Freedom of Speech
Summary and Keywords
John Stuart Mill is a liberal icon, widely praised in particular for his stirring defense of freedom of speech. A neo-Millian theory of free speech is outlined and contrasted in important respects with what Frederick Schauer calls “the free speech ideology” that surrounds the First Amendment of the U.S. Constitution, and with Schauer’s own “pre-legal” theory of free speech. Mill cannot reasonably be interpreted to defend free speech absolutism if speech is understood broadly to include all expressive conduct. Rather, he is best interpreted as defending an expedient policy of laissez-faire with exceptions, where four types of expression are distinguished, three of which (labeled Types B, C, and D) are public or other-regarding, whereas the fourth (labeled Type A) is private or self-regarding. Types C and D expression are unjust and ought to be suppressed by law and public stigma. They deserve no protection from coercive interference: they are justified exceptions to the policy of letting speakers alone. Consistently with this, a moral right to freedom of speech gives absolute protection to Type B public expression, which is “almost” self-regarding. Type A private expression also receives absolute protection, but it is truly self-regarding conduct and therefore covered by the moral right of absolute self-regarding liberty identified by Mill in On Liberty. There is no need for a distinct right of freedom of expression with respect to self-regarding speech. Strictly speaking, then, an expedient laissez-faire policy for public expression leaves the full protection of freedom of private expression to the right of self-regarding liberty.
An important application of the neo-Millian theory relates to an unjust form of hate speech that may be described as group libel. By creating, or threatening to create, a social atmosphere in which a targeted group is forced to live with a maliciously false public identity of criminality or subhumanity, such a group libel creates, or significantly risks creating, social conditions in which all individuals associated with the group must give up their liberties of self-regarding conduct and of Type B expression to avoid conflict with prejudiced and belligerent members of society, even though the libel itself does not directly threaten any assignable individual with harm or accuse him or her of any wrongdoing of his or her own. This Millian perspective bolsters arguments such as those offered by Jeremy Waldron for suppressing group libels. America is an outlier among advanced civil societies with respect to the regulation of such unjust hate speech, and its “free speech ideology” ought to be suitably reformed so that group libels are prevented or punished as immoral and unconstitutional.
Freedom of speech is declared a human right by multiple influential international documents and institutions. Its status as a human right is fairly recent, however, and far from universally recognized. Even in what are now considered advanced liberal democracies, the predominant view of freedom of speech from the days of John Milton’s Areopagitica (1644) into the early 20th century was arguably the absence of prior legal restraints on publication, with the understanding that published books, pamphlets, and other forms (including oral forms) of public expression could legitimately be punished by the government or ecclesiastical authorities for seditious libel, blasphemy, obscenity, and the like.
Sharp disagreement still exists within and between advanced democracies over the nature and scope of any moral right to freedom of speech. Should the right be seen as absolute, as some libertarians argue? Or should it be limited in various ways for the sake of other social values such as the public good? Should it or should it not give protection to so-called “hate speech,” for example? What sort of right is the right to free speech? Is it an immunity from government interference? Is it a claim not to be interfered with by other people so that others have duties not to prevent the speaker from speaking? Or is it a naked permission to speak such that other individuals, including public officials, are free to compete and perhaps even shout down the speaker? Or is it some complex combination of immunities, powers, claims, and permissions? Is speech a synonym for expression? Are words essential to speech so that speech must be distinguished from expressive conduct not involving words? Or is a symbolic action such as burning the flag in protest against government policy properly considered speech no less than speech acts in which words are used to communicate ideas? Does every human action have an expressive component? Or are some actions mute, so to speak? Answers to these and related questions remain controversial and continue to be debated.
A theory of freedom of speech, rooted in John Stuart Mill’s philosophy as I interpret it, will be sketched in which speech is equated with expression and understood to encompass all human conduct: every human action is taken to have an expressive aspect, although different people may take different messages from a given action. Mill is a widely celebrated proponent of freedom of speech, and he clearly defends, in chapter 2 of On Liberty, absolute liberty of thought and discussion. While many competing interpretations of his philosophy exist, the theory of freedom of expression as sketched here is argued to be compatible with his texts and has considerable appeal in any case. Among other things, it allows us to see how his defense of absolute liberty of thought and discussion can be reconciled with prevention and punishment of wrongful types of expression, including credible threats of harm, incitements to violence, malicious libel, and even wordless criminal actions with an expressive component, that do not count as discussion.
Some important contrasts are stressed between the neo-Millian theory and what Frederick Schauer (1992) calls the “free speech ideology” that surrounds the First Amendment of the U.S. Constitution. The American “ideology” continues to be developed by the U.S. courts in an ongoing process of constitutional interpretation that includes criticism of court decisions by various commentators. Roughly speaking, the ideology as developed so far seeks to give heightened (virtually absolute) protection to so-called political speech in order to foster robust democratic debate, with the caveat that some reasonable time, place, and manner restrictions are recognized as legitimate. As a result, the United States, unlike other advanced democracies, is very reluctant to suppress hate speech absent a clear showing that the speech poses an imminent threat of severe harm to determinate individuals, because hate speech is deemed political in content. Types or categories of speech whose content is not recognized as political (or perhaps as sufficiently political) receive less-heightened protection (for example, commercial speech, artistic expression) or no protection at all (for example, child pornography, copyright infringement).1
Some important contrasts are also drawn between the neo-Millian theory and Schauer’s own “pre-legal” theory of freedom of speech as presented in Schauer (1982) and subsequently elaborated in a series of works. Numerous other influential theories exist, such as those proposed by C. Edwin Baker (1989), Catherine V. MacKinnon (1993), Ronald Dworkin (1996, 2006), T. M. Scanlon (2003), and Seana V. Shiffrin (2014), which cannot be examined carefully here. In any case, Schauer’s theory is sophisticated and, moreover, he relates it to Mill’s theory as he interprets Mill. The contrasts identified between the readings of Mill should help to clarify what is distinctive about the neo-Millian theory presented.
Mill cannot reasonably be interpreted as a free speech absolutist if speech or expression is understood broadly. Rather, he is best interpreted as defending an expedient social policy of laissez-faire with exceptions, where four types of expression are distinguished, three of which (labeled Types B, C, and D) are public or other-regarding, whereas the fourth (labeled Type A) is private or self-regarding. Public expression directly causes non-consensual harm to others, or poses a significant risk of doing so. Harm is any form of perceptible damage to a person’s interests, including damage to her body, reputation, finances, contractual or other legitimate expectations, and so on. It excludes mere dislike, disgust, or disdain. Unlike public expression, private expression does not cause harm to others or, if it does, only with their genuine consent and participation. Others may be disgusted by public as well as by private expression but their disgust is distinct from harm in the sense of perceptible damage.
Types C and D public expression cause wrongful harm to others. They are unjust types of speech and, as such, ought to be suppressed by law and public stigma. They constitute justified exceptions to the policy of letting speakers alone. Consistent with this, a distinctive moral right to freedom of speech gives absolute protection to Type B public expression, which is “almost” self-regarding. Type B public expression does cause non-consensual harm to others, but the harm is reasonably judged permissible by society or its representatives because the benefits of the expression outweigh the harms. Type A private expression also receives absolute protection, but it is truly self-regarding conduct and is therefore covered by the moral right of absolute self-regarding liberty defended by Mill in On Liberty. There is no need for a distinct right to freedom of expression with respect to self-regarding speech. Strictly speaking, then, an expedient laissez-faire policy applies only to public expression and leaves the absolute protection of freedom of private expression to the right of self-regarding liberty.
Certain features of the neo-Millian theory are highlighted that distinguish it from both Schauer’s pre-legal theory and the American constitutional free speech ideology. In particular, the Millian theory emphasizes that unjust public expression of Types D and C always wrongfully interferes with other assignable individuals’ moral rights of liberty with respect to self-regarding conduct and Type B speech. A speaker who engages in unjust types of speech necessarily fails to fulfill his moral duty to allow other identifiable individuals to freely live their own lives in accord with their own judgment and inclinations, short of wrongful harm to others.
The neo-Millian theory is applied to the vexed problem of hate speech. Unlike the American free speech ideology, the Millian theory maintains that group libels, that is, public expressions of hate that brand the members of a particular ethnic, racial, religious, or sexual orientation group as intrinsically evil or unworthy of equal respect, are unjust and ought to be suppressed. By creating a social atmosphere in which the targeted group is forced to live with a maliciously false public identity of criminality or subhumanity, such a group libel violates the moral rights of self-regarding liberty and of freedom of Type B speech of every individual associated with the group, even though the libel does not directly threaten any assignable individual with harm or identify him or her in particular as deserving of punishment for any wrongdoing of his or her own. This Millian argument bolsters other arguments such as those offered by Jeremy Waldron (2012) for suppressing group libels. America is an outlier among advanced civil societies with respect to the regulation of unjust hate speech, and its “free speech ideology” ought to be suitably reformed so that group libels are prevented or punished as immoral and unconstitutional.
