Concept
Offensive Speech Stays Protected
Regulation of unprotected expression
Offensive Speech Stays Protected
Matal v. Tam states the premise of this section most directly. The government defended a ban on registering disparaging trademarks as a way to shield groups from demeaning messages, but the Court looked past the phrasing: the “unmistakable thrust” of the interest was preventing speech that expresses ideas that offend, and that idea, the Court said, “strikes at the heart of the First Amendment.”1 Speech that demeans on grounds such as race, ethnicity, gender, or religion is hateful, the Court conceded, yet the proudest boast of free speech doctrine is that it protects the freedom to express “the thought that we hate.”2 The case returns at the end of this section with its facts and holding. Here it supplies the premise.
The premise follows from how the unprotected categories work. Each class studied so far is defined by something the speech does: incitement produces imminent lawless action, fighting words provoke a fight, and true threats instill a reasonable fear of violence. Offense names no such harm. It is a listener’s reaction to what a message says, so if the state could punish speech because it offends, the listener’s displeasure would decide what may be said, and the government would be choosing which ideas are too distasteful to hear. Taste adds a second difficulty. Once officials begin purging vocabulary, nothing marks where to stop, and a word’s emotional force is part of what the speaker communicates, so removing the word removes part of the message.
“Hate speech” is a common label, but First Amendment law has no unprotected category by that name. Racist or bigoted expression loses protection only if it independently fits a recognized class such as incitement, true threats, or fighting words. Even the fighting-words doctrine depends largely on the context in which words are used, so no word automatically falls into it.3
Three settings test the premise. Cohen concerns a vulgar word whose offense lies in the manner of expression. Snyder concerns tort liability for emotional distress caused by speech on public issues. Matal concerns the denial of a government benefit because of a message. The limits that remain, such as obscenity, the captive audience, and fighting words, surface as each setting raises them.
Profanity and Cohen
Cohen v. California asks whether a state may punish a vulgar word because the word offends. Paul Robert Cohen walked through the corridor outside a municipal courtroom in the Los Angeles County Courthouse, with women and children present, wearing a jacket that read “Fuck the Draft.” He was convicted under a California statute forbidding anyone to maliciously and willfully disturb the peace of any neighborhood or person by “offensive conduct,” and he received thirty days’ imprisonment.4 He testified that he wore the jacket to tell the public how deeply he opposed the Vietnam War and the draft, and the record showed that neither he nor anyone responding to him committed or threatened violence.5 The case therefore presents a message with no menace attached. The state could point only to the word.
The Court reversed, and its first move shaped everything after it. Justice Harlan observed that the conviction rested on the asserted offensiveness of the words Cohen used, and that the only “conduct” the state sought to punish was the fact of communication, so the case concerned speech and not some separable act.6 The state could not punish Cohen for the idea his jacket expressed, which meant the conviction could stand, if at all, only as a regulation of the manner in which he spoke. A state that punishes manner alone must show that the manner falls within some recognized exception, and the opinion tests three candidates.
The second question does the work that separates Cohen from the fighting-words cases. The Court conceded that the four-letter word on the jacket “is not uncommonly employed in a personally provocative fashion,” yet held that here it was clearly not “directed to the person of the hearer.”11 The doctrine turns on circumstances and not on vocabulary, so the same word can be protected in one setting and a fighting word in another.12
Chaplinsky v. New Hampshire shows the other side of the line. In a heated argument with a city marshal, Chaplinsky called him a “God damned racketeer” and “a damned Fascist,” and he was convicted under a statute forbidding any offensive, derisive, or annoying word addressed to another person in a public place.13 The Court upheld the conviction and defined fighting words as those that “by their very utterance inflict injury or tend to incite an immediate breach of the peace.”14 That definition seems to focus on the words themselves, but the cases applying it ask whether the language was used in circumstances that were personally provocative and created an actual threat of imminent disorder.15 Since Chaplinsky the Court has seldom rested a decision on the fighting-words label, and it has generally found statutes defended on that rationale to be overbroad.16
The state’s fallback position was broader than any of the three categories. California could not defend the conviction on the theory that the jacket was inherently likely to provoke violence, the Court said, because that theory reflected at most an “undifferentiated fear or apprehension of disturbance,” and the state had shown no evidence that substantial numbers of citizens stood ready to strike out at whoever offended them with such language.17 That left the claim that a state, acting as guardian of public morality, may remove one offensive word from the public vocabulary. The Court answered by returning to the purpose of the guarantee, which it described as removing governmental restraints from the arena of public discussion and leaving the choice of which views to voice largely to each person, in the hope of a more capable citizenry and in the belief that no other approach would honor individual dignity and choice.18 The price is some verbal tumult, which the Court counted “not a sign of weakness but of strength.”19 Against that background Harlan gave three reasons for reversal, and each has its own mechanism.
