Protecting Listeners From Unwanted One-to-One Speech

I. The Value of the One-to-One vs. One-to-Many Line

“[N]o one has a right to press even ‘good’ ideas on an unwilling recipient,” the Supreme Court held in Rowan v. United States Post Office Department.¹Rowan v. U.S. Post Off. Dep’t, 397 U.S. 728, 738 (1970). At the same time, “[t]he fact that society may find speech offensive is not a sufficient reason for suppressing it. Indeed, if it is the speaker’s opinion that gives offense, that consequence is a reason for according it constitutional protection.”²Hustler Mag., Inc. v. Falwell, 485 U.S. 46, 55 (1988) (cleaned up). That is generally true even if the speaker’s opinion gives offense not just to “society” but to many of the speaker’s listeners.³Bolger v. Youngs Drug Prods. Corp., 463 U.S. 60, 72 (1983).

The best way to reconcile these principles, it seems to me, is to distinguish (1) one-to-one speech said to an unwilling listener from (2) one-to-many speech that reaches both potentially willing and unwilling listeners.⁴Eugene Volokh, One-to-One Speech vs. One-to-Many Speech, Criminal Harassment Laws, and “Cyberstalking”, 107 Nw. U. L. Rev. 731 (2013); Eugene Volokh, Freedom of Speech in Cyberspace from the Listener’s Perspective: Private Speech Restrictions, Libel, State Action, Harassment, and Sex, 1996 U. Chi. Legal F. 377, 421–23 (1996); Eugene Volokh, Thinking Ahead About Freedom of Speech and Hostile Work Environment Harassment, 17 Berkeley J. Emp. & Lab. L. 305, 311 (1996); Eugene Volokh, Freedom of Speech and Workplace Harassment, 39 UCLA L. Rev. 1791, 1863–67 (1992) (using the terms “directed” and “undirected” instead of “one-to-one” and “one-to-many”). Ashutosh Bhagwat well explains both the precedents and the policy arguments supporting the distinction. Most speech should generally be protected because it may persuade or inform some potentially willing listeners even if others are upset.⁵Ashutosh Bhagwat, Respecting Listeners’ Autonomy: The Right to be Left Alone, 98 S. Cal. L. Rev. 1129, 1145 (2025). But speech said solely to an unwilling listener, where it’s clear the listener is unwilling, is likely only to offend. The government can in many situations help protect listeners against such one-to-one speech, because that promotes the unwilling listener’s autonomy without interfering with communication to potentially willing listeners.⁶Id. at 1145–48.

And this helps explain the constitutionality of many common speech restrictions, including:

1. telephone harassment laws,⁷Volokh, One-to-One Speech, supra note 4, at 740.
2. do-not-call registries,⁸See, e.g., Patriotic Veterans, Inc. v. Zoeller, 845 F.3d 303, 306 (7th Cir. 2017).
3. harassment restraining orders that forbid speech to the protected person,⁹Volokh, One-to-One Speech, supra note 4, at 741.
4. application of university “hostile environment harassment” policies to people “following students around and yelling slurs or otherwise directing hostile speech at individual students who have demanded to be left alone,”¹⁰Bhagwat, supra note 5, at 1153.
5. application of workplace harassment law to one-to-one insults, or one-to-one repeated unwanted romantic advances,¹¹Volokh, Freedom of Speech and Workplace Harassment, supra note 4, at 1863–68.
6. residential picketing laws,¹²Bhagwat, supra note 5, at 1144-45. and more.

II.  Must Restrictions on Unwanted One-to-One Speech Be Content-Neutral?

This general conclusion, however, raises subsidiary questions. A particularly important one is whether restrictions on one-to-one speech must be content-neutral.

There is precedent suggesting this, as well as broader First Amendment principles supporting such a view. Frisby v. Schultz upheld a content-neutral residential picketing ban on the grounds that such picketing is essentially speech targeted to the unwilling listener in the home.¹³Frisby v. Schultz, 487 U.S. 474, 486, 488 (1988). But Carey v. Brown had earlier struck down a residential picketing ban that excluded labor picketing because that exclusion made the law content-based.¹⁴Carey v. Brown, 447 U.S. 455, 470–71 (1980). It was the content neutrality of the ban in Frisby that saved it.¹⁵Frisby, 487 U.S. at 481, 488.