New challenges for freedom of speech in the age of the Internet are also briefly discussed.
Speech in public is conduct that can harm others without their consent. Schauer emphasizes this point: “Speech is plainly not a self-regarding act, even assuming there be a category of acts that are self-regarding” (Schauer, 1982, p. 10). While he deliberately leaves open the question of what harm means, he illustrates the other-regarding nature of speech by indicating some of the numerous ways in which it causes non-consensual harm to others, to wit: a speaker maliciously spreads lies that damage another’s reputation; or he publishes a true account of his intimate relations with a former lover without her consent so that his revelations, though true, constitute a gross invasion of her privacy; or he uses a public forum to attack the high price of bread and incites a mob to burn down another’s bakery; or he publishes an unfavorable review that discourages people from going to a new play, thereby causing financial damage to the producers; or he intentionally reveals secret military plans about troop movements, allowing the enemy to slaughter some of the troops; or he recklessly publishes false charges of political corruption that impair the operation of government and even endanger national security.
Schauer goes on to point out that Mill apparently agrees with him: “A close reading of On Liberty reveals that, as between self-regarding and other-regarding acts, Mill treats speech as a member of the latter category” (Schauer, 1982, p. 11; see also Schauer, 2011, 2015a, pp. 134–135). Mill does say that “expressing and publishing opinions . . . belongs to that part of the conduct of an individual which concerns other people” (Mill, 1977, pp. 225–226). So, even for him, public expression is not properly classified as self-regarding conduct that, by his definition, doesn’t directly cause harm to others or does so “only with their free, voluntary, and undeceived consent and participation” (Mill, 1977, p. 225). Thus, Schauer is convincing when he claims that any principle of free speech must be distinct from a principle of self-regarding liberty: “If there is a Free Speech Principle, it protects certain conduct not because it is self-regarding, but despite the fact that it is other-regarding” (Schauer, 1982, p. 11, emphasis original).
Nevertheless, even though it is other-regarding conduct, Mill brings public expression within the ambit of his principle of self-regarding liberty, according to which any competent individual has a moral right to choose among his self-regarding actions as he pleases: “In the part [of his conduct] which merely concerns himself, his independence is, of right, absolute” (Mill, 1977, p. 224). As it stands, this is a puzzling move. If expression really were self-regarding conduct, then the competent individual would necessarily have a moral right to express himself as he wishes on all subjects at any time or place. And Mill does insist that the liberty of expressing and publishing opinions is “practically inseparable” from the liberty of thought itself, which is undeniably a self-regarding liberty and thus, “of right, absolute.”2
Even so, Mill slips in the caveat that the liberty of expression is “almost of as much importance as the liberty of thought” and “resting in great part on the same reasons” (Mill, 1977, p. 226, emphasis added). This suggests that he may see freedom of expression as “almost” absolute but not truly absolute. Or, to save the claim that freedom of expression is absolute, perhaps he means expression in some restricted sense such as discussion of opinions. He might take for granted that his readers understand that he is referring, for example, only to expression in the sense of speech acts, that is, oral or written communications involving words, and even then, only to a proper subset of them, to wit, speech acts comprising discussion of scientific, moral, theological, political, or aesthetic ideas and theories. It remains to be seen, however, whether expression even in some such restricted sense can plausibly be treated as if it is self-regarding conduct. In any case, Mill’s view requires clarification because it cannot seriously be maintained that freedom of expression broadly understood is, by right, absolute.
The puzzle of Mill’s assimilation of public expression to self-regarding conduct can be restated in terms of the expression’s different consequences for others. If it were really self-regarding, expression might cause others to feel offended or disgusted but it couldn’t cause them any non-consensual harm. Yet there is no doubt that speech often does cause non-consensual harm to others. This way of stating the puzzle is obscured a bit by Schauer’s refusal to commit to a definition of harm. His refusal reflects his position that “speech can cause the kinds of effects that would count as a harm under almost any theory” (Schauer, 1982, p. 218, note 1). As a result, he is apparently ready to accept a very broad notion of harm as any negative consequence, including mere displeasure, offense or emotional distress (see, for example, Schauer, 2003, pp. 646–647). By contrast, as already mentioned, Mill apparently works with a more restricted idea of harm as any form of perceptible damage to external objects of concern to competent individuals, including their bodies and reputations, for example, but also public goods that they have jointly produced through their individual contributions. This empirical idea, while still broad, excludes mere dislike, disgust, disapproval, emotional upset, or other negative attitude that is unaccompanied by any evidence of perceptible injury.
“I am not bound to abstain from an action because another person dislikes it,” Mill asserts, “however he may dignify his dislike with the name of disapprobation” (Mill, 1969a, p. 195). If there is no perceptible damage, he seems to be saying, there is no justification to prevent the mere dislike or to punish anyone for causing others to suffer it. Instead, because there is no perceptible injury obstructing them from reacting as they wish, competent others are reasonably expected to freely avoid what they dislike if they judge such avoidance necessary for their own good. Their equal right of self-regarding liberty includes a right to avoid any agent who, despite their advice and warnings, persists in self-regarding conduct that displeases them. By persisting in his disliked conduct, the agent implicitly consents to the loss of their friendship and other “natural penalties” that are inseparable from their dislike (Mill, 1977, pp. 277–279).
But this cannot be said of other-regarding conduct like public expression that causes non-consensual perceptible damage to others, because such damage does obstruct them from freely avoiding speech they dislike. So it is surprising that Mill seems prepared to treat speech as if it were always self-regarding. He cannot pretend that it only causes others to feel mere dislike or disgust as distinct from harm. This is not to say that it doesn’t cause dislike. After all, it causes harm and there is no denying that harm frequently leads to dislike. Perceptible damage or danger of same is a reason for feeling dislike or distress. Public speech can directly cause many forms of perceptible damage or danger of same, including damage to reputation, financial loss, and even physical injury through incitements. Thus, Schauer (2003) is on solid ground when he casts doubt on the soundness of “the lesser harm hypothesis,” according to which speech can be shown on the basis of observation to cause a smaller quantity of harms or a less severe kind of harm than non-speech conduct causes to others. His argument is convincing, even if harm is distinguished from mere dislike.
At the same time, the fact that expression as well as non-expressive conduct causes harm to others doesn’t imply that the harm is always disliked or that prevention of it is always justified. According to Mill, a competent person’s reason for feeling dislike is defeasible and overridden by his genuine consent to the perceptible damage, for example, and even non-consensual harms are permissible if they are outweighed by social benefits. To justify coercive interference with any type of conduct, he implies, the conduct must be shown to cause non-consensual harm to others that is wrongful or unjust.
Justice and Equal Rights
This is not the place to discuss Mill’s (1969b) unusual utilitarian liberal theory of justice in detail, beyond saying that he conceives of justice in terms of an optimal social code that distributes and sanctions universal equal rights and duties that competent judges, or majorities of them in case of disagreement, decide are best for protecting the vital interests shared by the members of society.3 Ideally, the relevant majorities would be moral agents with the capacities needed to determine optimal rights that maximize security for everyone’s vital interests. But Mill seems content to leave the determination to actual democratic majorities, even though their moral sentiments are contaminated and distorted by their selfish interests, so long as the democratic process involves suitable checks and balances designed to prevent abuses of power and encourage respect for minorities’ rights (Mill, 1977, p. 283; Riley, 2008). The one notable exception is the right of self-regarding liberty, which Mill insists actual majorities have failed to recognize because of their tendency to dislike diversity and force the individual to pursue her own life as most people pursue theirs. In his view, the construction of an optimal code of justice is by far the most important step that must be taken to promote the general happiness. Violations of any individual’s optimal moral rights, many of which are already discernible, including the right of choosing among self-regarding actions as one pleases, can never be outweighed by competing social benefits.
Coercive interference with speech can only be justified, Mill implies, by showing that the speech in question violates others’ moral or justified legal rights. If it is treated as if it is self-regarding conduct, however, expression can never violate others’ rights and thus is, “by right, absolute.” Self-regarding expression is evidently covered and given absolute protection by the right of self-regarding liberty itself.
