Abstract
In my recent book Freedom of Expression as Self-Restraint, I expound and defend the moral principle of freedom of expression. This article recounts a few of the main strands of the exposition in that book, and it touches upon the justification for the principle of freedom of expression. Supplementing the abstract ideas broached in the article are several illustrative examples that render the abstractions more accessible.
The moral principle of freedom of expression defended in my recent book Freedom of Expression as Self-Restraint (Kramer 2021) is as follows. Every system of governance is morally obligated to refrain from penalizing any communicative activities qua communicative activities. In other words, any restrictions imposed by a system of governance on some communicative activity are morally legitimate only if they have been imposed not because the activity is communicative but instead because it instantiates a type of misconduct whose wrongness is communication-independent. Central to such an account of freedom of expression, clearly, is the notion of communication-independence. Hence, this paper will begin its presentation of a few aspects of that account by explaining what communication-independence is.
Communication-independence
My references to communication-independent misconduct or to communication-independent wrongfulness might mislead some readers because those references might be taken to suggest that the classifiability of utterances as such misconduct (or as partaking of such wrongfulness) is not due to what the utterances have communicated. Let us have in mind here a couple of the sundry examples of communication-independent misconduct which I ponder in my recent book: the malicious shouting of “Fire” in a dark and crowded theater, and the incendiary urgings of an orator addressed to a mob outside the home of a corn-dealer. My discussion in this section will proceed with reference to those two examples—in order to make the discussion more concretely accessible—but its points are generalizable.
As is evident, the shout in the theater forcefully expresses a message to its addressees. It is designed to induce some immediate life-endangering actions on the part of those addressees, but that effect is deliberately achieved by way of the alterations in their beliefs and attitudes that are brought about through the transmission of the admonitory message. Consequently, the wrongfulness of the shout supervenes on what has been communicated in the circumstances wherein the shout has occurred. Any satisfactory account of that wrongfulness has to make reference to the content and context of the utterance. Those features of the utterance are what warrant its being classified as the inducement of a dangerous public disturbance (or as an attempt to induce a dangerous public disturbance). In other words, the classifiability of the malicious shout as an instance of communication-independent misconduct is due to the message which the shout has conveyed and to the circumstances in which the conveyance of that message has occurred.
Accordingly, if the adjective “communication-independent” were supposed to indicate that the wrongfulness of the act of shouting “Fire” is not due to the communicative import of such an utterance, then that adjective would be inapposite. However, when I characterize the wrongfulness of the utterance as communication-independent, or when I contend that the utterance can properly be subjected to sanctions as a communication-independent mode of misconduct, there is no suggestion that the wrongfulness is unattributable to the content and context of the message which the shout of “Fire” has expressed. Rather, the point is that the shout belongs to a category of misconduct—the category of inducement (or attempted inducement) of a perilous public disturbance—that is not inherently communicative. Any number of potential or actual instances of such misconduct are communicative, and any number of potential or actual instances of such misconduct are non-communicative. These heterogeneous instances are categorized together on the basis of the wrong-making properties which they have in common. Those properties are instantiated by any endeavor to initiate or exacerbate a situation of dangerous public disorder, whether the endeavor is communicative or non-communicative in its character. Precisely because the actions that instantiate those wrong-making properties can be either communicative or non-communicative, the instantiation of those properties is communication-independent. And because the wrongness of any such action does not hinge on whether the action is one of the communicative instances or one of the non-communicative instances in the category of misconduct to which it belongs, its wrongness is likewise communication-independent.
Let us now turn to the harangue by a firebrand orator who incites the members of a mob outside the home of a corn-dealer and goads them into undertaking a lethal rampage of violence against the corn-dealer. Patently, the fulminations of the orator are communicative; through them he intentionally conveys both a clear-cut message and some malign sentiments to the members of the mob. Because of those communicative contents and because of the circumstances in which those contents are imparted, the speaker’s tirade is constitutive of his direct participation in a lynching. Given the proximity of his utterances to the exertions of violence against the corn-dealer, and given the role of his utterances in deliberately spurring the mob to proceed with those exertions, his declamations are properly classified as some of the initial stages of the violence. Their classifiability as such stages—their classifiability as his direct participation in a lynching—is due to their communicative contents and context. Hence, when I claim that his diatribe constitutes a communication-independent mode of misconduct, there is no suggestion that the wrongness of the diatribe is somehow unattributable to what it has communicated. Its wrongness is communication-independent, but not in that ridiculous sense.