The first reason is that the state’s principle has no stopping point. If the state may remove this four-letter word because it offends, nothing in the principle distinguishes it from any other offensive word, so officials would have to rank words by how distasteful they find them, with no standard for the ranking except their own taste.20 Harlan put the difficulty in a sentence that has outlived the case: one man’s vulgarity is another’s lyric, and the Constitution leaves matters of taste and style so largely to the individual precisely because governmental officials cannot make principled distinctions in this area.21
The second reason is that words carry feeling as well as information. Much linguistic expression conveys both ideas that can be explained with detachment and emotions that cannot otherwise be expressed, and words are often chosen as much for their emotive force as for their cognitive content.22 A jacket reading “The Draft Is Unfair” or “Resist the Draft” reports a position, while Cohen’s jacket reported the same position together with the intensity of the wearer’s feeling about it, which is a different message. A state that bans the word edits the message by forcing the speaker to deliver a cooler version than the one he chose. A rule that protected only the detachable content of speech would protect the less important half of what many speakers mean to say.
The third reason concerns what a word ban can be used for. The Court refused to assume that particular words can be forbidden without a substantial risk of suppressing ideas along with them, and it warned that governments might seize upon the censorship of particular words as a convenient guise for banning the expression of unpopular views.23 The mechanism is easy to trace. A ban on a word looks neutral because it names no viewpoint, yet it falls hardest on speakers whose message depends on that word, and an officer who dislikes the message can charge the vocabulary instead of the idea. The Court therefore held that, absent a more particularized and compelling reason, the state may not make the simple public display of this single four-letter expletive a criminal offense.24
The captive-audience question explains why the presence of women and children in the corridor did not change the result. The Court accepted that government may protect the privacy of the home from unwelcome views, but it said that shutting off speech solely to protect listeners depends on a showing that “substantial privacy interests are being invaded in an essentially intolerable manner,” because a broader view would let a majority silence dissidents as a matter of personal predilection.25 Viewers in the corridor could escape further bombardment of their sensibilities by averting their eyes, and a courthouse hallway carries nothing like the privacy interest of a person’s home.26 A homeowner whose evening is filled with a loudspeaker outside her window stands where Cohen’s viewers did not, since she cannot look away from a sound. The state also could not recast the conviction as protecting courthouse decorum, because the statute applied throughout California and nothing in it warned that certain speech would be forbidden in certain places.27
Cohen also trimmed Chaplinsky. Chaplinsky had listed the lewd and the profane among words outside the First Amendment, and Cohen made clear that profanity is at least sometimes protected, that fighting words must be directed to the person of the hearer and not addressed generally to the world at large, and that preventing psychic offense is not enough to justify punishing words said to wound by their very utterance.28 Later cases apply the directed-to-a-person requirement. In Texas v. Johnson the Court rejected a fighting-words defense of a flag-burning conviction because no reasonable onlooker would have taken a generalized protest against federal policy as a direct personal insult or an invitation to exchange fisticuffs.29 Vulgar language therefore stays protected unless some other limiting principle applies, and the last major section of this chapter takes up the one setting, broadcasting, where the Court has treated indecency differently.