As it stands, Mill’s puzzling assimilation of public expression to self-regarding conduct seems contradictory. It has been argued elsewhere (Riley, 2005, 2008, 2015a), however, that he is best interpreted as endorsing an expedient social policy of laissez-faire with respect to “expression and publication of opinions.” Such a policy protects a wide range of public expression even though it is other-regarding conduct. The policy is roughly analogous to a laissez-faire policy with respect to economic exchange or trade; Mill says that trade, like expression, is “social” or other-regarding conduct (Mill, 1977, p. 293). Unlike the moral right of absolute self-regarding liberty, expedient laissez-faire policies for other-regarding actions such as public expression or trade don’t grant absolute liberty but instead contain exceptions such that coercive interference is legitimate if the other-regarding conduct causes a grievous type of non-consensual harm to others that is reasonably declared unjust by society or its representatives. Thus, the laissez-faire policy is constrained by optimal rules of justice that authorize exceptional uses of coercion to prevent or punish violations of rights.
Two Kinds of Non-Consensual Harm
An expedient laissez-faire policy must recognize that public expression can cause two different kinds of non-consensual harm to others. One kind is permissible. Included in this kind is what might be termed fair competitive harm, that is, non-consensual harm caused by a winning speaker to his losing competitors in an open public discussion or debate, where the participants are free to present and argue for their various ideas and opinions: the losers don’t consent to lose support from the audience but instead are forced to face the fact. Similarly, published criticism of the merits of artworks that are competing for an audience, such as books, films, paintings, and plays, is permissible even if it causes financial loss to the creators. Expression that causes only fair competitive harms is properly deemed permissible because the harms are thought to be outweighed in the aggregate by the social benefits of freedom for that sort of speech, including discovery and lively appreciation of warranted opinions, promotion of technological advance, exposure and criticism of government corruption, promotion of individuality or self-development, reassessment and reform of existing social customs, and so on (Mill, 1977, pp. 228–275; Schauer, 1982, pp. 15–86; Garton Ash, 2016, pp. 73–114). As Mill puts it, “society admits no right, either moral or legal, in the disappointed competitors, to immunity from this kind of suffering” (Mill, 1977, p. 293).
Also permissible is non-consensual harm caused by speakers to particular individuals to prevent or punish their unjust actions, as when white lies are told to deceive the Nazi searching for the Jew hiding in the house, or when fair sentences are duly imposed by judges on convicted criminals. Such speech is essential to maintain an optimal code of justice which, as remarked earlier, is, for Mill, the most important step that must be taken to promote general utility.
But expression can also cause another kind of non-consensual harm that is impermissible because it wrongfully harms others. Such speech is unjust. It includes telling the truth to the Nazi, for example, and giving a suspended sentence to a convicted serial killer. Much of it, however, causes what may be called unfair competitive harms. As Mill continues in the passage already quoted, society “feels called on to interfere [in a competition], only when means of success have been employed which it is contrary to the general interest to permit—namely, fraud or treachery, and force.” Speech that causes unfair harm or credible risk of same to other speakers, creative artists, or third parties must be prevented or punished for public discussion and criticism to be fair. Such speech includes malicious libel, gross invasion of privacy, and credible threats of violence against other speakers, for example, as well as speech that is delivered in a time, place, or manner that violates others’ rights not to be awoken in their homes in the middle of the night or to be the victims of incitements to wrongful actions. For a fair competitive marketplace of opinions to exist, speakers must obey social rules of justice that prevent or punish such unjust expression, just as sellers in a fair competitive economic market must obey rules of justice that exclude force, fraud, production methods that cause undue pollution, and so on.
As it turns out, the laissez-faire approach also allows us to see under what conditions public expression can be understood as “almost” self-regarding so that Mill’s otherwise puzzling move is vindicated. More specifically, liberty of Type B public expression is legitimately treated as, by right, absolute.
A Neo-Millian Framework
An expedient laissez-faire policy for expression can be clarified by developing and sharpening the foregoing remarks and integrating them with Schauer’s analysis of a right in terms of its coverage and protection (Schauer, 1982, pp. 89–92, 131–136). To do this, we need to address more carefully the issue of what counts as expression or speech and then indicate the reasons for giving different degrees of protection to different types of expression (Schauer, 1982, pp. 93–153, 201–206). It is convenient to organize the ensuing discussion in terms of Figure 1.
The Concept of Speech
What is meant by speech or expression? One answer is to define it broadly to include not only oral and written communications with words but also symbolic actions (such as burning the flag or defacing a public monument with red paint) that have a communicative or speech aspect, even if words are not involved, conjoined with a noncommunicative or non-speech aspect.
For an action to count as symbolic, Schauer asserts that the agent must intend to communicate a message perhaps accompanied with strong sentiments: “Any coherent formulation of a Free Speech Principle requires communicative intent as well as a perceived message . . . Without communicative intent, a communicated message, and a recipient of the communication there is no complete communicative act, and no occasion to talk about freedom of speech” (Schauer, 1982, p. 98). He then suggests that the speech and non-speech aspects of symbolic conduct can be separated and balanced in importance one against the other when deciding whether to protect the conduct. Such balancing is often employed by the U.S. Supreme Court to decide cases and controversies, and may be viewed as an important element of the constitutional doctrine of free speech, which American courts continue to develop through their interpretations of the First Amendment.
Under this balancing approach, symbolic conduct can legitimately be prevented or punished when the court decides that the conduct’s non-expressive aspect causes non-consensual harm to others (i.e., to society as represented by government) that outweighs the social value of the conduct’s communicative aspect, despite the incidental suppression of speech that goes along with this. It might thus be argued that the government has a compelling interest in interfering with flag burning or defacement of monuments, for example, even though the individual’s intention to send a political protest message will be suppressed as a result. True, given that freedom of speech is a positive good, a heightened justification for coercive interference with symbolic conduct seems necessary because of the unusual social benefits associated with the communicative aspect: “recognizing a Free Speech Principle entails establishing a burden of governmental justification greater for those governmental actions directed at communication than for governmental actions directed at other forms of conduct” (Schauer, 1982, p. 141; see also Schauer, 1984, 1992, 2015a, 2015b). In other words, the interference would not be justified if the noncommunicative aspect caused a similar degree of non-consensual harm to others as would be caused by a non-symbolic action, ceteris paribus. Notice that this presupposes that the speech associated with the communicative aspect is of a type that confers net social benefits, that is, it isn’t the sort of speech that is itself unjust.
There are a number of contentious features of this influential approach to symbolic conduct. First, it isn’t clear that the two components of symbolic conduct are separable. For example, it may be that burning the flag or defacing a monument is essential to the message that an individual intends to communicate to others: that way of delivering the message communicates the passion with which he intends to send it. Given that symbolic conduct consists of a message inextricably bound up with a non-speech component, it is improper to prevent or punish symbolic conduct unless the action as a whole is recognized by competent judges as causing wrongful harm to others. If the action as a whole is unjust, however, then coercive interference with the message that is inseparable from it is justified. Because speech acts per se can cause wrongful harm to others, there is no call to treat symbolic actions as requiring a different kind of justification than other actions require for coercive interference. Unjust actions cannot be justified or excused by virtue of the fact that messages, including those conveyed by speech acts, are involved in their performance.
Second, if the two components aren’t separable so that symbolic conduct is just a form of expression that is bound up with a non-communicative component, defining speech in terms of the agent’s intent to express himself to others doesn’t allow us to distinguish symbolic conduct as a proper subset of all voluntary conduct. As Schauer remarks: “The problem is that all voluntary conduct is self-expression” (Schauer, 1982, p. 93, emphasis original). Given that speech acts are only a form of self-expression, and not necessarily the most important form to the agent who may assign more expressive value to his attire, occupation, sexual activities, residence, and so on, the argument from self-expression turns a principle of free speech into a general liberty principle: “the argument from self-expression leads to the conclusion that all forms of self-expression are worthy of equivalent protection” (Schauer, 1984, p. 1291). If a free speech principle covers all voluntary conduct as expressive and thereby reduces to a general principle of liberty, however, then, according to Schauer, there is no reason to give “speech” heightened protection or to isolate a special principle of free speech. Given that he himself worries that there isn’t any justification for a special principle of free speech, however, this conclusion might turn out to be reasonable. It remains to be seen whether his worries on this point are well-founded.
Third, the agent’s intention to communicate a message that others will receive may not even be an essential feature of symbolic conduct. Those who witness the conduct or its consequences may take messages from it not intended by the agent. For example, ideas and sentiments may be suggested by paintings, sculptures, and music that the artist does not intend to communicate, and opinions and beliefs may be read into murders and other crimes that did not motivate the criminal. If symbolic conduct doesn’t depend on the intentions of the agent being understood by others, however, this strengthens the conviction that expression includes all voluntary conduct, even conduct that others interpret as sending them messages that the agent doesn’t intend to send. Thus, it makes sense to treat the right to freedom of expression as a complex right that covers all voluntary conduct, although universal coverage does not imply universal protection from coercive interference.