Instead, the communication-independence of the misconduct (or the communication-independence of the wrongness of the misconduct) consists in the fact that the direct participation of a person in a lynching can occur either through communicative actions or through non-communicative actions. An inflammatory tirade that stirs up a mob into a frenzy is one way in which somebody can directly contribute to the perpetration of a lynching, and the placing of a rope around the victim’s neck is another such way. Those two modes of direct participation, along with countless other possible modes of such participation, are grouped together in a category of evildoing on the basis of the wrong-making properties which all of them share. Because the fiery tirade belongs to that category of evildoing, it is aptly classifiable as communication-independent misconduct. It belongs to a genus that cuts across the distinction between the communicative and the non-communicative, and its membership in that genus of wrongdoing is what accounts for its legitimate prohibitability. Proper sanctions will be directed against the orator’s incitement not qua act of communication but instead qua act of direct participation in a murderous crime. Anyone who has participated directly in that crime in a non-communicative manner will be similarly liable to undergo such sanctions, for the range of the applicability of proper sanctions—like the category of wrongdoing to which they are punitive responses—cuts across the distinction between the communicative and the non-communicative.
In the respect that has just been summarized, rather than in the ludicrous respect that has been broached in the penultimate paragraph above, the misconduct constituted by the orator’s harangue is communication-independent. That type of misconduct (participation in a deadly act of violence) comprises communicative instances and non-communicative instances alike, and any legitimate prohibition of such misconduct is aimed against what those instances have in common rather than against any features that appertain solely to the communicative instances. Like the type of wrongdoing itself, then, a legitimate prohibition of it is communication-independent. Extending to the type’s communicative instances and non-communicative instances across the board, such a prohibition complies with all the constraints of neutrality that are integral to the principle of freedom of expression.
The neutrality required under the principle of freedom of expression
As has just been suggested, what the principle of freedom of expression requires in any society is that the relationship between the society’s system of governance and the communicative activities of its citizens be marked by several types of neutrality. Only insofar as each of those types of neutrality is attained, will a system of governance have exerted the self-restraint that is involved in never directing any penalties or disadvantages against communicative endeavors qua communicative endeavors.
Communication-neutrality
Most general among the ways in which every system of governance should be neutral with regard to the communicative conduct of its citizens is that it should never treat the communicative character of the conduct either as a ground for the imposition of sanctions (or other disadvantages) or as a ground for the conferral of benefits. Under the principle of freedom of expression, the terms and purposes of all governmentally imposed restrictions must conform to this requirement of general communication-neutrality. An insistence on such neutrality has underlain some of the most famous doctrines and distinctions in American constitutional law.
Let us briefly ponder afresh Oliver Wendell Holmes’s famous example of a man who maliciously shouts “Fire” in a crowded theater. Although such an action is communicative, the prohibition of it can be morally legitimate because the action instantiates a type of misconduct whose wrongness is not distinctively communicative. That is, the shouting is an attempt to induce serious public disorder in circumstances where the disorder is very likely to result in deaths or injuries and damage to property. Such an attempt could have been undertaken through non-communicative conduct such as the firing of a gun or (for that matter) the igniting of a fire. Hence, a law that proscribes such misconduct can be—and almost always will be—communication-neutral in its purpose. Such a law, banning the inducement of serious public disturbances, can provide for sanctions without treating the communicative character of this or that instance of the banned misconduct as a ground for the imposition of those sanctions.
Let us also glance here at a distinction on which I have expatiated in my book about freedom of expression—the distinction between advocacy and incitement, as it has been drawn by the U.S. Supreme Court in the celebrated 1969 case of Brandenburg v Ohio. When somebody’s utterances incite other people to engage in serious criminality, her statements are aimed at bringing about some major wrongdoing in circumstances where the imminent occurrence of the wrongdoing is likely as a result of the statements. That is, a communication that amounts to incitement is characterized by three main elements: the intendedness of the link between the communication and some serious misconduct, the imminence of the misconduct, and the likelihood of the occurrence of the misconduct. By contrast, when somebody advocates the perpetration of misconduct but does not commit incitement, at least one of the three elements just specified is missing. A classic example of incitement, which I have already contemplated in §1 above, is John Stuart Mill’s scenario of a fiery speaker who rails against the iniquity of corn-dealers while addressing a mob of angry people who have gathered outside the home of a local corn-dealer (Mill 1956, 67–8). In Mill’s scenario, and in other cases of incitement, the utterance of inflammatory statements by a speaker is constitutive of the speaker’s participation in the serious and imminent misconduct which the statements are designed and likely to bring about. Because of the proximity between the inciting utterances and the imminent criminality which they are intended to produce, the utterances are subsumed into the criminality as some of the initial stages thereof. Consequently, the imposition of sanctions on Mill’s fiery speaker is unequivocally consistent with the principle of freedom of expression. Punishment is administered not because of the communicative character of the speaker’s fulminations, but because of their having constituted his direct and deliberate involvement in the perpetration of a lynching. Accordingly, that punishment is communication-neutral.