We see something similar in Rowan v. United States Post Office Department.¹⁶Rowan v. U.S. Post Off. Dep’t, 397 U.S. 728, 738 (1970). Rowan upheld a statute that barred senders from sending material to householders, once the householder informed the post office that he “in his sole discretion believes [the mailings] to be erotically arousing or sexually provocative.”¹⁷Id. at 730. The statute was thus content-based on its face, but the Court stressed it was essentially content-neutral as enforced:

Both the absoluteness of the citizen’s right under [the statute] and its finality are essential; what may not be provocative to one person may well be to another. In operative effect the power of the householder under the statute is unlimited; he may prohibit the mailing of a dry goods catalog because he objects to the contents—or indeed the text of the language touting the merchandise. Congress provided this sweeping power not only to protect privacy but to avoid possible constitutional questions that might arise from vesting the power to make any discretionary evaluation of the material in a governmental official.¹⁸Id. at 737.

Yet if content neutrality is indeed required in such situations, then many restrictions on one-to-one speech would be hard to defend. Telephone harassment laws, for instance, often specially target lewd or indecent harassing calls.¹⁹See, e.g., Wash. Rev. Code Ann. § 9.61.230 (2024). Workplace harassment law ends up specially targeting one-to-one speech that is personally insulting.

Likewise, when various laws target one-to-one speech intended to “harass” or “abuse,” they must be treated as content-based. As the Court held in Reed v. Town of Gilbert, “[s]ome facial distinctions based on a message are obvious, defining regulated speech by particular subject matter, and others are more subtle, defining regulated speech by its function or purpose. Both are distinctions drawn based on the message a speaker conveys, and, therefore, are subject to strict scrutiny.”²⁰Reed v. Town of Gilbert, 576 U.S. 155, 163–64 (2015). When the “regulated speech” is defined by a purpose to harass or abuse, that definition generally targets speech that has a harassing or abusive “message.” The definition is therefore content-based.

More broadly, when even a “generally applicable law” is “directed at [a speaker] because of what his speech communicated”—when the speaker violates the law “because of the offensive content of his particular message”—that too is treated as a “content-based regulation of speech.”²¹Holder v. Humanitarian L. Project, 561 U.S. 1, 28 (2010). This would cover most harassment laws, at least when speech is found to be harassing because of its offensiveness rather than because it’s too loud or ties up telephone lines.

Indeed, relatively few of these laws actually set up Rowan-like rules that (1) require the listener to first tell a speaker, “stop speaking to me,” but then (2) make that order binding regardless of what the speaker wants to say. The laws are indeed aimed at “address[ing] the ‘first blow’ of curse words spoken only once.”²²Bhagwat, supra note 5, at 1154. At the same time, they aim to avoid giving someone an absolute veto on future communications: consider, for instance, workplace harassment, where the law can’t let employees categorically forbid any future communications (including on legitimate work-related topics) by coworkers.

Now perhaps that’s the wrong approach—perhaps the law should indeed insist on content neutrality even as to restrictions on unwanted one-to-one speech. Or perhaps content-based restrictions should indeed be subjected to strict scrutiny but might in some situations be upheld.

But I think it might be better to recognize that at least some such content-based restrictions are permissible when it comes to one-to-one speech, even if they wouldn’t be permissible as to one-to-many speech. The Court has acknowledged that content-based restrictions may be constitutional when “substantial privacy interests are being invaded in an essentially intolerable manner.”²³Erznoznik v. City of Jacksonville, 422 U.S. 205, 209–10 (1975) (quoting Cohen v. California, 403 U.S. 15, 21 (1971)). Perhaps the “privacy interests” here should be read as not just focusing on privacy in the home, or true captivity of a sort where it is “impractical for the unwilling viewer or auditor to avoid exposure.”²⁴Id. at 209. Rather, perhaps they should also be seen as including intrusions on the listener’s autonomy rights that Professor Bhagwat rightly identifies: the targeting of a particular likely unwilling listener for one-to-one speech may be what is “essentially intolerable.”

R.A.V. v. City of St. Paul²⁵R.A.V. v. City of St. Paul, 505 U.S. 377 (1992). may provide a helpful framework for dealing with this. The Court in R.A.V. held that content-based restrictions must generally be subject to strict scrutiny even when they are limited to subsets of unprotected categories of speech. For instance, a ban on racist fighting words would be presumptively unconstitutional even though a ban on all fighting words would be valid.²⁶Id. at 386. But the Court also held that this principle has certain exceptions, again where the content discrimination is entirely within an unprotected category; the relevant exceptions are:

1. “[w]hen the basis for the content discrimination consists entirely of the very reason the entire class of speech at issue is proscribable,”²⁷Id. at 388. for instance when the law restricts “only that obscenity which is the most patently offensive in its prurience,” or “only those threats” that are especially disruptive;²⁸Id.
2. when “a particular content-based subcategory of a proscribable class of speech” is “swept up incidentally within the reach of a statute directed at conduct rather than speech”;²⁹Id. and
3. when “the nature of the content discrimination is such that there is no realistic possibility that official suppression of ideas is afoot.”³⁰Id. at 390.

