Journal of Controversial Ideas

Arguments that try to justify free speech restrictions that go beyond the restrictions already recognized by even very permissive and libertarian accounts of free speech as well as by First Amendment law usually proceed by conjuring up all kinds of “harms” that speech considered to be protected might nonetheless produce. Thus, these harms are deemed so serious as to justify speech restrictions. Normative argument for the claim that they are sufficiently serious is seldom, if ever, produced; and the fact that free speech theory denies that serious harm is sufficient for justifying restrictions is almost always ignored.1 Moreover, credible evidence for the alleged serious harms is always lacking. Finally, while those arguments pose as viewpoint-neutral, their one-sidedness in assessing harms is conspicuous and appears to be designed for the very purpose of facilitating viewpoint-based censorship. The same one-sidedness is also evident in the fact that authors providing such arguments worry almost exclusively about the harms produced by free speech, but consistently ignore certain harms of suppressing or compelling speech.

The two recent arguments for further speech restriction critically examined here confirm this picture. Ann E. Cudd tries to make various kinds of clearly protected free speech responsible for “trauma.” However, she misrepresents the psychological studies she relies on and her account legitimizes anti-speech violence on a massive scale, which renders it morally absurd. Melina Constantine Bell tries to combine John Stuart Mill and psychological studies to argue that sexist and racial jokes and slurs produce severe harm and should therefore be restricted. Yet the studies are flimsy and the picture of Mill unrecognizable. I will address, as a much-needed corrective to the one-sidedness of those who warn about the alleged harms of free speech, the harms imposed by compelled speech, using the topical example of compelling people to use female pronouns for males who claim to be women. I show that this practice is abusive and wrongful. I conclude with a reminder about the nature of liberal democracy. Its raison d’être is not protection from harm per se but the safeguarding of freedom. There are no convincing reasons to further restrict or even compel speech, but every reason to defend free speech.

Another author intent on restricting speech due to its “harmfulness” is Melina Constantine Bell. There are sections of the article to be discussed here where Bell claims that racial or sexist slurs harm only minorities, or harm them more. She provides no evidence for this claim. There is also a section entitled “Systemic oppression and epistemic injustice.” This section repeats well-worn claims made by Critical Race Theory and by certain feminist theories that posit a “White patriarchy.” It neither provides any evidence to support its claims, nor engages with objections, nor, therefore, advances the debate. I will not deal with these sections.

However, there is also a section that promises to identify the “tangible, concrete harms” that certain forms of speech normally deemed to be protected inflict on “historically marginalized social groups” (independent of whether or not such harms can also befall majorities). The “harms” she refers to “involve ‘experience [that] is severe, prolonged, or constantly repeated’ such that ‘the mental suffering it causes may become ... incapacitating, and therefore harmful’” (mere taking offense does not count for her).23 In order to identify such harms, she relies, like Cudd, on psychological studies. Two such studies concern the effects of sexist humor. Bell summarizes the results as follows:

Sexist humor … can tangibly and directly harm women by eliciting depression, eating disorders, disruption of focused attention, appearance anxiety, and body shame. … In one study, a group of female participants exposed to sexist comedy skits expressed a greater state of self-objectification compared to women exposed to neutral comedy skits, while male participants’ ratings of self-objectification did not differ based on which skits they viewed. Other experiments found that women engaged in more body surveillance, … [which] can disrupt focused attention, usurping attentional resources and reducing performance on other cognitive tasks. In other research, women performed worse on math tests while wearing swimsuits, relative to a comparison group wearing sweaters. … One study … found that prior experimental “exposure to sexist jokes led to greater tolerance of the supervisor’s sexist behavior [in a vignette on workplace interaction] in comparison to exposure to neutral jokes or comparable non-humorous sexist statements.” Additionally, after watching sexist comedy skits, sexist men demonstrated significantly greater willingness to cut funding for a women’s organization, compared to other types of organizations.24

First, cutting funding is not even a harm in Bell’s own sense. Moreover, given that Bell purports to work within a Millian framework, it should be noted that utility might be increased by cutting funding for one organization and giving it to another. Money is a scarce resource. Second, it would have been preferable if the studies had been less biased in their methodology. There were no control groups composed of men against whom sexist humor was directed or of men who had to perform math tasks while wearing swimsuits. In short, the studies do not show that sexist humor or wearing swimsuits affects women more than men. Third, it would have been worthwhile to find out, for instance, whether a group of female participants exposed to the movies Erin Brockovich or Pretty Woman, where Julia Roberts constantly shows off her perfect figure in provocative outfits, also “expressed a greater state of self-objectification” compared to women watching, let’s say, a documentary on screw manufacture. If so, would that, in Bell’s opinion, constitute an argument for prohibiting such movies—or the women’s magazine Cosmopolitan, for that matter?