The upshot is that Figure 1 may be taken to represent all voluntary conduct, including speech acts using words, on the assumption that all voluntary conduct is expressive. Notice that this renders moot any distinction between expression and conduct. It is of no practical importance whether we label a particular action as speech or non-speech conduct. While this may seem counterintuitive, it will prove convenient because it allows us to sidestep much fruitless controversy over the problem of distinguishing speech from action. As Schauer (1982, 1984, 1992, 2003, 2015a, 2015b) has long argued, it is difficult if not impossible to identify a proper subset of conduct called expression or speech that has some unique feature that justifies treating it as special in comparison to non-speech conduct in the sense that a more demanding test must be passed before coercive interference with speech is legitimate. At the same time, the assumption that all voluntary conduct might be treated as expressive doesn’t commit us to any conclusion that a distinctive Free Speech Principle is impossible.
Different Degrees of Protection
The protection given by a right to freedom of expression, understood to cover all voluntary conduct as expressive, arguably comes in only two degrees, namely, absolute protection or no protection. There seems no convincing justification for intermediate degrees of protection. Speech is either fully protected because it doesn’t cause any wrongful non-consensual perceptible damage to others, or it is not protected at all because it does. In terms of Figure 1, an expedient laissez-faire policy for expression leaves speakers alone to freely engage in Types A and B speech, but it prevents or punishes speech of Types C and D.
Different Types of Expression
Type D expression is unjust expression that receives no protection because it causes, or poses a significant danger of, wrongful harm to others by virtue of its content, whatever the time, place, or manner. This type of speech always violates or endangers the moral or justified legal rights of others, or wrongfully damages or endangers the interest of the public at large. It includes speech acts such as malicious libel, gross invasion of privacy, commercial fraud (false labeling or advertising of products), credible threats of violence, and disclosure of military secrets. Also included are speech acts or visual displays that are essential ingredients of certain criminal actions, such as blackmail, theft of copyrighted materials, sexual abuse of real children or animals in films or photographs (child or animal pornography),4 murder or robbery carried out by one person on the orders of another, vandalism of government property or of another’s private property (although burning one’s own flag or defacing one’s own house with red paint is permissible), and the like. The use of words or pictures cannot justify or excuse the crime, or exempt from punishment anyone who participates in it.5 Even if words or pictures are not used, unjust expression is an inseparable component of criminal actions in which the agent intends to send a message, such as a message of hatred for a murder victim because of the victim’s race, sex, religion, or ethnic background, but also a message of hatred because of the victim’s relationship to the killer such as spouse or even fellow member of the human species. And even if the wrongdoer doesn’t intend to send it, others’ taking such a message from his criminal action can exacerbate but cannot alter its injustice.
Type C expression is unjust expression that receives no protection because it causes wrongful harm to others as a result of its time, place, or manner of delivery, even though in other contexts speech acts of the same content, or actions in which the same words or pictures or expressive intentions play an indispensable role, are not unjust but instead are protected. This type of speech violates the rights of others in certain contexts whereas in other contexts it only causes permissible non-consensual harms to others. In other words, the expression is an inseparable part of a wrongful action in the one case whereas it isn’t in the other case. Any regulations of it should be crafted to make as clear as possible the contexts where it is prohibited. It includes instigating “an excited mob assembled before the house of a corn-dealer . . . to some mischievous act” by voicing the opinion that “corn-dealers are starvers of the poor” or by handing out placards on which is written the opinion that “private property is robbery,” for example, even though the same opinions “ought to be unmolested when simply circulated through the press” (Mill, 1977, p. 260). Another example is broadcasting opinions on art or religion very loudly in a residential neighborhood at times when people have a right not to be disturbed so they can rest, even though the same opinions “ought to be unmolested” when published in a magazine or posted on the Internet. Similarly, raucous political protest marches, demonstrations, and “occupations” are allowed when conducted in accord with licenses issued by the relevant authorities, and may even be tolerated to some extent in the absence of licenses, but at some point are legitimately terminated and punished because at that point they are competently judged to impose wrongful harms and inconveniences on other people. Many further examples might be left to common sense without any need for explicit regulations. Expressing oneself by wearing a mask while carrying an unloaded old firearm may be permissible as an invited guest at a Halloween party in a private home but “may justly incur punishment” when performed in a bank during working hours.
Type B expression never causes any wrongful harm to others and thus ought to be protected. It consists of speech acts and any other voluntary actions whose expressive elements are not of an unjust content, delivered in times, places, and manners that respect the moral rights of others. As already indicated, it includes robust discussions of ideas and doctrines as well as published criticisms of artworks. Also included are published cartoons, lampoons, satires, and the like. These may damage the amour propre of a humorless egoist or of his followers. But such expression cannot harm the self-esteem or violate the rights of anyone with a sense of humor, even if he dislikes the expression. Permissible too are advice and warnings of danger “obtruded” on a reluctant individual by speakers who are uncertain that the individual knows what he is doing (see Riley, 2017b).
Type B expression causes non-consensual but permissible harms to others, and so it is not self-regarding conduct. Speakers who are predominantly motivated by self-interest to win a fair competitive debate and thereby gain a wider audience for their own ideas and opinions can hardly be expected to consent to be losers. Nor can ambitious creative artists be expected to consent to harsh criticisms of their works and the loss of popular support likely to follow. Similarly, humorless egoists will not consent to tolerate disrespectful cartoons or lampoons of those whom they admire or revere. And Nazis and convicted criminals cannot be expected to consent to being deceived or sentenced for their crimes. So, Type B speech remains other-regarding conduct, even though it doesn’t cause any wrongful harm to others. It still seems incorrect for Mill to treat even this type of speech as “almost” self-regarding. How can he justify doing so?
Type B expression is plausibly treated as “almost” self-regarding because we think that others ought to consent to suffer the permissible non-consensual harms, even though they refuse to consent out of self-interest. By contrast, we don’t think that others ought to consent to suffer the violations of their rights caused by unjust speech. So there is no case for treating speech of Types D and C as almost self-regarding. But Type B speech only causes permissible non-consensual harm, and the permissibility signals that there is a strong case for ignoring the fact that consent conflicts with selfish interest in this context.
If a moral agent with a disinterested concern to protect the shared vital interests of all by equal right imagines herself with her own preferences in the position of the losing competitive speaker or the Nazi, for example, she would consent to suffer this sort of harm. In a fair competition, she might be pained a bit because of her wasted exertions on behalf of her own opinions and ideas, but such pains would be more than offset by her satisfaction associated with the social benefits produced by freedom of speech of Type B. Her focus is not on some selfish desire to win at all costs but rather on those social benefits such as the acquisition of warranted opinions, exposure of political corruption, and widespread cultivation of individuality. Similarly, she is not preoccupied with amour propre but instead seeks a progressive society composed of individuals who have sufficiently developed their self-confidence to laugh at themselves and at human foibles. And a moral agent in the Nazi’s position would consent to be deceived so as to prevent the violation of the innocent Jew’s right to life. In short, the permissible harms caused by Type B expression to losing competitors, humorless egoists, or wrongdoers can be seen as consensual from a moral point of view. Speech that causes only this kind of harm may be classified as almost self-regarding, even though selfish agents do not in fact consent to suffer it.
Notice that Type B speech can be received by competent third-party bystanders in a truly self-regarding way. In other words, although it produces non-consensual harms to other speakers, humorless egoists, or wrongdoers, its receipt by third parties, and use by them in the formation of their own thoughts and opinions, doesn’t itself impose any further non-consensual harm on others. The content of the speech act, or of the expressive component of the symbolic action, is permissible, and so is its time, place, and manner of delivery, and the bystanders’ “consumption” of the speech doesn’t change any of that. The effect on bystanders plays a major role in Mill’s defense of freedom of discussion and debate: “it is not on the impassioned partisan, it is on the calmer and disinterested bystander, that this collision of opinions works its salutary effect” (Mill, 1977, p. 257). Indeed, to prohibit the bystander from consuming this kind of speech and using it as he pleases to facilitate his own thinking is wrongful interference with the person’s right of self-regarding liberty. Such wrongful interference, whether by public officials or private individuals, ought itself to be prevented and punished by law or public opinion.
By contrast, preventing bystanders from having an opportunity to receive unjust speech of Types D or C is no injustice to them. The idea that such speech must be permitted so that onlookers can receive it merely to facilitate their own thinking has no merit. Protecting instead of suppressing malicious libel, credible threats, or incitements to violence clearly ignores the injustice caused to the relevant parties and can only facilitate the violation of the parties’ rights.