Emotional Distress and Snyder
Snyder v. Phelps asks how far the First Amendment limits a state tort suit when the plaintiff’s injury was caused by speech. The Free Speech Clause can serve as a defense in such suits, including suits for intentional infliction of emotional distress, and the Court applied it to a jury verdict against protesters at a soldier’s funeral.31 Under Maryland law the plaintiff in that kind of suit must show that the defendant intentionally or recklessly engaged in extreme and outrageous conduct that caused severe emotional distress.32 The Court proceeded on the premise that the picketing satisfied the tort, so the only question before it was constitutional.33 No one suggested that the speech was obscene or fighting words, so ordinary First Amendment protection applied.34
Members of the Westboro Baptist Church, founded by Fred Phelps, picketed the funeral of Marine Lance Corporal Matthew Snyder, who was killed in Iraq. They stood on public land about 1,000 feet from the church, following police guidance, and for roughly 30 minutes before the service they peacefully displayed signs such as “Thank God for Dead Soldiers,” “Fags Doom Nations,” “America is Doomed,” “Priests Rape Boys,” and “You’re Going to Hell.”35 Matthew’s father saw only the tops of the signs as he drove to the funeral and learned what they said from a news broadcast that night.36 A jury found for Snyder on claims of intentional infliction of emotional distress, intrusion upon seclusion, and civil conspiracy, and it awarded $2.9 million in compensatory and $8 million in punitive damages. The district court reduced the punitive award to $2.1 million, and the Fourth Circuit reversed on First Amendment grounds.37 Chief Justice Roberts, writing for an 8 to 1 Court, affirmed, holding that the First Amendment shields Westboro from tort liability for its picketing in this case. The judgment under review was a civil damages award and not a criminal conviction, and the funeral setting did not turn the speech into unprotected conduct.38
Whether the First Amendment bars the verdict turns largely on whether the speech is of public or private concern, judged by all the circumstances of the case.39 The test tracks what each kind of speech puts at risk. Speech on public issues sits at the heart of First Amendment protection, since debate on those issues is meant to be uninhibited, robust, and wide open, and a threat of damages there pushes speakers toward self-censorship on exactly the subjects the Amendment most wants discussed.40 A suit over purely private speech threatens no free debate of public issues and creates no comparable pressure to stay silent, so First Amendment protections in that setting are often less rigorous.41 The inquiry therefore asks which kind of speech the tort suit would burden before it asks whether the defendant’s words were cruel.
Speech deals with a matter of public concern when it can fairly be considered as relating to any matter of political, social, or other concern to the community, or when it is a subject of legitimate news interest, meaning a subject of general interest and of value and concern to the public.42 The Court admitted that the boundaries of this test are not well defined, and it offered a method in place of a formula: examine the content, form, and context of the speech as revealed by the whole record, with no factor dispositive and with attention to what was said, where it was said, and how it was said.43 A reviewing court makes its own independent examination of that record, so a jury’s characterization of the speech does not settle the matter.44
Applied to Westboro’s signs, the content question was easy to answer. The placards addressed the political and moral conduct of the United States and its citizens, the fate of the Nation, homosexuality in the military, and scandals involving the Catholic clergy, which the Court called matters of public import even though the messages might fall short of refined social or political commentary.50 A few signs, such as “You’re Going to Hell,” could be read as aimed at the Snyders, yet the overall thrust and dominant theme of the demonstration spoke to broader public issues.51 The context did not change the answer. The church had been speaking on these subjects long before it learned of Matthew Snyder, and no earlier relationship or conflict suggested that its public message masked a private attack.52
The contrasting case is Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., in which a credit agency sent a report about one company to only five subscribers, who were bound not to pass it along. The Court treated the report as speech solely in the individual interest of the speaker and its specific business audience, so it concerned no public issue.53 The two cases differ in audience and subject. Westboro’s signs were designed to reach as broad a public as possible on questions the community debates, while the credit report was a private commercial message to a closed group.54
Justice Alito, dissenting alone, applied the same public-concern inquiry and reached the opposite result on the record. Because Westboro chose a private funeral over countless other venues, he reasoned, a reasonable bystander would have connected the signs to the deceased, and some signs would have been read as suggesting, falsely, that Matthew Snyder was gay. In his view the picketers specifically attacked a private figure because he was a Catholic and a Marine, and speech about a private person’s own conduct is not a public concern.55 He added that actionable speech should not be immunized because it is interspersed with protected speech, and that a public street near a funeral is no free-fire zone for verbal attacks.56 The majority answered that the dominant theme controls, and it observed that adjacency to a public street does not insulate speech from liability but heightens the concern that the speaker is communicating views to the public in a traditional public forum.57 The disagreement is over how to characterize the record, which is why the Court insists on independent review of the whole record in these cases.
