

There would have been no injustice, then, in a much more general revision of libel law. A more general revision would have saved the Court, moreover, from the hopeless task I just mentioned, the task it undertook, in the years after Sullivan, when it decided which plaintiffs were in fact subject to the higher burden of proof that the Sullivan rule created. As Lewis explains, Brennan’s original distinction between public officials and all other plaintiffs gave way, first, in the Court’s decision that Wally Butts—head coach of the University of Georgia who the Saturday Evening Post said gave away his team’s game plans to the Alabama coach, Bear Bryant, before the big game between the two schools—was a public “figure” though not an “official,” and was therefore subject to the Sullivan rule.35 Brennan’s distinction gave way more comprehensively in the later Gertz case, when the Court created an intermediate category of protection: it held that though a liberal lawyer who had been libeled by a John Birch Society publication was not a public figure, and so did not have to meet the strict Sullivan test of proving actual malice, he nevertheless had to prove that the defendant was at least negligent in publishing falsehoods about him, because the statements he complained of concerned a matter of political interest. 36

So the Court has found it difficult to make the various discriminations its rules now require, and its categories seem arbitrary from the perspective of the instrumental view of free speech they are supposed to reflect. Movie stars, for example, have been classified as public figures, and so must satisfy the actual malice standard when they sue tabloids for false reports about them, though, as Lewis points out, celebrity gossip hardly contributes to the efficiency of the political process in discovering truth or wisdom. Lewis would regret any general extension of the Sullivan rule to all libel plaintiffs, however, because he thinks that the Court would not be as zealous in protecting that rule if the rule had much more general application. But, as he recognizes, the significant extension of protection since the original decision has not so far caused any reduced enthusiasm for the rule. There is the important possibility, moreover, that a general extension would set in motion an even more radical reform of American libel law which, in the end, would benefit both libel plaintiffs and press freedom.

The Sullivan rule has not proved as effective a protection of the press’s freedom to report on politics as commentators initially expected. As Lewis says, celebrations stopped when lawyers and the press realized that the rule gave well-financed public-figure plaintiffs an opportunity to inflict great damage on the press by claiming that it was, indeed, malicious or reckless. The much-publicized suits by Ariel Sharon against Time and General William Westmoreland against CBS in 1984 illustrate the difficulty: though Time won a qualified jury victory, and Westmoreland finally abandoned his suit, both Time and CBS were seriously damaged by the publicity, cost, and dislocations of long trials in which their honesty and competence were the chief topics of investigation.37

In another long and expensive trial, Herbert v. Lando, CBS argued that it should not be required to produce time-consuming volumes of confidential reports, internal memoranda, and other material that the plaintiff said he needed to review in order to prove actual malice.38 The Supreme Court said that the press could not have it both ways: if the law required the plaintiff to prove that the defendant knew what it said was wrong, or that it published in disregard of its truth, then it would be unfair to allow the defendant to withhold any information the plaintiff might need to prove that. So the press is still in jeopardy of suffering great financial loss if it publishes what opponents with rich supporters are prepared to argue is false and malicious.

Lewis describes sensible proposals that several commentators have offered for revising libel law to help solve that problem, while also giving people reasonable protection against false and malicious statements.39 Under the present law libel plaintiffs seek to redeem their reputation by winning huge jury awards whose purpose is not to compensate them for any actual financial loss they may have suffered but to punish the press. Reformers hope to separate these two ingredients of a libel suit by allowing a plaintiff to sue, not for monetary reward, but simply for a judicial declaration that what the press said was false. Under one version of that proposal, someone who thinks he has been libeled must inform the publisher of facts which he believes show that what it published was false. Unless the publisher prints a sufficiently prominent correction, the plaintiff may sue for a judicial declaration of falsity, and an order requiring the defendant to report that declaration.40 Since there would be no prospect of damages, the truth or falsity of what was published would be the only issue, and no question of the defendant’s malice or negligence would be raised. The Sullivan rule would have no application. In most cases the trial would be expeditious and inexpensive, and an unfairly libeled plaintiff would be able to secure a judicial declaration that he had been unjustly attacked.

It is a further question how far such a declaration would help restore a plaintiff’s reputation. Sensational statements in one newspaper, even one with a reputation for inaccuracy, are widely copied in other media, and a judicial declaration that the statement was an error might not be as widely reported, even if the offending paper was itself required to report that decision. So even if states were to adopt a new form of legal action, allowing a plaintiff to sue just for a judicial declaration of falsehood, they might continue to allow a plaintiff to sue for damages for loss of reputation as well, if he so wished. But there would then be no reason not to apply the Sullivan rule to all damage claims, so that both public figures and private individuals would have the option of suing for a judicial declaration, in which case they need prove nothing except the falsity of the publication, or for damages, in which case they must show that the press published the falsehood knowingly or recklessly.41

Reform of that general character would be more likely, in fact, if the Court did insist that the same rules apply to everyone. So far no state has changed its libel law in the way the reformers recommend. In 1985, Charles Schumer, a congressman from Brooklyn, introduced a federal bill along the lines of these suggestions, but his bill failed. The Supreme Court cannot itself order a change of that complex character in state libel law, though it might perhaps go some distance toward it, by ruling, for example, that a state must accept a prominently printed correction as a full defense. Congress or the states would be more likely to make the change themselves, however, if the Court had declared that ordinary people as well as public figures must meet the Sullivan test in order to collect damages for libel. Private individuals would then be anxious to find some quicker and less expensive means of vindicating their reputations, and a suit for a judicial declaration would be a valuable alternative. No legal scheme can provide an ideal solution to the inevitable conflict between free speech and the protection of private reputation. Nor is there any reason to think that the Supreme Court will expand the Sullivan rule in the way I suggest. But a unified system which treats all plaintiffs and all defendants alike seems more in the interests of everyone—press, public, and private citizens—than the present set of complex and unstable rules.

Sullivan was a monumental decision, for all the reasons Lewis skillfully reports. Brennan’s decision freed the American press to play a more confident role in protecting democracy than the press plays anywhere else in the world. It does not detract from his achievement that the intellectual premises of his argument must now be expanded, in the face of very different threats to liberty from those he confronted in that case, and that the libel law he constructed, radical for its time, is now sufficiently well established to be simplified. Sullivan was a crucial battle in the defense of our first freedom. But now we have new battles to fight.

This Issue

June 11, 1992