Type A expression is what may be called private expression, which is truly self-regarding conduct. It includes, for example, talking to oneself, writing in a secret diary, intimate conversations between lovers, and small talk among family and friends, where small talk comprises such things as discussing the weather, parties, and sports as well as harmless but tasteless jokes and gossip. Also included are the harmless voluntary actions taken by individuals who create objects of fine art, as opposed to public criticism of these activities and objects.6 Indeed, private expression covers expressive aspects of any self-regarding actions, whether or not the relevant messages are intended by the agent. But these various elements of Type A speech are protected by the right of self-regarding liberty, even apart from any question of a right to free speech.
The right of self-regarding liberty also authorizes voluntary associations, including private clubs, churches, and corporations, to restrict expression by members in ways that do not cause non-consensual injuries to non-members. Contractual conditions of membership may include bans against revealing corporate or club business secrets, for example, which are properly enforced by the state in addition to its own prohibitions against speech of Types D and C. The enforceable bans may extend for some time after individuals have resigned from the voluntary group. Voluntary associations may also prohibit speech of Types B and A by members inside the walls of the association, as when universities forbid insulting speech acts that disrupt the classroom, or when churches and monasteries require silence during prayers or meals. But these bans are not expediently enforced by the state, and expulsion of members who refuse to abide by the relevant rules is left to the association. Moreover, the voluntary associations have no legitimate authority to prohibit these types of speech outside the association, whether by members or nonmembers.
Two Distinct Rights
There is no need for a distinct right of free speech with respect to Type A speech, because such private expression is self-regarding conduct and thus already covered and given absolute protection by the moral right of self-regarding liberty. Schauer, despite his emphasis on the other-regarding nature of some speech, claims that “most speech is not harmful, but most non-speech actions are not harmful as well” (Schauer, 2015a, p. 135). While his claim is a welcome change from the dogmatic assertion, once prominent, that self-regarding conduct is nonexistent, the question of the relative quantities of self-regarding versus other-regarding expression may be left open, even if (contrary to the assumption underlying Figure 1) we accept a view that some voluntary conduct is not expressive. The more important point is that a great amount of speech is public in the sense of causing non-consensual perceptible damage to others. Such other-regarding expression, including expression of Types D, C, and B, is not covered or protected by the right of self-regarding liberty.
It follows that there is no error involved in conceiving of a moral right to freedom of public expression, which is distinct from the right of self-regarding liberty. This distinctive right of free speech, in other words, a Free Speech Principle, covers speech of Types D, C, and B but not Type A speech, and gives protection (and absolute protection) only to Type B speech as that type alone is “almost” self-regarding. By giving full protection to Type B speech, the Free Speech Principle, like the principle of self-regarding liberty, promotes substantial social benefits as already indicated. In particular, these two principles both promote the cultivation of individuality, or self-development of one’s own intellectual, moral, and imaginative capacities, in the direction of a Greek ideal of noble moral character (Mill, 1977, p. 266; Riley, 2013, 2015a, pp. 96–107, 113–116, 261–268). Unlike the self-regarding liberty principle, however, the distinct Free Speech Principle that protects Type B expression does involve the imposition of non-consensual perceptible damage on other people, and yet this harm is not of an unjust kind. Noble agents would consent to suffer such harm, and so the right to freedom of speech of Type B is “almost” but not truly a right of self-regarding liberty, “resting in great part on the same reasons.”
The distinctive right to freedom of speech can be properly stated in either of two ways. If public expression is understood very broadly, even to include all voluntary other-regarding conduct as expressive, then the right covers all expression broadly understood, even all voluntary other-regarding conduct, but gives protection only to Type B expression. If, however, public speech is understood as restricted to Type B expression, or understood even more narrowly (as Mill may have understood it) as Type B speech acts (excluding symbolic actions), then the right covers only Type B speech and gives absolute protection to it. In any case, Type B speech includes public discussion of opinions on all scientific, moral, theological and political subjects.
By analogy with the right of self-regarding liberty, under which the absolute liberty of Type A private speech is subsumed, the distinctive right of public speech is a moral or justified legal claim on society to guarantee the speaker’s liberty in the positive sense of choosing as she pleases among Type B other-regarding speech acts and actions whose expressive components are inextricably bound up with their non-expressive components. Others, including government officials, have correlative duties not to coercively interfere with the speaker.7 Indeed, government is obligated to expediently enforce the correlative duties of others, including government officials themselves (through a system of checks and balances) as well as citizens and voluntary groups such as corporations and churches, not to coercively interfere with the speaker’s Type B speech or to force her to engage in speech of Types D or C. Strictly speaking, the right of public speech may be described as a complex right that includes a liberty or privilege to speak Type B speech backed up by a claim not to be coercively interfered with by others, as well as an immunity from others waiving their correlative duties—others have no legitimate power to waive them. Also included is a power to combine with others to engage in Type B speech, with the caveat that no individual has the power to permanently alienate her right to speak.8
Distinctiveness of the Neo-Millian Framework
Schauer, among many others, views the right of free speech somewhat differently. He starts with a general liberty principle, that is, a presumption that the individual has a natural liberty or privilege in the sense of no duty not to perform any action. But the government can legitimately interfere with the individual’s liberty using a very weak test, namely, a rational basis criterion according to which government need only show to the courts that coercive interference serves a reasonable purpose, that is, a purpose that is not blatantly irrational or insane. So, unless it can be shown that a proper subset of conduct is special and deserves heightened protection from government interference, the individual cannot be said to have a right or claim to liberty with respect to that special subset. Schauer is skeptical that, even if speech is somehow distinguished in ordinary language from non-speech conduct, there is any feature of speech as a class that justifies giving it heightened protection as compared to non-speech action as a class given that, for example, it seems false that speech causes fewer harms or a less severe kind of harms than non-speech conduct does. So he questions whether speech as a class should be treated as special and given heightened protection by right; in other words, he doubts that there is a convincing justification for an independent Free Speech Principle, even though the First Amendment requires its interpreters to assume otherwise in the American context.
Despite his repeated references to self-regarding conduct, Schauer doesn’t focus on Mill’s idea that an equal right to liberty of self-regarding conduct is justified because it, unlike a right to liberty of other-regarding conduct, produces social benefits such as the widespread cultivation of individuality or self-improvement without causing any non-consensual harm to other people (Riley, 2015a, 2017a, 2017b). Nor is a right to self-regarding liberty recognized by American courts so that it makes no appearance in judicial interpretation of the First Amendment. And so Schauer and other scholars of the First Amendment, despite their appreciation of Mill as a liberal icon (Schauer, 1992, 2011), do not consider a neo-Millian analysis of the sort sketched earlier, in which a distinctive right to freedom of Type B expression is based on the pre-constitutional and pre-legal grounds that Type B expression is “almost” self-regarding (though really other-regarding) conduct.
For Mill, the justification for the equal right to freedom of Type B expression is very similar to the justification for the equal right of self-regarding liberty itself. The distinctive right to liberty of Type B speech tends to produce social benefits such as the widespread cultivation of sufficient independence of mind to dissent from received opinions; and the ensuing variety of opinions expressed in public discussions, critiques, and demonstrations in turn tends to produce further social benefits such as discovery of warranted opinions, improved social customs, and so on. True, these various social benefits, which are more or less the same as those produced by the right of self-regarding liberty, are purchased at the cost of non-consensual perceptible damage to others. But the non-consensual harms are of a kind that is morally permissible so that moral agents, unlike selfish ones, would consent to them.
