Concept
Incitement to Lawless Action
Regulation of unprotected expression
Incitement to Lawless Action
Advocacy of force or law violation may be punished only when it is directed to inciting or producing imminent lawless action and is likely to incite or produce such action. That is the rule of Brandenburg v. Ohio, and this section builds toward it. Brandenburg, a Ku Klux Klan leader, had been convicted under Ohio's criminal syndicalism statute for “advocat[ing] . . . the duty, necessity, or propriety of crime, sabotage, violence, or unlawful methods of terrorism as a means of accomplishing industrial or political reform” and for voluntarily assembling with a group formed to teach those doctrines.1 The Court reversed. The statute failed to distinguish the abstract teaching of the moral propriety of force from preparing a group for violent action and steeling it to such action, and a statute that blurs that line sweeps in speech the Constitution has immunized from governmental control.2 That holding gives incitement its place among the unprotected categories, and it attaches two conditions that must both be met before speech may be punished.
The rule strikes a balance. Speech that urges others to break the law can be part of a crime, and government has a real interest in stopping it. Yet a strong presumption favors speech, and that presumption justifies protecting even advocacy of illegality unless a substantial likelihood of imminent harm exists. Justice Brandeis put the point this way: “even advocacy of [law] violation, however reprehensible morally, is not a justification for denying free speech where the advocacy falls short of incitement and there is nothing to indicate that the advocacy would be immediately acted on.” Advocacy of law violation, even civil disobedience, is also a powerful way of expressing a message. But the Court has never held such speech completely protected, so government is not confined to punishing the criminal acts themselves.3
The Court reached this line only after trying several others. It articulated the clear-and-present-danger test during World War I and the years just after. In the 1920s and 1930s it decided criminal syndicalism cases, which involved statutes that made it a crime to advocate overthrowing the government by force or violence, without invoking that test; it upheld the laws and their applications so long as the government's law and prosecution were reasonable.4 In the 1950s it recast the danger test as a risk formula that weighed the gravity of the evil against its likelihood. Since the late 1960s it has defined incitement narrowly in order to maximize protection of speech.5 Of the earlier approaches, only the reasonableness test has been expressly repudiated, and Brandenburg did the repudiating when it overruled Whitney v. California, the 1927 decision that had sustained California's Criminal Syndicalism Act.6
The first subsection below traces the danger test from Schenck to Dennis and Yates. The next two take Brandenburg's conditions in turn: the first asks what the speaker was aiming at, and the second asks what the speech was likely to do. The last subsection applies the standard to organizations and to those who support them.
Clear and Present Danger
Schenck v. United States introduced the test. The defendants were charged with conspiring under the Espionage Act of 1917 to cause insubordination in the military and naval forces and to obstruct recruiting, by printing and circulating to men who had been called and accepted for military service a document alleged to be calculated to cause that insubordination.7 Writing for the Court, Justice Holmes asked in every case whether the words were used in such circumstances, and were of such a nature, as to create a clear and present danger of the substantive evils Congress has a right to prevent. He called it a question of proximity and degree.8
The language comes from the law of criminal attempt. As a state judge, Holmes had asked whether a defendant's acts came near enough to the completed offense to be punishable, which he called a question of degree.9 Circumstances therefore control the result. The character of any act depends on the setting in which it is done, so a person who falsely shouts fire in a theater and causes a panic gets no protection, and words that might pass in ordinary times may be punished in a nation at war.10
Despite its careful wording, the formula protected no one in the cases that announced it. Within days of Schenck the Court affirmed convictions in Frohwerk v. United States and Debs v. United States, both opinions by Holmes.11 Frohwerk rested on newspaper articles. The Court could not say they were incapable of supporting a conviction, partly because it did not know how strong the government's evidence had been. It reasoned that a jury might have found the paper circulated in quarters where “a little breath would be enough to kindle a flame,” a fact known to and relied upon by those who sent it out.12 In Debs the speaker had praised individual draft resisters for their courage and sacrifice.13
The early results make more sense against the test most lower courts were then applying. Under the bad-tendency test, speech could be punished if its natural tendency and probable effect was to bring about the substantive evil the legislature might prevent, and any tendency to produce bad acts, however remote, sufficed to validate a repressive statute.14 Holmes rejected that test as the standard, and his formula was meant to be tighter. In practice, though, the first cases came close to the older approach: tendency in wartime circumstances was enough, and every defendant lost. The speech-protective version of the danger test arrived later, in dissent and concurrence.