In that context, Bell also interprets one of the studies as showing that “sexist jokes imply that it is acceptable to make demeaning statements about women aloud, reinforcing notions that it is culturally acceptable to regard women as unequal,” and that therefore “women who hear these jokes might internalize their messages and self-objectify rather than rejecting or criticizing the jokes.”25 Maybe. But it is safe to assume that the same is true, and likely to an even greater extent, for statements about the inferiority of women that are made with religious authority, for example in a sermon. If so, would Bell consider this as a reason to restrict religious speech? And how compatible is that going to be with Mill’s liberalism—or with liberalism, period?

But putting these questions aside—what harm did the studies actually demonstrate? One study answers this as follows: “Our study contributes to this literature by demonstrating that sexist humor can trigger a transitory state of self-objectification in women, which has been shown to have far-reaching and detrimental consequences for women in daily life.”26 Transitory! But how “far-reaching and detrimental” can the consequences of this transitory state be? Sure, it can lead to transitory “disruption of focused attention, appearance anxiety, and body shame,” as Bell states—but calling this “far-reaching” appears to be an exaggeration precisely in light of the transiency involved. Moreover, there is no reason to assume that such a transient state leads to the depressions and eating disorders that Bell is talking about. The studies have certainly not shown this. In fact, one study explicitly admits: “No research to date has directly examined long-term emotional consequences of exposure to sexist humor at work.”27 In other words: there’s no evidence.

The same holds for the alleged effects on men’s willingness to “tolerate” sexist behavior. First of all, the “tolerance” refers to the test subjects’ assessment of a supervisor’s sexist behavior in a story on workplace interaction. However, it is quite conceivable that after having heard a joke about beating up bosses, test subjects will (perhaps still being in a jocular mode) show more “tolerance” toward a fictitious villain beating up his boss in a story about violent office interaction. Whether having heard such a joke will make a difference when they witness a colleague beating up their boss in real life is an entirely different matter. In other words, whether the experimental reaction under consideration is indicative of the reaction in real life would require further argument and evidence—which is missing. Second, tolerance of “sexist” behavior can at best cause or constitute harm if the sexist behavior itself is harmful. It does not automatically become harmful by calling it (not even by being) “sexist.” Here, too, further evidence would be required. Finally, the study actually derives the allegedly “increased tolerance for sexist behavior” from decreased ratings of the offensiveness of a supervisor’s sexist behavior.28 But this is a non sequitur as far as the behavior of the allegedly “tolerant” men is concerned. Men and women alike will probably find murder at the workplace more offensive than battery, but that hardly implies that they won’t object to both. Moreover, the alleged effect was only seen in men who were already “high in hostile sexism” to begin with,29 and for those men sexist humor does not appear to make a lasting difference: “Even among men high in hostile sexism, exposure to sexist humor did not affect the evaluative content of men’s stereotypes about women relative to comparable non-humorous disparagement or neutral humor. Thus, to date, there is no evidence that exposure to sexist humor uniquely affects stable, internal knowledge structures, such as stereotypes and attitudes toward women.”30

In short, none of the studies Bell refers to show that sexist jokes are more “harmful” than a mother telling her daughter to lay off the cookies and check out the Roberts diet. If the latter kind of speech isn’t harmful enough to warrant restriction, neither is the former.31

Bell also claims that “African Americans might be tangibly harmed by racist speech.”32 To back up this claim, she refers to a single study that she characterizes as an “overview”33 and interprets as concluding that “racial discrimination itself can cause direct physical harm in the form of tangible adverse health outcomes.”34 Actually, the study admits that “findings linking reports of racism alone to physical health—particularly among African Americans—have been mixed.”35 More importantly, our issue here is speech, and Bell concedes that the examples of the study do not refer to speech but rather to “discrimination in housing, policing, and medical care.” Yet she assures the reader that “it would be surprising if racist jokes and insults, which fall within this broad definition of discrimination, failed to affect African Americans in similar ways, especially given the demonstrated connection between sexist humor and harm to women.”36 Actually, in the light of the dissimilarities between these two sets of phenomena, this would not be surprising at all, especially given, as we saw, that the studies Bell referred to failed to demonstrate a connection between sexist humor and harm to women, let alone net harm to human beings in general (something Millians should be concerned about).37

