Criminalizing Offensive Speech: The Fighting Words Exception to the First Amendment

In an earlier post, I discussed one category of unprotected speech that may be criminally punished consistent with the First Amendment, speech integral to criminal conduct. My former colleague Jonathan Holbrook has written about another category of unprotected speech, true threats. Today’s post discusses another type of unprotected speech that may be criminalized: fighting words. The First Amendment generally protects even highly offensive and insulting speech. When does offensive language lose its presumptive First Amendment protection and give rise to criminal liability? This issue potentially arises with a variety of crimes, such as disorderly conduct, resisting arrest, harassing phone calls, and other abusive language or breach of peace offenses.

The Birth and Evolution of the Fighting Words Exception. In Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942), the Supreme Court first recognized so-called “fighting words” as a category of unprotected speech. Under Chaplinsky, fighting words are those that “which by their very utterance inflict injury or incite an imminent breach of peace.” Id. The defendant in Chaplinsky was charged with and convicted of violating a New Hampshire statute forbidding (in relevant part) calling or addressing another person “any offensive, derisive, or derogatory word” in public. Id. at 569.  The defendant called the complainant “a God damn racketeer” and “a damned fascist” while condemning a city official and the city government. In affirming the conviction, the Court observed that “resorts to epithets or personal abuse is not in any proper sense communication of information or opinion safeguarded by the Constitution,” and could properly be criminalized. Id. at 572.

Chaplinsky originally staked out a broad exception from First Amendment protection for insulting language—any words inherently hurtful or likely to cause a violent response were unprotected fighting words. That approach has been considerably narrowed over time. A few years after Chaplinsky, the Court decided Terminiello v. Chicago, 337 U.S. 1 (1949). There, the defendant was convicted of breach of peace based on a speech he gave to a large crowd in which he vigorously condemned certain political and racial groups, resulting in some unrest among the crowd. In vacating the conviction on free speech grounds, the Court seemingly narrowed fighting words to situations “shown likely to produce a clear and present danger of a serious substantive evil that rises far beyond public inconvenience, annoyance, or unrest.” Id. at 4. This arguably eliminated the “inflict injury” prong of the Chaplinsky formulation.

That change was clear by the time Gooding v. Wilson, 405 U.S. 518 (1972), was decided. There, the defendant called one officer a “white son of a bi*ch” and another a “son of bi*ch” while threatening to kill them both during a protest against the war in Vietnam. Reversing the defendant’s conviction for using abusive words tending to cause a breach of peace, the Court cited Chaplinsky for the proposition that fighting words are limited to those words that are likely to provoke an immediate, violent response from the person to whom the words are directed. Similarly, in Cohen v. California, 403 U.S. 15 (1971), while reversing the defendant’s conviction for disturbing the peace based on his wearing a jacket with the words “Fu*k the Draft” inside a courthouse, the court noted that fighting words must be personally insulting words directed at an individual. The Court has also indicated that the context in which offensive words are spoken matters. See R.A.V. v. City of St. Paul, 505 U.S. 377, 432 (1992) (Stevens, J., concurring) (“Whether words are fighting words is determined in part by their context.”)

Putting it all together, fighting words must be personal insults, directed at an individual, likely to provoke an immediate and violent reaction from the person hearing the words, in light of the overall context in which the words are spoken. It is generally not sufficient that there is some potential that the words might provoke a violent response from someone; there should be evidence in a fighting words prosecution that the words would provoke an immediate, violent reaction from the hearer. Gooding at 528.

Even then, the Supreme Court has not found another offensive word, phrase, or act of expressive conduct to qualify as a fighting word since Chaplinsky was decided in 1942. Consider the following cases, all of which found that the offensive speech (or expressive conduct) at issue did not qualify as fighting words and was entitled to First Amendment protection.

  • Lewis v. City of New Orleans, 415 U.S. 130 (1974) (“goddamn mother fu*king police”)
  • Brown v. Oklahoma, 408 U.S. 914 (1972) (“mother fu*king fascist” and “black mother fu*king pig”)
  • Texas v. Johnson, 491 U.S. 397 (1989) (burning a U.S. flag)
  • Virginia v. Black, 538 U.S. 343 (2003) (burning a cross without intent to intimidate)
  • Synder v. Phelps, 562 U.S. 443 (2011) (homosexual slurs and insults of dead U.S. soldiers)

The fighting words standard may be even higher when the offensive speech is directed at a law enforcement officer. “[A] properly trained officer may reasonably be expected to ‘exercise a higher degree of restraint’ than the average citizen, and thus be less likely to respond belligerently to ‘fighting words.’” Lewis at 135 (Powell, J., concurring) (internal citation omitted). See also City of Houston v. Hill, 482 U.S. 451, 461 (1987) (“[T]he First Amendment protects a significant amount of verbal criticism and challenge directed at police officers.”)