Even if he became convinced that a compelling justification exists for a distinct Free Speech Principle prior to the commands of the First Amendment or of statutes, Schauer apparently has a non-Millian conception of a right to free speech. In particular, he emphasizes that a distinctive right of free speech is a liberty or privilege backed by a fundamental constitutional immunity from government coercion but not from coercive interference by nonstate actors such as private citizens, companies, churches, and the like: “The separation between the individual and government is central to the Free Speech Principle, and this feature is lost when, following Mill, we conflate social intolerance and governmental intolerance” (Schauer, 1982, p. 122). Schauer admits that a legal claim-right against nonstate actors “might be valuable” but stresses that “legal rights in the private law sense are very different from moral or political rights against state interference” (Schauer, 1982, p. 125). The idea seems to be that rights against state action are fundamental constitutional rights such as those embodied in the First Amendment whereas private law rights are not so fundamental. Thus: “A free speech claim against a private individual is different in kind from a free speech right against the state” (Schauer, 1982, p. 125). Schauer seems to think that it is far more difficult to identify wrongful coercive interference by a private agent than by government, and that the kind of argument needed to justify a free speech right against nonstate actors is “substantially different” from the kind of argument needed to justify a Free Speech Principle against state action.9
Shiffrin’s view seems to match Schauer’s in this regard. After she states that her “thinker-based theory [of freedom of speech] supports the moral right against private censorship as well as public [state] censorship,” and she allows that “a constitutional system that enshrined protections against both forms of censorship might better protect the underlying moral rights than our current [American] system,” she goes on to suggest that there may be reasons to follow the American practice of protecting the moral right against government censorship at the constitutional level but protecting the right against private censorship “more contingently, at the legislative level” (Shiffrin, 2014, pp. 108–109). She points to “the immensity of the state and the power it may quickly muster,” for example, and, like Schauer, seems to believe that there are peculiar difficulties surrounding identification of coercive interference by nonstate actors. And yet she admits that there are “easy cases of parallelism” between state coercion and private coercion, “such as when strangers threaten to harm advocates if they continue to voice their political views” or when powerful corporations issue “sanctions against employees for disfavored, off-duty speech irrelevant to bona fide job requirements” (Shiffrin, 2014, p. 109). Indeed, although not mentioned by either Shiffrin or Schauer, society can, as Mill suggests, legitimately consider regulating promotional public advertising by companies seeking to maximize profits from individuals’ self-regarding activities, including gambling, prostitution, and consumption of some types of porn, of which majorities disapprove, even though competent adults must be free to engage in the self-regarding activities. Deliberate exaggeration of the net social benefits of these activities biases public information in favor of the profit-seekers and is likely to mislead some individuals to their own detriment. By restricting such promotional advertising, society can provide an opportunity for individuals to make up their own minds free from distortions injected by self-interested private advertisers.
A constitutional system that gives fundamental protection against both state censorship and private censorship does seem better than the American system for enforcing the moral right to freedom of Type B expression. The First Amendment is flawed insofar as it fails to require government to prevent or punish wrongful coercive interference by nonstate actors with Type B speech. Wrongful coercive interference is no more difficult to identify in the private context than in the public. This is not to deny that private companies and other voluntary associations have a right to restrict Type B and even Type A expression within the walls of the company, club, church, and so on, as long as members are free to resign their membership as they wish.
Notice that a Millian system, which guards against social or private intolerance no less than against state intolerance, doesn’t treat public expression as special. Type B speech, though given absolute protection by right, receives the same level of protection from coercive interference as self-regarding conduct. And unjust speech of Types D and C, while unprotected, is assessed under the same standards of injustice, or of wrongful non-consensual harm, as other-regarding conduct in general. The latter point follows necessarily once we assume that all voluntary conduct is expressive.
Unjust Public Expression Obstructs Self-Regarding Liberty
Remarkably, the Millian approach brings to prominence the fact that unjust expression of Types D and C always violates some other person’s right of self-regarding liberty, whether or not it violates her other moral rights. For example, malicious libel coercively interferes with her self-regarding freedom to pursue her own life in her own way short of non-consensual harm to others, and this interference is a wrongful non-consensual harm to her in addition to any wrongful damage caused to her reputation, finances, and so on. The malicious lies wrongfully weaken her control over her self-regarding sphere, which is, by right, absolute. Similarly, credible threats and incitements violate her right to pursue her own life as she wishes, in addition to endangering if not violating her property rights and even her right to life. In some cases, unjust speech violates only her right of self-regarding liberty without also violating other rights. Gross invasion of privacy can be an example of this (Riley, 2005).
An important application of the Millian approach, which illustrates how it differs in practice from the current American free speech ideology, relates to hate speech, roughly understood to comprise speech acts and expressive actions that cast a false and malignant light on others by identifying them as evil or subhuman on the basis of their racial, sexual, religious, ethnic, national, or other group characteristics.10 If hate speech rises to the level of credible threats of violence against determinate individuals, then its classification as unjust Type D speech is unexceptional. While reasonable criteria are needed to determine when a credible threat exists, there is no doubt that a credible threat of wrongful harm to others is justifiably prevented or punished. Such a threat need not involve speech acts using words. Expressive conduct by those who associate themselves with odious groups and organizations that have a history of discrimination and violence against innocent minorities can constitute a credible threat against identifiable members of those minorities (Riley, 2008). The state legitimately denies a license to a potentially dangerous neo-Nazi group seeking to march in Nazi regalia through a Jewish community, for example, and it properly inflicts extra punishment on trespassers for associating themselves with the Ku Klux Klan by burning a cross on a black family’s lawn.11 Similarly, it legitimately denies entry to foreigners convicted of hate crimes elsewhere who seek to visit or work in the home country.
But what about non-threatening hate speech? Insofar as this means racist jokes or ethnic insults or religious sneers and the like among family and friends, it is protected private Type A speech. Anyone who dislikes or feels contempt for such self-regarding conversations can freely warn his associates not to engage in it and, if his warnings are ignored, choose to avoid those speakers and subject them to other natural penalties.
Similarly, if hate speech is taken to include published cartoons, lampoons, and the like that poke fun at prominent opinions or persons (living or dead, human or divine), then such public hate speech is protected as non-threatening Type B expression. Even so, toleration doesn’t imply approval or endorsement of such expression.
But what of public hate speech that defames others by tying them to a fake group identity, although it doesn’t directly pose a credible threat of egregious perceptible damage to any assignable individual or even accuse him or her of wrongdoing of his or her own? What of hate speech of this sort communicated through repeated radio and television broadcasts, permanent websites, and public billboards erected along highways? Such group libel doesn’t directly threaten any particular individual with violence, for example, or with financial loss, or damage to her own reputation, or loss of her own citizenship. Under the First Amendment, hate speech of this sort generally receives heightened protection as robust political speech, which, though rude and offensive, is essential to a lively working democracy. American courts give such hate speech something like absolute protection so that it is, in effect, treated as Type B speech.
Jeremy Waldron in effect challenges this American view, which is extraordinary among advanced civil societies, by arguing that such group libel wrongfully harms any member of the target group by depriving him or her of “dignity in the sense of a person’s basic entitlement to be regarded as a member of society in good standing, as someone whose membership of a minority group does not disqualify him or her from ordinary social interaction” (Waldron, 2012, p. 105). Despite its considerable appeal, however, Waldron’s argument fails to make clear why this form of hate speech must be judged incompatible with dignity thus understood. What rules out the counter-argument that robust political speech including group libels, though offensive, is speech that a person with dignity must learn to tolerate? Proponents of the American view typically insist that a dignified person can and should ignore the insults, slurs, and false assertions directed at his group because he is by assumption not actually facing any credible threat of wrongful harm such as physical injury or loss of his citizenship. The targeted individual can presumably reply in kind to his assailants so that a robust democratic discussion remains possible: more speech is the remedy because, in the aggregate, the offensive remarks about groups supposedly counteract and neutralize one another. Timothy Garton Ash opposes legal regulation by arguing along these lines, for example. In his view, regulation has proven ineffective and, if introduced, only impairs a kind of “robust civility” that admits diverse ideas of what counts as offensive speech (Garton Ash, 2016, pp. 208–241, 253–282).
Nevertheless, while ambiguous in some respects, Waldron’s argument hints at a justification that can be gleaned from Mill’s work for concluding that group libel is not merely offensive but also unjust speech that ought to be suppressed. A main theme of On Liberty is that public expression can cause wrongful harm to others by coercively interfering with their self-regarding liberty (including their freedom of Type A speech) and, by extension, with their freedom of Type B speech. By stigmatizing and humiliating individuals simply because they are members of a supposedly evil or subhuman group (usually a minority but conceivably a majority), group libels can create a coercive social atmosphere in which the moral rights of the individuals to choose among their self-regarding actions and to engage in Type B speech as they please are denied. In such an atmosphere, the individual may be subjected to verbal insults, he may be refused service in various establishments, certain jobs and careers may be closed to him, his business may be boycotted, and so on, if he voices certain political or religious opinions, wears certain articles of clothing, or simply looks like he’s a member of the targeted group. Thus, innocent people are intimidated and forced to cease and desist from their preferred self-regarding conduct and Type B expression in order to function and to avoid being harmed by ignorant and belligerent others in society.
Group libel, especially as conveyed through the mass media, public billboards, public demonstrations, and the like, is a coercive form of hate speech that is properly seen as unjust because it directly creates and maintains, or poses a significant risk of doing so, an atmosphere of unfreedom in which individuals are forced to give up their liberties of self-regarding conduct and Type B public expression. This can occur even though the group libel itself hasn’t threatened to harm any assignable member of the targeted group. It may be objected that group libels don’t necessarily produce an atmosphere of unfreedom, and that they can provoke more tolerant majorities to rise up and reaffirm support for the equal rights of the unjustly maligned group. But the risk of the unfree atmosphere remains, and the suggestion that large numbers of people will rise up in defense of targeted groups may underestimate the “free rider” problems associated with such collective actions.