Talking about Mill: using him of all people to argue for further speech restrictions is a rather bold move, but not a successful one. Bell states that “the main public purpose of free speech protections, for Mill, is to safeguard a public space for opinions to be shared, for debate to take place, and for rational and reasonable people to both argue for and amend their positions. Bigoted insults do not deserve protection under this rationale because they cannot reasonably be understood as opinions tendered for consideration and they are not answerable.”38 But why not? After all, Bell herself explicitly states that bigoted insults express an idea, namely: “You do not deserve my respect or regard; you are not my social or moral equal.”39 If Anne Frank had conveyed this message to Heinrich Himmler, she would most certainly have been right. Moreover, given that many current-day political theorists incorrectly believe in moral equality,40 denying moral equality, even by way of a slur, might stimulate debate and help people to amend their positions. Prohibiting the expression of such ideas of moral inequality, even, for that matter, of race-based inequality, would amount to precisely the viewpoint-based censorship that Bell distances herself from officially41 and that Mill rejects.

Perhaps Bell thinks the same ideas could be expressed without slurs. But, first, in arguments and debates (let alone in novels or poetry, which can also express ideas) form and content cannot so easily be separated.42 Second, it is not so clear what is so especially bad about slurs. Bell thinks the insult “N____ whore!” cannot be “answered.” Yet in many contexts it clearly can; what is less clear, however, is why one should think that calling someone a “prostitute of color” would be better. To be sure, Bell thinks that “bigoted insults are a form of cultural propaganda that ordinarily bypasses rational processing. Hearing an unsubstantiated statement repeated tinges it with familiarity, making hearers more likely to believe it whether or not it is true.”43 That is not uniquely true of “bigoted slurs,” though, but also of the claims made by anti-racist activists about the supposed racism of people who may or may not in fact be racist. Does Bell want to restrict those too?

Moreover, Bell greatly underestimates the benefits of free speech for “the oppressed” and overrates the importance that “bypassing of rational processing” has for Mill. She states:

Mill recognizes that writing a newspaper opinion that corn merchants are starvers of the poor is an importantly different matter from shouting the same opinion to an angry mob gathered at the merchant’s business or home … [E]xperimental evidence suggests that expressing views such as “women only have value as wives or sexual partners” is significantly less likely to cause harm to women than expressions of comparable views using humor, which bypasses rational consideration, as does an incendiary statement to an angry mob. … So … the manner of expressing … opinions could be shaped to optimize their potential contributions to rational discussion while minimizing their likelihood to harm socially vulnerable, disadvantaged, or oppressed people, whether the foreseeable harm is imminent or festering, caused simply or by aggregation of contributory causes.44

First, how is that going to be achieved? By having political discourse be moderated by a Victorian chaperone (“Oh yes, my dear, I know that you think that members of that group are degenerate fascist scum—but could you say that rather less stridently”)? Second, the suggestion sounds as though harm to people who are not “vulnerable, disadvantaged, or oppressed” does not count, which is rather biased. Moreover, even accepting the special status of such groups, it would appear that someone is going to have to decide a priori who they are, for if you allow for a controversial debate about this question, then they would still meanwhile be legitimate targets of slurs. After all, granting some but not other groups immunity during the debate would be to beg the question. And it would certainly be a mistake to have this question decided (or begged) by enemies of free speech, for they do not appear to be particularly prone to register all the evidence or to allow others access to it.

Bell is no exception. She says that the view (at least if expressed by a “white man”) that “men are naturally better at mathematics, … or the view that African-American poverty is a consequence of poor work ethic,” articulate “bigoted stereotypes.”45 One would assume that a view can only be “bigoted” if it is unsubstantiated. She provides no evidence that these views are, which indicates that it is not the white man, but perhaps she who expresses bigoted stereotypes, in this case of the leftist variety.46