The same might apply with regard to subcategories of likely unwanted one-to-one speech, if Professor Bhagwat and I are right that such speech is essentially constitutionally unprotected. Indecent harassing phone calls, for instance, may well be especially likely to be unwanted, and a restriction on such calls may indeed be unlikely to involve “official suppression of ideas.”

Likewise, a prohibition of one-to-one speech intended to abuse or harass might be justified on the same theory, and might also be “swept up incidentally within the reach of a statute directed at conduct rather than speech,” given that such harassment laws often do target nonspeech conduct (such as physical stalking) as well as speech. R.A.V. itself gave hostile environment harassment law as an example of a law that may “incidentally” “swe[ep] up” “sexually derogatory ‘fighting words,’ among other words,” because it bans a wide range of conduct as well as speech.³¹Id. at 389. Likewise, the law may incidentally sweep up derogatory unwanted one-to-one speech more broadly. (For reasons I explain elsewhere, this rationale does not extend to offensive one-to-many ideological expression, even when it’s viewed as sexist, racist, and the like.³²Volokh, Freedom of Speech and Workplace Harassment, supra note 4, at 1848–55.)

III.  When Must the Government Tolerate One-to-One Speech to Government Officials?

Though one-to-one speech to unwilling listeners may generally be forbidden, the analysis must be different when the speech is addressed to government employees on the job, especially public-facing employees. I agree with Professor Bhagwat that listeners generally have considerable autonomy interests in not hearing unwanted speech—interests that the government may protect. But when one works for the public,³³Query whether the same principle should also apply to public-facing employees of some private companies as well. one must accept the risk of disapproving speech from the public:

[R]eceiving mail from disgruntled constituents is usual for a politician. A person “who decides to seek governmental office must accept certain necessary consequences of that involvement in public affairs . . . [and] runs the risk of closer public scrutiny than might otherwise be the case.” Here, given Michael’s status as a selectman and the content of the letters, it cannot be said that Michael’s “substantial privacy interests [were] invaded in an essentially intolerable manner.”³⁴Commonwealth v. Bigelow, 59 N.E.3d 1105, 1113 (Mass. 2016) (citations omitted).

This is particularly clear for elected officials,³⁵Id. at 1108, 1112 (town council member); U.S. Postal Serv. v. Hustler Mag., Inc., 630 F. Supp. 867 (D.D.C. 1986) (Congressman); Hicks v. Faris, No. 1:20-CV-680, 2024 WL 4011824, at *14 (S.D. Ohio Aug. 30, 2024) (county treasurer); see also United States v. Yung, 37 F.4th 70, 78–79 (3d Cir. 2022) (dictum) (city councilman). candidates for office,³⁶State v. Drahota, 788 N.W.2d 796, 798, 804 (Neb. 2010) (candidate for state legislature); United States v. Sryniawski, 48 F.4th 583, 587 (8th Cir. 2022) (same). or high-level political appointees.³⁷United States v. Popa, 187 F.3d 672, 673 (D.C. Cir. 1999) (U.S. Attorney). But it may be true for lower-level public-facing employees as well, such as police officers³⁸State v. Fratzke, 446 N.W.2d 781, 782, 785 (Iowa 1989). or others.³⁹State v. Golga, 239 N.E.3d 1165 (Ohio Ct. App.) (water department). Some cases do allow punishing offensive speech to such employees,⁴⁰State v. White, No. 2024CA00044, 2025 WL 354802 (Ohio Ct. App. Jan. 29, 2025) (police officer); United States v. Waggy, 936 F.3d 1014, 1015 (9th Cir. 2019) (Veterans Administration employee). but I think they’re mistaken.⁴¹Cf. Hagedorn v. Cattani, 715 F. App’x 499, 507 (6th Cir. 2017) (viewing the Rowan principle as applicable to speech to a mayor’s personal email account because it is the “functional equivalent of a home mailbox”).

IV. The Borders of “One-to-One”

Finally, “one-to-one” and “one-to-many,” like many such useful general phrases, may not fully capture the legal principles that courts should and do apply. To give one example, say someone is speaking simultaneously to three listeners, all of whom have asked the speaker to stop bothering them. That’s technically one-to-three speech, not one-to-one speech. But it should be restrictable as tantamount to one-to-one speech, precisely because it is addressed solely at unwilling listeners.