Recent Cases on Fighting Words. The Court of Appeals had occasion to address the fighting words exception in a recent case in the context of G.S. 143-318.17, the prohibition on disrupting a public meeting. In State v. Barthel, COA25-159, ___ N.C. App. ___ (Nov. 5, 2026), the Court of Appeals struck down the defendant’s convictions for disrupting a public meeting and resisting arrest on First Amendment grounds. The defendant attended an Avery County Board of Commissioners meeting, which was set to begin with a public comment period. While a commissioner began announcing the rules for public comment, the defendant stood in the back of the room and unfurled a banner with a picture of a county employee. The banner named the depicted person and referred to her as “Avery County’s Most Unprofessional Employee.” It also said, “I’m not a gynecologist, but I know a c*nt when I see one!!” (the c-word was uncensored on the banner). The defendant did not otherwise speak or act to disrupt the meeting and did not block anyone’s view. Law enforcement immediately approached the defendant and asked him to take the banner down. When he refused, officers attempted to handcuff the defendant, and he resisted those efforts by “tensing up his body.” Barthel Slip op. at 3. The officers eventually moved the defendant out of the boardroom and into the hallway. There, the defendant continued arguing with the officers and told them that he was being unlawfully detained. Ultimately, officers placed him under arrest, charging him with disrupting a public meeting and resisting arrest. The defendant was tried and convicted of those charges in district court and again on de novo appeal in superior court over his objections that his conduct was protected First Amendment speech.

On appeal, a unanimous panel of the Court of Appeals agreed with the defendant. In the words of the court:

[The defendant’s] words do not qualify as fighting words because there was ‘no likelihood’ that [the commissioner named on the banner] ‘would make an immediate violent response.’ And offensive as they were, Defendant’s ‘personal insults’ describing a commissioner as a ‘c*nt’ and ‘unprofessional employee’ remain, under these facts, mere ‘distasteful modes of expression.’ Id. at 10-11 (internal citations omitted).

The Barthel opinion goes on to discuss the fact that the speech at issue occurred in a limited public forum, where the government may impose reasonable, content-neutral restrictions on speech but may not engage in viewpoint discrimination, but this fact was ultimately of no help to the prosecution. [My colleagues Kristi Nickodem and Kristina Wilson wrote a blog post diving into the public forum issues and rules for public comment periods over on Coates Canons, available at this link, if you are interested in more on those points.]

Because the defendant’s speech was protected and because it was the officers who caused the disruption of the meeting in response to the defendant’s banner, the defendant’s conviction for disrupting a public meeting was vacated. Because the defendant’s alleged act of resisting arrest was based on his unlawful arrest for disrupting a public meeting and the defendant used reasonable force to resist that unlawful arrest, his conviction for resisting an officer was likewise set aside. See State v. Mobley, 240 N.C. 476, 478-79 (“[E]very person has the right to resist an unlawful arrest” with reasonable force.)

As Barthel notes, the North Carolina Supreme Court has only recognized offensive speech as fighting words once, and that was in the context of proceedings to remove an elected district attorney from office (not a criminal case). Where a white district attorney “loudly and repeatedly” called a Black man the n-word while trying to instigate a fight with the man, his words lost First Amendment protection as fighting words. In Re: Spivey, 345 N.C. 404, 408 (1997). Again, though, the offensive word itself was not enough—evidence was presented that the defendant was intending to start a fight with the man and tried to do so. See also U.S. v. Bartow, 997 F.3d 203, 209 (4th Cir. 2021) (striking down the defendant’s conviction for abusive language for use of the n-word towards black employees at a store where there was no evidence of a likely or actual immediate violent response by the people hearing the slur) (“[E]ven the most egregious racial slur is not a fighting word per se. The circumstances in which the word is used matter a great deal.”)

The Court of Appeals has occasionally affirmed convictions for use of fighting words. In State v. Cunnigham, 34 N.C. App. 72 (1977), the court upheld a conviction for disorderly conduct where the defendant accosted a parking officer about a ticket, repeatedly cussed at him, and threatened to run him over with his car. The court noted there that “the likelihood of violent retaliation may have been slight,” given that the person hearing the words was a public officer, but nonetheless found that an average person would have been likely to react to such language violently. Id. at 76. Cunningham has twice been cited by North Carolina courts in unpublished cases to uphold disorderly conduct convictions against First Amendment challenges for speech occurring in a school setting but is arguably out of step with most modern cases addressing fighting words. See Matter of A.R.-V., 265 N.C. App. 601 (2019) (unpublished) (upholding a juvenile adjudication for disorderly conduct based on profanity directed towards school resource officer); State v.  Tucker, 164 N.C. App. 232 (2004) (unpublished) (upholding a disorderly conduct conviction for loudly yelling and cussing in the school principal’s face in front of the defendant’s child and her teacher within hearing of others at the school).

Final Thoughts. As the cases discussed above indicate, the bar to establish offensive speech as unprotected fighting words is a high one and is unlikely to be met by words alone, no matter how offensive the message may be. The application of fighting words has been consistently narrowed for decades, with some commenters seeing it as a relic of earlier times when public brawling was more socially acceptable and others questioning whether it still exists at all in any meaningful way. See U.S. v. Bartow, 997 F.3d 203, 208 at n. 2 (4th Cir. 2021) (compiling criticism of the doctrine). Of course, even if a defendant’s speech does not meet the high bar required to qualify for the fighting words exception, it may fall within another category of unprotected speech, such as true threats, speech integral to criminal conduct, or incitement to violence. For defenders handling a case involving offensive speech, consider making a pretrial motion to dismiss on First Amendment grounds, as the defendant in Barthel did. If that motion is denied, remember that the denial of such a motion is not something that can be preserved following a guilty plea; the defendant will have to try out the case and renew the constitutional argument on appeal. See G.S. 15A-1444(a2) (limiting the grounds on which a defendant can appeal following a guilty plea).

I can be reached as always at dixon@sog.unc.edu with any questions or comments.

About the Author