Second, given Bell’s professed concern for “the oppressed,” she is remarkably negligent in appraising the benefits that free speech might have for them. She quickly dismisses the idea that “the civil rights movement depended on the freedom of activists to speak against the status quo,” by stating that the First Amendment “never protected the free speech of enslaved persons.”47 Putting aside the fact that the term “civil rights movement” does not refer to slaves, Amendments 4, 5, 6, and 7 also failed to “protect”—that is, they were not applied to—enslaved persons; however, to suggest that enslaved persons would have been made better off by getting rid of requirements for search warrants and legal procedure altogether would appear to be quite a leap. After all, abolitionists, that is, those trying to free the slaves, did benefit from the First Amendment (as well as from the other Amendments just mentioned). Of course, sometimes they were sent to jail for their abolitionist speeches anyway, but the First Amendment surely made that more difficult and thus benefited slaves indirectly by making it easier to promote the idea that slavery should be abolished. Bell also quotes Delgado’s and Stefancic’s claim that the First Amendment “is far more valuable to the majority than to the minority, more useful for confining change than for propelling it.”48 That is irrelevant. After all, suppression of free speech might also be more useful for “the majority” than for “the minority.” Obviously, therefore, the relevant question is whether, comparatively, suppression of free speech is more useful to those in power (whether they are a majority or not) than free speech. Common sense, logic, and history have provided a clear answer to this question.49

Third, and coming back to Mill, he nowhere suggests that what bothers him about someone shouting inciting utterances to an angry mob is that it “bypasses rational processing.” After all, Mill was not only opposed to manslaughter carried out in the highly emotional heat of the moment, but also against rationally calculated cold-blooded murder. Thus, what bothers him about uttering the opinion that corn-dealers are starvers of the poor in front of an angry mob, as opposed to expressing it in a newspaper article, is that the former utterance, in contradistinction to the latter, “constitute[s] ... a positive instigation of some mischievous act.”50 Of course, slurs can constitute such instigation under certain very special circumstances, but in most circumstances they simply won’t. To put it metaphorically: there is a huge difference between lighting a match in the gunpowder room and lighting a match on a park bench on campus. Mill was very well aware of such differences and, accordingly, Bell’s attempts to blur them unsurprisingly fail to provide a Millian case against free speech.

A conspicuous one-sidedness demonstrated by enemies of free speech is that while they constantly go on about the offense given or the harms produced by certain forms of free speech, they virtually never seriously consider the offensiveness and harmfulness of the suppression of free speech.51

To be sure, Bell does admit that “it might be just as important to the white students shouting insults that they are free to do so as it is to the African-American student to avoid those insults.” Yet she quickly assures the reader: “If these hypothetical persons’ interests are equally matched, however, the tie is easily broken by the other important dimension of free speech: the public interest in truth-seeking, equal concern for constituents, and aggregate human happiness. Public values line up squarely against permitting such an utterance in this context.”52 This reply is dismissive and entirely inaccurate.

First of all, “if these hypothetical persons’ interests are equally matched,” then the tie will necessarily be broken in favor of the majority insulting the minority,53 for with Mill’s utilitarianism we are seeking the greatest happiness for the greatest number, aren’t we? Second, Bell is engaged in double counting here. In the Millian framework, “aggregate happiness” is not a separate point but already comprises everything of moral importance, including the upholding of happiness-producing values.54 Third, we already saw that it is misguided (and rejected by the Supreme Court) to think that giving offense and thus using slurs cannot play a role in truth-seeking. Fourth, the state can demonstrate “equal concern for its constituents” by equally allowing its constituents free speech.55 Fifth, there are also quite a number of values on the side of not suppressing free speech, not even hate speech. One is honesty; and in fact minorities benefit from knowing what exactly others think about them instead of having to engage in guesswork: it allows them to make more informed decisions. Other authors have argued that suppression of free speech undermines the democratic legitimacy of laws that could not be verbally attacked as viciously before their implementation as their opponents wanted (and the attack can of course also be indirect via a verbal attack on the law’s supposed beneficiaries).56 Self-expression and autonomy are further important values, in particular in a liberal society, and speakers’ self-expression and autonomy would obviously be suppressed right along with their speech.57 Thus, Bell (and others) have not only failed to demonstrate that the harms they expansively connect with free speech are actually there; they have also failed to demonstrate that those harms are greater than the harms produced by the suppression of free speech.

Another blind spot of enemies of free speech when it comes to harms and offense concerns compelled speech. “Woke” academia and its followers in the wider culture are not only unduly relaxed about suppressing speech but also about compelling it. The harms and the offense that might inflict are never even mentioned.