That version took shape in Abrams v. United States. The majority affirmed convictions for leaflets attacking the government's war policy, and Holmes, joined by Brandeis, dissented. We must tolerate even opinions we loathe, he wrote, “unless they so imminently threaten immediate interference with the lawful and pressing purposes of the law that an immediate check is required to save the country.”15 Behind that limit lies the marketplace of ideas. Time has upset many fighting faiths, and the best test of truth is the power of a thought to get itself accepted in competition with other thoughts.16 If the harm is not immediate, listeners have time to answer the speaker, and the response the system trusts is more speech. Only an emergency that makes it dangerous to leave the correction of evil counsels to time justifies an exception.
Brandeis supplied the framework, and the case he answered was Gitlow v. New York. Gitlow held that the Due Process Clause of the Fourteenth Amendment carries the First Amendment's speech guarantee to the states, so that state and federal laws restricting speech face the same standards.17 Gitlow had been convicted under an anarchy law for publishing a Manifesto. His sole contention was that, with no evidence of any concrete result or of circumstances showing a likelihood of one, the statute punished the mere utterance of doctrine having no quality of incitement.18 The majority answered with deference to the legislature: where a statute itself condemns a kind of advocacy, the speech may be punished if its natural tendency and probable effect was to bring about the evil the legislature may prevent.19
Two years later, in Whitney, Brandeis wrote a concurrence that Holmes joined and that read like a dissent. He rejected the notion, first presented in Gitlow, that enacting a statute forecloses the clear-and-present-danger inquiry: “the enactment of the statute cannot alone establish the facts which are essential to its validity.”20 Deference to a legislative determination of danger would make judicial review essentially meaningless, a theme Chief Justice Burger later adopted: “Deference to a legislative finding cannot limit judicial inquiry when First Amendment rights are at stake.”21 Brandeis nonetheless voted to affirm, because Whitney had not argued that no clear and present danger existed, and that left his opinion a dissent in all but name.22
Advocacy that falls short of incitement, with nothing to indicate that it would be immediately acted on, fails these inquiries however reprehensible it may be morally.26
Imminence does the most work in the sequence, because it decides whether discussion remains available as a remedy. When the harm cannot occur before people have time to answer the speaker, the cure for bad ideas is more speech and repression has no justification. When the harm will arrive before discussion can begin, the marketplace has no time to operate.27 The same words can fall on either side of that line depending on the audience and the moment, so a court must read them in context. The Court soon applied the Brandeis insistence on incitement and imminence in Herndon v. Lowry. The evidence showed that the defendant had advocated a doctrine embracing unlawful action, but nothing showed that he had incited anyone to take such action or that unlawful action was imminent as a result, and the Court reversed his conviction.28 In Bridges v. California it described the “minimum compulsion” of the First Amendment as requiring that “the substantive evil must be extremely serious and the degree of imminence extremely high before utterances can be punished.”29
A rival approach predates these opinions. In Masses Publishing Co. v. Patten, a 1917 district court decision interpreting the Espionage Act, Judge Learned Hand wrote: “If one stops short of urging upon others that it is their duty or their interest to resist the law, it seems to me one should not be held to have attempted to cause its violation.” He drew a line between legitimate agitation and direct incitement to violent resistance, and only the latter could be punished consistent with the First Amendment.30 Hand's incitement test is in fact an absolutist approach, because advocacy of illegal conduct that falls short of incitement is absolutely protected.31 It looks to the speaker's language, where the danger test looks to the circumstances in which the speech is heard.
Each test has a weakness that explains why neither prevailed alone. The danger test requires a prediction about effects, and the early convictions show how loosely a court can make that prediction. Hand's test sorts speech by what the speaker urged and ignores whether anyone was positioned to act on it.
Dennis v. United States arrived in the McCarthy era, an age of suspicion in which merely being suspected of belonging to a communist or radical group could cost a person a job or place a name on a blacklist.32 The defendants, leaders of the American Communist Party, were convicted of conspiring to advocate overthrow of the government by force and violence. Chief Justice Vinson, writing for a four-Justice plurality, adopted a reformulation in Judge Hand's terms: “whether the gravity of the ‘evil,’ discounted by its improbability, justifies such invasion of free speech as is necessary to avoid the danger.”33 The formula works as a sliding scale. The greater the gravity of the act advocated, the less clear and present the danger need be, and advocacy of revolution is grave enough that the danger need not be very clear or very present. The test thus became a disguised balancing test that weighed the seriousness of the danger against competing interests in free speech.34
The government need not show that the putsch was about to be executed; it assumed only the relatively light burden of showing tendencies and probabilities.35 Critics called the result “the remote bad tendency test dressed in modern style,” because any radical political doctrine would always appear as a threat to the nation and so as the most serious of evils.36 Dennis shows why imminence cannot be bargained away. A test that weighs only the gravity of the feared evil gives a majority a ready basis for silencing an unpopular group in a time of fear.