The place and manner of the picketing reinforced the majority’s conclusion. Church members had the right to be where they were. They alerted local authorities, followed police guidance on where to stand, picketed about 1,000 feet from the church and out of sight of those inside, and engaged in no shouting, profanity, or violence.58 Any distress the signs caused turned on the content and viewpoint of the message and on no interference with the funeral, and a group of parishioners standing at the same spot with signs reading “God Bless America” would have faced no liability.59 That comparison shows why the verdict operated as a content-based restriction: the same conduct in the same place drew damages only because of what the signs said. Speech at a public place on a matter of public concern receives special protection and cannot be restricted simply because it is upsetting or arouses contempt.60
The jury had been told it could hold Westboro liable if it found the picketing outrageous, and the Court held that this standard cannot carry liability for speech of this kind.61 The rule grows out of Hustler Magazine v. Falwell, where a parody portrayed a nationally known minister’s first sexual experience as a drunken encounter with his mother in an outhouse. The Court held there that a public figure may not recover for intentional infliction of emotional distress from such a publication without showing a false statement of fact made with actual malice.62 Its reason was that outrageousness, applied to political and social discourse, lets a jury impose liability on the basis of its tastes or views, or its dislike of a particular expression.63 Snyder shows how that works at a funeral. The tort asks whether conduct is extreme and outrageous, and when the conduct is a message, a juror answering that question must react to the message. A jury is unlikely to be neutral toward content that offends it, so the standard risks becoming an instrument for suppressing vehement, caustic, and unpleasant expression. The Court called that risk unacceptable, because public debate must tolerate insulting and even outrageous speech to give the First Amendment breathing space.64 Falwell’s plaintiff was a public figure and Snyder’s was a private person, so the decisive fact in Snyder was the subject of the speech, and the plaintiff’s status did no work.65
Snyder also argued that he was a captive audience at his son’s funeral, which would have supported the intrusion-upon-seclusion verdict. The Court answered that the burden normally falls on the viewer to avert his eyes, and that government may shut off discourse solely to protect others from hearing it only on a showing that substantial privacy interests are being invaded in an essentially intolerable manner.66 The doctrine has been applied sparingly, to offensive mail delivered to a home and to picketing before or about a residence. Here Westboro stayed well away, Snyder saw only the tops of the signs, and nothing interfered with the service.67 The civil conspiracy claim fell with the two torts it rested on.68
The Chief Justice described the holding as narrow, limited by the particular facts: matters of public import, public property, a peaceful manner, and full compliance with local officials’ guidance, with speech planned to coincide with the funeral but not itself disrupting it.69 The opinion leaves content-neutral regulation of the setting available. The Court pointed to earlier decisions upholding a ban on picketing before or about a residence and a buffer zone around an abortion clinic entrance, and Maryland now bars picketing within 100 feet of a funeral, a rule Westboro’s picketing would have satisfied.70 To the extent such laws are content neutral, they raise very different questions from a tort verdict that depends on the message.71 Justice Breyer, concurring, stressed the same limits. A state can sometimes regulate picketing even on matters of public concern, the Court’s opinion does not imply that a state is always powerless to protect private individuals, and the picketing could not be seen or heard from the ceremony itself.72 Justice Alito replied that buffer laws are no substitute for the tort, since the attack on the Snyders complied with Maryland’s law.73
Distress claims built on messages aimed at one person raise a different question. In Counterman v. Colorado, a defendant sent a barrage of Facebook messages that suggested harm would befall C.W., who became anxious, stopped going out, and cancelled performances. Colorado charged him under a statute making it unlawful to “repeatedly communicate with another person” so as to cause “a reasonable person to suffer emotional distress,” and his defense was that the messages were not “true threats.”74 That question belongs to the true-threats doctrine, taken up in a later section, and not to the public-concern analysis of Snyder.