Likewise, say Wendy Smith’s ex-husband Harry Smith posts a Facebook message on his own page saying, “My ex @WendySmith is a slimy trollop.” (This @ syntax is specifically designed to notify the Facebook user WendySmith about the post; Twitter and Instagram have the same feature.) It is thus more or less like an e-mail to Wendy (one-to-one speech), coupled with a post about her to the author’s friends (one-to-many speech). If Wendy gets a harassment restraining order barring further correspondence from Harry, it would be constitutionally permissible for that order to be interpreted as banning such mentions; Harry would still be able to communicate with his friends by posting the same item without the @ (“My ex Wendy Smith is a slimy trollop”).⁴²See, e.g., ARM v. KJL, 995 N.W.2d 361, 368–69 (Mich. Ct. App. 2022).

The hardest question arises when speech appears to be largely aimed at a particular unwilling listener but also reaches some other listeners. This is what the Court faced in Frisby v. Schultz, where it reasoned that residential “picketing is narrowly directed at the household, not the public”:

The type of picketers banned by the Brookfield ordinance generally do not seek to disseminate a message to the general public, but to intrude upon the targeted resident, and to do so in an especially offensive way. Moreover, even if some such picketers have a broader communicative purpose, their activity nonetheless inherently and offensively intrudes on residential privacy. . . .

Because the picketing prohibited by the Brookfield ordinance is speech directed primarily at those who are presumptively unwilling to receive it, the State has a substantial and justifiable interest in banning it.⁴³Frisby v. Schultz, 487 U.S. 474, 487–88 (1988).

Here the speech wasn’t “foisted (exclusively) upon unwilling listeners”⁴⁴Bhagwat, supra note 5, at 1147.—presumably at least some residential picketers also want to reach the resident’s neighbors.⁴⁵See Schultz v. Frisby, 807 F.2d 1339, 1341 (7th Cir. 1986), vacated, 818 F.2d 1284 (7th Cir. 1987). Rather, the Court says the speech was targeted “primarily” at the resident and acknowledges that it might have also had “a broader communicative purpose.”

Distinguishing the “primary” audience from the “secondary” is of course subjective, plus it’s not clear why even secondary audiences should be ignored. For instance, if animal rights protesters are picketing outside a fur store, is their speech “directed primarily” at buyers, who are likely “unwilling to receive” the message (especially if the message is framed harshly)? After all, fur buyers presumably know well where the fur comes from—and like it. Or is the speech directed at least equally to neighbors and passersby, or to the likely relatively rare ambivalent customer?

Likewise, most people who go to abortion clinics are likely unwilling to hear from anti-abortion protesters and counselors, but some might be open to their arguments.⁴⁶See McCullen v. Coakley, 573 U.S. 464, 473 (2014) (“In unrefuted testimony, petitioners say they have collectively persuaded hundreds of women to forgo abortions.”). Most people who go to churches, synagogues, or mosques that are being picketed are unwilling to hear from protesters,⁴⁷For cases upholding right to picket outside places of worship, see generally Survivors Network of Those Abused by Priests, Inc. v. Joyce, 779 F.3d 785 (8th Cir. 2015); Gerber v. Herskovitz, No. 22-1075, 2023 WL 2155050 (6th Cir. Feb. 22, 2023). but again some might be persuadable.

I’m not sure how this line is to be properly drawn. Perhaps courts should view Frisby as limited to “residential privacy,” given its reliance on the precedents saying that, “[a]lthough in many locations, we expect individuals simply to avoid speech they do not want to hear, the home is different.”⁴⁸Frisby, 487 U.S. at 484 (citing Erznoznik v. City of Jacksonville, 422 U.S. 205, 210–11 (1975), and Cohen v. California, 403 U.S. 15, 21–22 (1971)). On the other hand, there will always be arguments for extending this sort of extra protection beyond the home to medical facilities,⁴⁹Hill v. Colorado, 530 U.S. 703, 718 (2000). funeral homes,⁵⁰Phelps-Roper v. Ricketts, 867 F.3d 883 (8th Cir. 2017); Phelps-Roper v. Strickland, 539 F.3d 356 (6th Cir. 2008). high schools,⁵¹Blythe v. City of San Diego, No. 24-CV-02211-GPC-DDL, 2025 WL 108185, at *4 (S.D. Cal. Jan. 14, 2025). places of worship,⁵²Id. at *1. and more. Here, I just want to acknowledge the difficulty that this issue raises.

  Conclusion

The one-to-one/one-to-many distinction is critical to understanding how and when unwilling listeners may be protected. I hope this short article has helpfully elaborated on a few questions the distinction raises.