Yates v. United States moved the law part of the way back by stressing advocacy of action over advocacy of abstract doctrine, an emphasis that pointed toward incitement.37 That emphasis set up the combination the Court arguably reached in Brandenburg, which pairs the incitement focus of Hand's approach and of Yates with the probability and imminence that the Holmes-Brandeis danger test required. Brandenburg never mentions clear and present danger, and it cites Dennis and Yates as authority, two decisions that are arguably inconsistent with each other.38
Directed to Inciting
Brandenburg states the modern rule in a single sentence, and its two conditions are joined by and. This subsection takes the first, which asks what the speech is aimed at. The facts show what that condition screens out. Brandenburg and others had gathered before newsmen, who filmed the scene. One film showed twelve hooded figures, some carrying firearms, around a large wooden cross that they burned. No one was present other than the participants and the newsmen. Most of the words uttered were incomprehensible when the film was projected, but scattered phrases could be understood that were derogatory of Black people and, in one instance, of Jews.40 In the second film six hooded figures appeared, and one of them, later identified as Brandenburg, repeated a speech very similar to the first. That version omitted the reference to the possibility of “revengeance” and added a sentence saying that Black people should be returned to Africa and Jews to Israel. Though some of the figures carried weapons, the speaker did not.41
Ohio prosecuted Brandenburg under its Criminal Syndicalism statute, which also punished voluntarily assembling with a group formed to teach criminal syndicalism. He was fined $1,000 and sentenced to one to ten years' imprisonment.42 The Court reversed in a short per curiam opinion, one issued for the Court as a whole instead of being signed by a single Justice. Its analysis targets the statute and not the content of the filmed speeches, because neither the indictment nor the trial judge's instructions refined the statute's definition of the crime in terms of mere advocacy not distinguished from incitement to imminent lawless action.43 The Court confronted a statute that, by its own words and as applied, purported to punish mere advocacy and to forbid, on pain of criminal punishment, assembly with others merely to advocate the described type of action.44
That is a defect in the statute itself, the kind that makes a law invalid on its face. When the Court invalidates a law on its face, the law cannot be used to punish anyone under any circumstances.45 The reason is structural. If the crime is advocacy, every prosecution can succeed by proving only that the speaker argued for force, and that showing is satisfied alike by a philosophical argument and by a command to act. A law drawn that loosely reaches speech the Constitution protects.
The words “directed to” place the speaker's aim among the things the government must prove: the speaker must have meant the words to bring about immediate unlawful action.46 The Court restated the point in Counterman v. Colorado, observing that the First Amendment bars punishment for incitement, civil or criminal, unless the speaker's words were intended, and not merely likely, to produce imminent disorder.47 Intent alone does not complete the condition, because the words must also urge the action. One way to unpack the standard treats it as three requirements: the state must prove that the speaker subjectively intended incitement, that the words used were likely in context to produce imminent lawless action, and that the words objectively encouraged and urged incitement. The first and third concern aim and expression and belong to this condition, and the second is the subject of the next subsection. The third requirement, with its focus on the objective words the speaker used, derives from Hess v. Indiana.48
Purpose carries this much weight because most arguments for change, pressed hard enough, invite listeners to act, and a rule keyed to a speech's tendency would reach most forceful dissent. The Court drew its line from Noto v. United States: “the mere abstract teaching . . . of the moral propriety or even moral necessity for a resort to force and violence, is not the same as preparing a group for violent action and steeling it to such action.”49 Teaching that force can be right leaves each listener free to reject the idea. Preparing a group for violence and steeling it to act is an effort to produce the act.
This requirement marks the sharpest break from the era of deference. Whitney had sustained a state syndicalism act, and the Court of that era decided syndicalism cases by asking only whether the law and the prosecution were reasonable.50 Brandenburg declared that the contrary teaching of Whitney could not be supported and overruled it.51 Ohio's statute, like California's, defined the crime by the subject of the advocacy, and the reasoning that overruled Whitney condemned it.