Waldron’s argument goes through on Millian grounds.12 Instead of being able to pursue her own life and engage in protected speech as she wishes, the targeted individual is forced to endure the harmful effects of a fake but entrenched public identity reflecting the malicious lie that all members of a given minority group are evil wrongdoers, subhumans, or undeserving of the status of equal citizens. As Waldron puts it: “group-libel laws aim at protecting the basics of each person’s reputation [I would prefer to say liberty] against attempts (for example) to target all the members of a vulnerable racial or religious group with some imputation of terrible criminality” (Waldron, 2012, p. 47).
True, there is some discretion involved in how society should respond to such wrongful group libel. Perhaps official warnings and civil fines are sufficient to discourage it when its appearance is confined to isolated instances. At some threshold point, however, society must terminate repeated radio and television broadcasts of these malicious lies about groups, remove websites and blogs that spread such lies on the Internet, forbid public billboards, and the like, so as to prevent the creation of the atmosphere of unfreedom.
Coercive Public Speech, When Permissible, When Not
Natural penalties, which are inseparable from dislike and are “the only . . . inconveniences” to which an agent of self-regarding conduct and, by extension, of Type B expression “should ever be subjected” (Mill, 1977, pp. 278–279), include the loss of friends and acquaintances, given that attempts to persuade the agent to cease his disliked conduct have failed. Such natural penalties, though harmful to the agent, are the result of others’ spontaneous self-regarding conduct, because the agent in effect consents to suffer the harm by persisting in his disliked conduct despite attempts to persuade him to desist. But Mill tells us that, although others rightfully have liberty to avoid the agent and to warn third parties to avoid him, they do not have any right “to parade the avoidance” (Mill, 1977, p. 278). By implication, natural penalties do not include parading disapproval of the agent’s self-regarding conduct and, by extension, of his Type B speech, where parading is understood as ostentatious public displays designed to stigmatize and humiliate the agent so as to force him to cease the conduct. Parading is coercive public expression, exemplified by organized campaigns of public humiliation, boycotts of businesses, public marches and demonstrations, and the like. It is a form of deliberate social punishment that is legitimate only in response to immorality, that is, unjust expression of Types D and C that violates others’ moral rights.
Coercive public speech directed to prevent or punish injustice is permissible Type B expression, as when lies are told to deceive the Nazi seeking to murder Jews. Boycotts of businesses that discriminate against minorities are permissible, for example, and so are demonstrations with the announced goal of humiliating those who engage in unjust group libels. It is entirely appropriate to create an atmosphere of unfreedom for expression of Types D and C. But coercive public expression becomes unjust itself when directed against self-regarding conduct (including Type A speech) or Type B speech. Coercive hate speech that violates the rights of minorities by stigmatizing all members of the minority group as wrongdoers is unjust and ought itself to be prevented or punished. A liberal democratic constitution ought to demand that government use coercion to silence such hate speech. This censorship is consistent with what Mill says about freedom of expression.
Conclusion: New Challenges in the Digital Age
Evidently, the rise of the global Internet and extensive use of smartphones creates new problems for freedom of expression. Not only do the digital technologies facilitate group libels and incitements to violence against minorities within a pluralistic nation-state, they also facilitate as never before the rapid international transmission of such unjust hate speech. Unjust speakers in one society can attempt to create an atmosphere of unfreedom in another society by stigmatizing the Type B religious expression of the majority in the other society, for example, sanctioning them as evil Muslims or as wicked atheists. The speakers can even incite violence in that other society, committed by members of religious minorities or of the majority against innocent members of the other group. Violent reactions to coercive hate speech emanating from a foreign source should, of course, be put down. And yet it must not be forgotten that the hateful foreign speakers are also unjust. They cannot be allowed to hide their Type D or C expression behind an ideology of free speech absolutism.
Garton Ash thoughtfully discusses and amply illustrates problems of this sort, including the posting on the Internet of a hateful anti-Muslim video that seems eventually to have incited a mob to attack the U.S. consulate in Benghazi, killing the ambassador to Libya and others (Garton Ash, 2016, pp. 62–72). One of his main themes is that individuals and groups must learn to tolerate speech they dislike. This is a valuable lesson with respect to speech of Types A and B. But speakers must also learn that speech of Types D and C is unjust. Free speech absolutism is a false doctrine when extended to cover such unjust speech. More speech is not always desirable. Sometimes it exacerbates injustice.
There is no denying the difficulty of reaching an international consensus on anything like a Millian analysis of freedom of speech. Universal failure to recognize either the moral right of self-regarding liberty or the different types of expression prevails. As matters stand, for example, some societies but not others accept that people have moral rights to freely express whatever religious opinions or to publish whatever silly cartoons they please. Some societies but not others recognize that a hateful group libel is not merely offensive but an unjust form of public expression because it violates others’ moral or justified legal rights, or poses a credible risk of doing so. Different societies are likely to establish sub-optimal and conflicting hate speech laws. Moreover, there is little prospect of finding international mechanisms that oblige governments not only to refrain from unjust speech themselves but also to prevent nonstate actors from engaging in it. We can dream with Garton Ash of some sort of international consortium of governments, private companies, and other voluntary associations with authority to remove hateful websites and blogs from the Internet. The consortium would also need authority to terminate repeated broadcasts, related billboards, and other more or less permanent public signs of hate speech within any society. Unfortunately, this is likely a pipe dream for a long time to come.
I wish to thank the editors and two anonymous external reviewers for their comments on an earlier draft of the article. I am also grateful to Molly Rothenberg for her comments and for help with drawing Figure 1. All responsibility for the views expressed remains mine.
Baker, C. E. (1997). Harm, liberty, and free speech. Southern California Law Review, 70(4), 979–1020.Find this resource:
Baker, C. E. (2002). Media, markets and democracy. Cambridge, U.K.: Cambridge University Press.Find this resource:
Baker, C. E. (2006). Media concentration and democracy: Why ownership matters. Cambridge, U.K.: Cambridge University Press.Find this resource:
Blasi, V. (2011). Shouting “Fire!” in a theater and vilifying corn dealers. Capital University Law Review, 39(3), 535–569.Find this resource:
Citron, D. K. (2014). Hate crimes in cyberspace. Cambridge, MA: Harvard University Press.Find this resource:
Cronin, M. M. (Ed.). (2016). An indispensable liberty: The fight for free speech in nineteenth century America. Carbondale: Southern Illinois University Press.Find this resource:
Dworkin, R. (2009). Foreword. In I. Hare & J. Weinstein (Eds.), Extreme speech and democracy (pp. v–ix). Oxford: Oxford University Press.Find this resource:
Hare, I., & Weinstein, J. (Eds.). (2009). Extreme speech and democracy. Oxford: Oxford University Press.Find this resource:
Levy, L. W. (1963). Freedom of speech and press in early American history: Legacy of suppression (2d ed.). New York: Harper.Find this resource:
Rabban, D. M. (1999). Free speech in its forgotten years: 1870–1920. New York: Cambridge University Press.Find this resource:
Redish, M. H. (1984). Freedom of expression: A critical analysis. Charlottesville, VA: Michie.Find this resource:
Redish, M. H. (2013). The adversary First Amendment: Free expression and the foundations of American democracy. Stanford, CA: Stanford University Press.Find this resource:
Saunders, K. W. (2017). Free speech and democracy: A comparative analysis. Cambridge, U.K.: Cambridge University Press.Find this resource:
Sumner, L. W. (2004). The hateful and the obscene: Studies in the limits of free expression. Toronto: University of Toronto Press.Find this resource:
Sunstein, C. R. (1993). Democracy and the problem of free speech. New York: Free Press.Find this resource:
Sunstein, C. R. (2009). Republic.com 2.0. Princeton, NJ: Princeton University Press.Find this resource:
Sunstein, C. R. (2017). #Republic: Divided democracy in the age of social media. Princeton, NJ: Princeton University Press.Find this resource:
Baker, C. E. (1989). Human liberty and freedom of speech. Oxford: Oxford University Press.Find this resource:
Dworkin, R. (1996). Freedom’s law: The moral reading of the American Constitution. Oxford: Oxford University Press.Find this resource:
Dworkin, R. (2006). Is democracy possible here? Principles for a new political debate. Princeton, NJ: Princeton University Press.Find this resource:
Garton Ash, T. (2016). Free speech: Ten principles for a connected world. New Haven, CT: Yale University Press. See also Free Speech Debate.Find this resource:
MacKinnon, C. A. (1993). Only words. Cambridge, MA: Harvard University Press.Find this resource:
Magarian, G. P. (2017). Managed speech: The Roberts Court’s First Amendment. New York: Oxford University Press.Find this resource:
Mill, J. S. (1969a). Whewell on moral philosophy. In J. M. Robson (Ed.), The collected works of John Stuart Mill (Vol. 10, pp. 165–201). Toronto: University of Toronto Press.Find this resource:
Mill, J. S. (1969b). Utilitarianism. In J. M. Robson (Ed.), The collected works of John Stuart Mill (Vol. 10, pp. 203–259). Toronto: University of Toronto Press.Find this resource:
Mill, J. S. (1977). On liberty. In J. M. Robson (Ed.), The collected works of John Stuart Mill (Vol. 18, pp. 213–310). Toronto: University of Toronto Press.Find this resource:
Riley, J. (2005). J. S. Mill’s doctrine of freedom of expression. Utilitas, 17(2), 147–179.Find this resource:
Riley, J. (2008). Racism, blasphemy and free speech. In C. L. Ten (Ed.), Mill’s “On Liberty”: A critical guide (pp. 62–82). Cambridge, U.K.: Cambridge University Press.Find this resource:
Riley, J. (2010). Mill’s extraordinary utilitarian moral theory. Politics, Philosophy & Economics, 9(1), 67–116.Find this resource:
Riley, J. (2012). Happiness and the moral sentiment of justice. In L. Kahn (Ed.), Mill on justice (pp. 158–183). London: Palgrave Macmillan.Find this resource:
Riley, J. (2013). Mill’s Greek ideal of individuality. In K. N. Demetriou & A. Loizides (Eds.), John Stuart Mill: A British Socrates (pp. 97–125). London: Palgrave Macmillan.Find this resource:
Riley, J. (2015a). Mill’s “On Liberty.” London: Routledge.Find this resource:
Riley, J. (2015b). Is Mill an illiberal utilitarian? Ethics, 125(3), 781–796.Find this resource:
Riley, J. (2016). Mill on utilitarian sanctions. In C. Macleod & D. E. Miller (Eds.), Companion to John Stuart Mill (pp. 342–357). Oxford: Blackwell.Find this resource:
Riley, J. (2017a). Mill’s radical liberalism: A study in retrieval. London: Routledge.Find this resource:
Riley, J. (2017b). Mill’s absolute ban on paternalism. In K. Grill & J. Hanna (Eds.), Routledge handbook on the philosophy of paternalism (pp. 350–390). London: Routledge.Find this resource:
Scanlon, T. M. (2003). The difficulty of tolerance: Essays in political philosophy. Cambridge, U.K.: Cambridge University Press.Find this resource:
Schauer, F. (1982). Free speech: A philosophical enquiry. Cambridge, U.K.: Cambridge University Press.Find this resource:
Schauer, F. (1984). Must speech be special? Northwestern University Law Review, 78(5), 1284–1306.Find this resource:
Schauer, F. (1992). The First Amendment as ideology. William and Mary Law Review, 33(3), 853–869.Find this resource:
Schauer, F. (2003). The phenomenology of speech and harm. Ethics, 103(4), 635–653.Find this resource:
Schauer, F. (2011). On the relation between chapters one and two of John Stuart Mill’s “On Liberty.” Capital University Law Review, 39(3), 571–592.Find this resource:
Schauer, F. (2015a). Free speech on Tuesdays. Law and Philosophy, 34(1), 119–140.Find this resource:
Schauer, F. (2015b). On the distinction between speech and action. Emory Law Journal, 65(2), 427–454.Find this resource:
Shiffrin, S. V. (2014). Speech matters: On lying, morality, and the law. Princeton, NJ: Princeton University Press.Find this resource:
Steiner, H. (1998). Working rights. In M. H. Kramer, N. E Simmonds, & H. Steiner, A debate over rights (pp. 233–301). Oxford: Oxford University Press.Find this resource:
Sullivan, K., & Feldman, N. (2016). First Amendment law (6th ed.). St. Paul, MN: Foundation.Find this resource:
Waldron, J. (2012). The harm in hate speech. Cambridge, MA: Harvard University Press.Find this resource:
(2.) There is no doubt that thinking, or forming one’s thoughts and opinions, is self-regarding conduct because one’s opinions are unknown to others and cannot harm them unless the opinions are expressed or communicated. Mill doesn’t deny that one’s ideas and opinions are causal contributors to one’s voluntary actions, both self-regarding and other-regarding. But unexpressed thoughts per se cannot harm other people, whereas other-regarding conduct including public expression causes non-consensual harm to others or poses a credible danger of doing so. So a right to liberty of self-regarding conduct including forming one’s own thoughts is distinct from, and compatible with either acceptance or rejection of a right to liberty of other-regarding conduct including expression. By prohibiting certain harmful other-regarding actions, society doesn’t interfere with the liberty of thought but rather with the liberty of acting on certain thoughts in ways that cause non-consensual perceptible damage to others.
(3.) Mill is not a standard utilitarian. For my interpretation of his extraordinary utilitarianism and the way it consistently privileges a liberal code of justice and equal rights over competing social considerations, see Riley (2010, 2012, 2016, 2017a). Contrary to received academic opinion, his non-standard utilitarianism illustrates how absolute rights, such as a right of self-regarding liberty and a right of freedom of Type B public discussion, can be consistently grounded in a consequentialist moral theory.
(4.) An anonymous referee asks whether “child pornography involving computer-generated images” counts as Type D expression. No, because such expression doesn’t cause any harm to sentient creatures, although it may cause dislike and disgust.
(5.) There are some exceptions. Those who buy child or animal porn are legitimately subject to punishment, for example, but those who pay blackmail are only harming themselves. The porn cannot be consumed independently of the non-consensual harm caused to the children or animals whereas paying blackmail is self-regarding activity that doesn’t deserve any punishment. See also Riley, 2017b.
(6.) This is not to say that activities related to the fine arts are necessarily self-regarding. Forcing slaves to construct great pyramids or monuments is unjust other-regarding activity, for example, as is forcing minors and animals to participate in porn films.
(7.) This is not a duty to view or listen to the speech but rather a duty not to prevent or punish the speech. The speaker has a right to achieve an outcome in which she engages in her Type B speech. But others may freely choose to avoid any Type B speech they dislike, although they have a duty not to coercively interfere with it.
(9.) Although he usually speaks of a right to free speech as an immunity from government interference, Schauer implicitly refers to a complex right that includes other Hohfeldian positions, at least if he accepts Hillel Steiner’s (1998) “working conception of rights.” Thus, for example, to have an immunity is also to have a claim: an immunity against state officials implies that the state has no power to alter the constitutional (and, by analogy, moral) position of the holder of the immunity, a no-power implies no-privilege or no-permission to exercise the power, a no-privilege is equivalent to a duty not to exercise the power, and the duty of state officials not to exercise the power they lack implies that the holder of the immunity has a claim correlative to that duty.
(10.) Speech that puts individual wrongdoers in a bad but true light, as when a judge sentences a duly convicted criminal and expresses disgust at the individual for committing the crimes, is not “hate speech” as defined. Rather, it is justified Type B speech. Moreover, if there were evidence that “Swedes [or blacks or Muslims] really are more inclined to crime than others,” then expressing that fact would not be hate speech. Hate speech must cast a false and malignant light on a group. I thank an anonymous referee for pressing me to clarify this point.
(11.) An anonymous referee wonders whether the state could legitimately prohibit the neo-Nazis from marching “in a neo-Nazi neighbourhood.” Assuming there is such a community, the marchers do not pose a credible threat of harm to others, although third parties might invade the community wrongly seeking to provoke a riot. Indeed, it’s unlikely that anyone in the community would protest so that non-consensual harm to them wouldn’t be an issue, provided the march is scheduled at a suitable time and place and is more or less peaceful. So, the local government could legitimately grant the neo-Nazis a license to march in this context. Similarly, a trespasser should not receive extra punishment for burning a cross on property owned by the Ku Klux Klan, even if the organization decided to press charges for the trespass.
(12.) To be fair, Waldron does briefly mention Mill and his distinction between natural penalties and social punishment such as organized “boycotts or ostracism” (Waldron, 2012, p. 229). But he doesn’t clarify the distinction or make it an important element of his main argument. Nor does he mention the right of self-regarding liberty or relate it to Mill’s view of the right to freedom of expression.