On an exam, a funeral, a grieving relative, and a large jury verdict invite the question whether the message was cruel enough to count as outrageous. Start earlier. Classify the speech by its content, form, and context. If it is a public concern delivered peacefully in a public place, its controversial or inappropriate character is irrelevant to that classification, and an outrageousness finding cannot sustain tort liability.75 The limits then come into view: a message about a private matter, a content-neutral location rule, or conduct that interferes with the funeral service itself would each be analyzed differently.76
Disparagement and Matal
In Matal v. Tam the Court decided whether the government may withhold a benefit from speech because the speech offends. Simon Tam, lead singer of the dance-rock band The Slants, whose members are Asian-American, applied to register the band’s name as a federal trademark. The name is a derogatory term for persons of Asian descent, and the band chose it in order to reclaim the word and drain its denigrating force.77 The Patent and Trademark Office refused under the Lanham Act’s disparagement clause, which bars registration of a mark that may disparage persons, living or dead, institutions, beliefs, or national symbols, or bring them into contempt or disrepute.78 The Court held that the clause violates the Free Speech Clause, because it offends the bedrock principle that speech may not be banned on the ground that it expresses ideas that offend. The previous subsection reached a related principle through a damages verdict. This case reaches it through the denial of a registration, a benefit the government confers. A benefit is not a penalty, so the Court had to explain why the principle still governs, and the explanation has three links: who is speaking, whether the clause discriminates by viewpoint despite applying to everyone, and whether any asserted interest can justify it.
The first link asks who is speaking, because the Free Speech Clause limits what the government may do to private expression and places no comparable limit on what the government says for itself. The government therefore defended the clause by arguing that registering a mark turns it into government speech, meaning a message the government itself conveys. The Court rejected the argument and held that trademarks are private speech.79 The reasoning follows the structure of a registry. A registration system records marks that private parties devise and submit, and if recording a claim adopted its content as the government’s own message, the government could silence any speaker simply by refusing to register the speaker’s mark.80 The same logic would reach other registries, copyright among them, so the Court declined to stretch the government-speech idea that far.81
A city’s permanent monuments in a public park show the opposite situation. In Pleasant Grove City v. Summum the city accepted certain monuments for its park and declined a rival group’s monument. Because the city selects the monuments it displays and presents them as its own expression, the refusal is the government choosing its own message and does not suppress anyone else’s, so the Free Speech Clause does not govern it. The Slants chose their own words and the government merely recorded them, while the city chose the monument and owns what it says. The question is who actually composes and controls the message, and the label the government attaches to its program does not answer it.