The condition has two ancestors. Judge Hand's approach in Masses Publishing Co. v. Patten looked to whether the speaker's language directly urged action, and Holmes's clear-and-present-danger test looked to the consequences of speech in its circumstances. The Court combined them. The incitement language comes from the Hand side, and the imminence and likelihood come from the danger side.52 Because the requirements are cumulative, the government must prove all of them, and the third requirement of the three-part analysis, the objective urging of incitement, is the one that carries the Hand tradition forward.53
The usual error is to treat the ugliness of the message as the test. The film showed hooded men, weapons, a burning cross, and racist and antisemitic remarks, yet the conviction was reversed, because no attempt had been made at trial to distinguish incitement from advocacy, and any law punishing mere advocacy of Ku Klux Klan doctrine and assembly of Klan members to advocate their beliefs was unconstitutional.54 The assembly count fell with the speech count, since statutes touching the right of assembly must observe the same distinction between mere advocacy and incitement to imminent lawless action.55
Satisfying the first condition does not end the inquiry. A speaker who urges concrete unlawful action still cannot be punished unless the speech threatens imminent unlawful conduct and there is a probability that lawless action will occur as a result.56 That second condition is the subject of the next subsection.
Likely and Imminent
The second condition looks to what speech could actually do where it was delivered. A state may punish advocacy only if it is likely to incite or produce imminent lawless action, and because the clause refers back to the imminent lawless action named in the first condition, the government must show a real probability of lawless action soon. A speaker's hope for it falls short.57 The Court has put the point generally: regardless of the inciting character of the speech, a defendant cannot be prosecuted unless the speech threatens imminent unlawful conduct and unless there is a probability that lawless action will occur as a result.58 Imminent means close enough in time that the harm may occur before anyone has a chance to answer, and likely means that the circumstances, realistically judged, make the harm a genuine probability. Brandenburg's own rally shows why the words matter, because no one was present but the participants and the newsmen filming them, a setting in which it is hard to see who might be swept into lawless action.59 The holding rested on the statute's failure to separate advocacy from incitement, so the rally illustrates the second condition without having applied it to a finding of fact.
Each word does its own work. Imminence preserves the remedy the First Amendment trusts. When the harm is distant, listeners have time to answer the speaker and officials have time to act against actual preparation or attempt. When the harm is close, a speaker who provokes an unthinking, immediate reaction leaves no time for reasoned debate, and one cannot rely on more speech in the marketplace of ideas to correct the error.60 Likelihood closes a separate gap. A speaker may intend and even urge unlawful action while addressing people who will not or cannot carry it out, and then the danger that justifies punishment does not exist. This is why soapbox orators and professors love the Brandenburg test: because no one ever listens to them, the government cannot demonstrate that their speech has any probability of doing anything other than causing people to laugh.61
Hess v. Indiana applies the second condition to words that sound menacing on the surface. After police cleared a street of antiwar demonstrators, Hess, standing at the curb, said, “We'll take the fucking street later.” He was convicted of disorderly conduct, and the state courts found that his remark was intended to incite further riotous behavior and was likely to produce it. The Supreme Court reversed in a per curiam opinion. At best the statement could be taken as counsel for present moderation, and at worst it was advocacy of illegal action at some indefinite future time. Nothing in the record or in a rational reading of the words showed that they were intended to produce, and likely to produce, imminent disorder, so they could not be punished on the theory that they had a tendency to lead to violence. The Court added that the remark was not directed to any person or group of persons, so Hess had not advocated action that would produce imminent disorder.62 The word later carries the analysis, because talk of what a group might do at an unspecified future time leaves room for argument, for policing, and for second thoughts. The case also shows that a lower court's finding of likelihood does not end the inquiry, since the Court examined the record itself and found no support for it.
A different failure appears in NAACP v. Claiborne Hardware Co., where the Court considered a damages award against participants in a boycott of local merchants. The trial court had based liability in part on a speech by an NAACP official who said, “If we catch any of you going in any of them racist stores, we're gonna break your damn neck.” The Court held that the speech was protected under the Brandenburg test and could not be the basis for liability.63 It acknowledged that, in the passionate atmosphere of the speeches, the words might have been understood as inviting an unlawful form of discipline, and it held the speech protected anyway. The addresses were emotional pleas for unity and for the use of the community's political and economic power, and the identified violence came weeks or months afterward. The opinion left one door open: had the strong language been followed by acts of violence, a substantial question would arise whether the speaker could be held liable for their consequences.
Set beside Hess, the case shows the second condition failing in two ways. Hess's words pointed to an indefinite future and so lacked imminence on their face. The Claiborne speaker's words carried a sharper sting, but they were an emotional appeal to supporters, and the violence that followed came long afterward, so nothing in the record showed speech that produced or was likely to produce immediate lawless action. An earlier decision reads speech in context in a similar way. In Bond v. Floyd the Court held that the Georgia legislature could not refuse to seat Julian Bond because he supported a statement sharply critical of the Vietnam War and the draft, and it concluded that his statements were advocacy of ideas protected by the First Amendment.64
The words in the second scene are no harsher than Leonard's. The difference lies in the audience and the clock. Leonard's speech is protected because no one in the park is positioned to act on it that day, however radical the content.