Consider the scene in George Orwell’s novel 1984, where the henchman O’Brien forces Winston under the threat of torture (and using torture) to say—even to think—that two plus two is five. Obviously, the threat of torture and the actual use of torture here is a particularly grave wrong-making feature of O’Brien’s forcing Winston to say that two plus two is five. But it is not the only wrong-making feature. Imagine O’Brien were a famous and powerful male movie producer who tells the aspiring actress who unsuspectingly came to his hotel suite that he will ruin her career unless she says that two plus two is five. “But it isn’t.” “Yes, it is, and I want you to say it, or you can forget your Hollywood dreams. Say it!” Intimidated, she does say it. This man is a swine. He abuses his power and violates the actress’s freedom of conscience and her freedom of speech by forcing her to say something that she knows to be wrong (it would still be swinish if she didn’t know it but only believed it to be wrong). The situation doesn’t improve if the abuse of power stems from a run-of-the-mill employer, from a university administration, from newspapers and journals, or a powerful media platform (think Twitter) threatening to throw anybody off who doesn’t say “Two plus two is five” when addressing people who idiotically have the statement “Two plus two is five” on their profiles.

Yet, allegedly, the wrongfulness miraculously disappears—poof!—if what the male film producer, the employer, or the media platform demand under threat of these sanctions is this: that people refer to male persons who declare themselves to be women with the female pronouns “she “or “her.” (Suppression is also involved, of course: sanctions ensue if one correctly calls the “transidentifying” male a man or uses his birth name in the case where he changed it to a female name). But the wrongfulness doesn’t disappear: it stays right there.58 In English language, confirmed by dictionaries, the term “woman” designates adult human females. Males aren’t females. Accordingly, males identifying as a woman aren’t women, period.59 Thus, for powerful actors to force people to refer to men claiming to be women with female pronouns is to force them to deny truth and the reality right in front of their eyes; it is a rights-violating abuse of power. It also produces offense and stress and potentially harm for those who are subjected to this abuse of power. Moreover, if this film producer is a special type of “transidentifying” male, namely what psychologists and sexologists researching gender dysphoria and transsexualism have called an autogynephile (a man “aroused by the idea of being a women”60), then the actress’s referring to him with female pronouns might well sexually arouse him (and explain his motivation for his wrongful action in the first place).61 His acts would then also amount to sexual harassment. Moreover, if we lower our standards as to what counts as “violence” in the same way as many enemies of free speech do (and certainly transgender ideologues who constantly claim that “misgendering is violence”62), then, by the same standards, his act would be sexual assault. In addition, internet platforms enforcing a policy requiring the use of female pronouns to refer to males if those males “prefer” this would not only aid and abet sexual harassment (as they currently do), but indeed sexual assault committed by a tiny majority against a vast majority (talk about “aggregate happiness”). This sacrificing of the rights and interests of a vast majority for the satisfaction of a tiny trans minority amounts to trans supremacism. Apparently, sometimes oppression is quite all right.

While enemies of free speech like to constantly take themselves to be detecting all kinds of “offense” given or “harms” produced by free speech, the blurred vision if not blindness regarding the harms and rights-violations involved in suppression of speech and compelled speech is conspicuous. The diagnosis: thoughtlessness or hypocrisy, probably a combination of both. The cure: free speech. It is the medium in which moral integrity and thoughtfulness thrive best.

It is also the tool through which we preserve freedom, the value at the heart of a liberal democracy. In such a system people are, for good reasons, protected against severe rights-violations but not against being offended or even against being harmed, even severely harmed. In particular, they are not—certain limits notwithstanding—protected against being harmed (by going out of business or losing a job) through the achievements of competitors (market economy, an element of liberalism); they are not protected against being severely harmed by laws passed (democracy); and they are not protected against being harmed by free speech (another element of liberalism). Indeed, while it is possible to give people equal freedoms, equal rights to engage in potentially harmful acts (like opening a restaurant that will ruin the competition or making a joke leading to transitory self-objectification), it is not possible to equally protect them from harms. Yes, you can equally protect them from slurs and insults, but the very act of suppressing slurs and insults is to harm (that is, set back the interests of) those who would rather receive insults than be kept from uttering them, and it harms them for the benefit of others. There is no equality here. Moreover, modern liberal democracies have not only done better than other regimes in securing the freedom of their citizens, but also in protecting them from the very severe harms regularly suffered in failed states or left- or right-wing authoritarian regimes. Free speech, however, is indispensable for sustaining liberal democracies. Therefore, anyone who is really concerned about avoiding overall net harms had better protect free speech. That those concerned about liberty should do so goes without saying.63