Two companion arguments failed for related reasons. The subsidy cases involved cash subsidies or their equivalent, such as funds for family planning services in Rust v. Sullivan and cash grants to artists in National Endowment for the Arts v. Finley, and the Patent and Trademark Office pays applicants nothing, since it charges fees to file and to maintain a registration.82 The government replied that registration confers valuable non-monetary benefits, but nearly every government service, from police protection to the adjudication of private lawsuits, costs public money, and that fact cannot turn every service into a subsidy.83 The proposed government-program doctrine, which merged the government-speech and subsidy cases, fared no better, because even in limited public forum cases that allow some content- and speaker-based restrictions, viewpoint discrimination is forbidden.84
The second link concerns viewpoint discrimination, which occurs when the government favors one side of a debate over another, and the government’s best reply was that the clause applies evenhandedly. The clause denies registration to any mark that is offensive to a substantial percentage of the members of any group, so it reaches marks that damn Democrats and Republicans or capitalists and socialists alike. The Court’s answer was that, in the sense relevant here, “[g]iving offense is a viewpoint.”85 Justice Gorsuch took no part, and the other eight Justices agreed that the clause is unconstitutional viewpoint discrimination, although they divided on how much of the opinion spoke for the Court.86 Justice Kennedy’s concurrence, joined by three other Justices, states the test in its plainest form: within the relevant subject category, has the government singled out a subset of messages for disfavor based on the views expressed?87 The clause defines its subject as persons, institutions, beliefs, and national symbols, and within that subject an applicant may register a positive or benign mark but not a derogatory one, which is the essence of viewpoint discrimination.88
Evenhandedness cannot cure that defect, and the reason is visible in what the government’s argument would permit. A subject first defined by content and then regulated by allowing only one sort of comment is not viewpoint neutral, and a law that forbids all sides to criticize their opponents is more viewpoint based, not less.89 Carried to its end, the government’s logic would make a rule neutral even if it let public officials be praised but not condemned.90 A registry that rewards positivity decides which arguments may reach the public in the form the speaker chose, and by mandating positivity the law might silence dissent and distort the marketplace of ideas.91
The government’s second neutrality argument was that the clause looks to the expected reaction of the applicant’s audience and ignores the applicant’s own views. That description fits the facts. Tam intended the name to celebrate what Asian-Americans contribute to the country, and the application was refused because the government expected the mark to offend at least some Asian-Americans.92 Tying a restriction to audience reaction does not shield it from the charge of viewpoint discrimination, because the danger is that the government removes certain ideas from public debate, and that danger grows when the audience may find an idea offensive at first hearing.93 A first reaction may prompt reflection and a more tolerant position, and a clause that filters marks by expected offense cuts that process short before it starts. Under such a test a bigot’s use of a slur and a band’s use of the same word to reclaim it can receive identical treatment, since the inquiry asks how a mark will land with listeners and cannot ask what the speaker means.
That power survives the decision. Justice Kennedy noted that the law may protect consumers and trademark owners when a mark is confusing or misleading, and that such considerations leave untouched the speech principles barring the viewpoint discrimination at issue.94 The usual exam error is to treat any refusal of registration as presumptively valid because registration is a government benefit. The correct analysis asks what criterion produced the refusal, and a criterion that turns on the offensiveness of the message cannot be defended as a neutral condition of the benefit.
The third link asks whether any interest can justify a viewpoint-based clause. The government argued that trademarks are commercial speech entitled only to relaxed scrutiny, and the Court did not decide that question, because the clause cannot survive even that standard, under which a restriction must serve a substantial interest and be narrowly drawn so that it extends only as far as the interest it serves.95 The first asserted interest, shielding underrepresented groups from demeaning messages, has as its unmistakable thrust an interest in preventing speech that expresses ideas that offend, and the Court said that idea strikes at the heart of the First Amendment, since the proudest boast of free speech doctrine is protecting the freedom to express “the thought that we hate.”96 The second interest, protecting the orderly flow of commerce, fails because the clause is not narrowly drawn to drive out marks that support invidious discrimination. It reaches “Down with racists,” “Down with sexists,” and “Down with homophobes” alike, which led the Court to call it “a happy-talk clause” and to ask whether commerce would be disrupted by a mark reading “Slavery is an evil institution.”97 The commercial label also would not have rescued the clause had the Court reached it, because viewpoint-based discrimination draws heightened scrutiny whatever the category, and commercial speech is no exception to the requirement of viewpoint neutrality.98
Two years later, in Iancu v. Brunetti, the Court applied the same reasoning to the Lanham Act’s bar on registering “immoral” or “scandalous” marks, which the Patent and Trademark Office had used to refuse a clothing brand’s mark.99 Justice Kagan found that the pair of terms, on its face, distinguishes between ideas aligned with conventional moral standards and ideas hostile to them, and the Court held that “a law disfavoring ‘ideas that offend’ discriminates based on viewpoint, in violation of the First Amendment.”100 Because both decisions treat offense as a forbidden criterion, Matal has potentially broad implications for efforts to restrict hate speech, which the next section takes up.101