Brandenburg leaves open how imminence and likelihood are measured when the threatened harm is grave. The test as written says nothing about the gravity of the harm, and it does not say whether a more serious danger requires less imminence and likelihood or whether some showing of both is needed whatever the harm.65 The question has consequences, because a sliding scale would bring back the Dennis formula. Brandenburg cited Dennis and Yates as authority even though the two are arguably inconsistent with each other.66 The most speech-protective answer treats both requirements as necessary whatever the harm, and the wording of the test supports that reading, but the opinion's silence leaves room for the other.
Because the likelihood requirement looks to the audience, most advocacy of illegal ideas never comes near punishment. A speaker who offers reasons why officials deserve punishment leaves listeners time to hear a rebuttal and decide for themselves, while a shouted command to attack bypasses deliberation. The next subsection carries the same standard to organizations, where the question becomes what a person's membership in, or support for, a group that advocates unlawful action reveals about the member's own intent.
Membership and Material Support
Brandenburg separates the abstract teaching of the moral propriety of force from “preparing a group for violent action and steeling it to such action,” and a statute that fails to draw that line sweeps in speech the Constitution protects.67 The Ohio statute punished both advocacy and voluntarily assembling with a group formed to teach criminal syndicalism, so the distinction had to hold for group activity as well as for individual speech.68 Organizations raise a further question that the incitement cases leave open: what does a person's tie to a group reveal about that person's own intent? The answer turns on whether the government punishes a status or an act.
A tie to an organization says little about what its member intends. People join groups for social, political, or reformist reasons, and some join hoping to steer the group away from the plans its leaders favor. A rule that punishes membership as such attributes the group's aims to every member, so guilt arrives by association instead of by personal choice. The Court has accordingly recognized that a person who associates with a group that uses unlawful means does not thereby necessarily forfeit the First Amendment's protection for freedom of association.69 Scales v. United States placed the line at specific intent. Knowing membership in even a subversive organization cannot be punished by itself, while active membership joined with specific intent to further the organization's illegal ends can be. Specific intent supplies what status lacks, because it ties the individual to the crime he is punished for advancing, so the law reaches participation and purpose and leaves belief alone. Material support, taken up next, sits uneasily beside that rule.70
Material support tests the same line between status and act in a harder setting. The federal statute makes it a crime to knowingly provide “material support or resources” to a designated foreign terrorist organization, and the definition expressly includes training, expert advice, or assistance.71 In Holder v. Humanitarian Law Project the plaintiffs, United States citizens and associations, wanted to train members of the PKK in using humanitarian and international law to resolve disputes peacefully, to teach them how to petition the United Nations for relief, and to engage in political advocacy on behalf of Kurds in Turkey and Tamils in Sri Lanka.72 They said they sought only to advance the group's lawful, nonviolent purposes. The Court held the statute constitutional as applied to the particular activities the plaintiffs described, and it left the more difficult cases that may arise under the statute for another day.73
The peaceful aim of the plaintiffs did not save them, and the reason lies in what the statute targets. Congress found that designated organizations are “so tainted by their criminal conduct that any contribution to such an organization facilitates that conduct.”74 On that premise the harm comes from the help itself. A group's resources are interchangeable, so training in lawful advocacy still gives it skills and standing and frees other resources for violence. The supporter's own purpose therefore does not remove the danger, which is why the statute can reach a person with entirely peaceful aims. That is the opposite of the rule for membership, where the individual's specific intent decides whether punishment is allowed. On the Court's reading the statute applies only where the group helps to coordinate the defendant's activities, so a speaker who argues for the group's cause entirely on his own falls outside the offense.75
The majority upheld the law without reference to Brandenburg or the other incitement cases.76 Justice Breyer, dissenting, argued that Brandenburg should have been applied and was not met. No one contended that the plaintiffs' speech could be prohibited as incitement, he observed, because they sought to advocate peaceful, lawful action, and coordination with a political group, like membership, implicates the freedom of association.77 His reading would carry the Scales requirement into the support setting, reaching pure speech and association only where the defendant knows or intends that the activity will assist the organization's unlawful actions, but a majority did not adopt it. A student must therefore be able to draw the line between status and active participation. Passive membership and independent advocacy of a group's cause remain protected. Active membership with specific intent to further illegal ends and knowing material support to a designated terrorist organization may be punished, even when the support is meant to advance only lawful goals.