What the Brandenburg criterion for incitement enables us to do, in any context where somebody (like Mill’s rabble-rouser) has exhorted people to commit violence or other serious wrongdoing, is to determine whether the exhortations can be penalized in a communication-neutral manner or not. If the utterances in question do meet the Brandenburg test for incitement, then a relevant communication-neutral legal prohibition—a prohibition on rioting or vandalism or murder or arson, for example—can be applied to those utterances in a communication-neutral fashion. Contrariwise, if the speaker’s declamations do not meet the Brandenburg test and are therefore properly classifiable as mere advocacy rather than as incitement, then any sanctions imposed in response to them are not communication-neutral. Even if the legal prohibition that provides for the sanctions is itself communication-neutral (like the prohibition on rioting or vandalism or murder or arson), any application of it to an act of mere advocacy is not communication-neutral. Unlike incitement, mere advocacy is not subsumed into the misconduct for which it calls; it is not sufficiently proximate to that misconduct to be so subsumed. Hence, if it is subjected to legal sanctions, it is subjected as an act of communication rather than as one of the initial stages of the communication-independent misconduct. Administered in such circumstances, the sanctions are not communication-neutral and are thus in violation of the principle of freedom of expression.
Two types of content-neutrality
Another type of neutrality required under the principle of freedom of expression is that of content-neutrality, which obtains in two main varieties: neutrality of subject matter and neutrality of viewpoint. We should examine each of those two main varieties in turn.
Subject-neutrality
A system of governance that duly abides by the constraint of neutrality of subject matter does not differentiate among the topics that can be broached whenever any general communicative activities are permitted. Helpful here will be a scenario in which this subject-neutrality has been forsaken. Suppose that a local ordinance permits the placing of advertisements (in return for set fees) on the sides and backs of municipal buses, and suppose that the ordinance excludes from this arrangement any advertisements that address politically sensitive topics. That exclusion does not discriminate among viewpoints at all; it applies to any political advertisements irrespective of the points on the political spectrum whence they emanate. Nonetheless, it plainly does discriminate on the basis of subject matter. With regard to a prominent public setting, its purpose is to delimit the range of topics that can be raised by potential communicators and pondered by potential addressees of the communications.
Such a content-specific restriction contravenes the principle of freedom of expression, even though a municipal government could legitimately exclude advertisements altogether from the sides and backs of its buses. What contravenes the principle of freedom of expression is not the extent of the restriction but instead its fine-grainedness. Of course, that feature of the restriction could have been accentuated. For example, instead of excluding all political advertisements, the ordinance might have excluded only advertisements pertaining to abortion or only advertisements pertaining to matters of immigration. Either of those exclusions would have been less wide-ranging than the exclusion of all politically sensitive topics, but—precisely for that reason—either of them would have been more fine-grained as an effort to direct the flow of discourse in a prominent public setting. As I have argued at length in Chapter 4 of Freedom of Expression as Self-Restraint, selectivity can be as problematically inimical to freedom of expression as is overbreadth. 1
Viewpoint-neutrality
Arguably even more important than neutrality of subject matter, as an aspect of the compliance by a system of governance with the principle of freedom of expression, is neutrality of viewpoint. Suppose that, instead of excluding all political advertisements from the sides and backs of municipal buses, a local ordinance were to exclude all advertisements that express opposition to the policies of the United Kingdom’s Labour Party. Or suppose that, instead of excluding all advertisements that pertain to the matter of abortion, a local ordinance were to exclude all advertisements that call for the outlawing of the practice of abortion. Neither of these examples involves a restriction on the range of topics that can lawfully be addressed in a prominent public setting, but each of them involves a restriction on the range of viewpoints that can lawfully be espoused within that setting. Opponents of the Labour Party or of abortion are not legally allowed to proclaim their views in the form of advertisements on public buses, whereas admirers of the Labour Party and proponents of the “pro-choice” side in controversies over abortion are legally allowed to proclaim their views in such a fashion.
Once again, what makes the exclusions objectionable is their selectivity. A municipal government is not morally obligated to make the sides and backs of public buses available as spaces for advertising, but a policy in favor of making them available as such spaces (in return for paid fees) cannot legitimately discriminate with reference to the viewpoints of would-be advertisers. As I have maintained in the aforementioned chapter of my 2021 book, the selectivity of such discrimination is particularly deplorable because it deviates so egregiously from the self-restraint that is morally required of every system of governance under the principle of freedom of expression. On the one hand, such self-restraint would be grossly abandoned through any sweeping restrictions on communications in contexts where no significant restrictions are morally permissible. For example, if a municipal government were to enact a law that prohibits people from conversing in public parks or in railway stations, its heavy-handed attempt to regiment public discourse would be blatantly in violation of the principle of freedom of expression. On the other hand, not only would departures from subject-neutrality or viewpoint-neutrality be morally illegitimate in the contexts just mentioned, but they would also be morally illegitimate in contexts where some blanket restrictions would be morally legitimate. Just above, we have been pondering a context of the latter kind: whereas a local government can permissibly disallow any advertisements on the sides and backs of its municipal buses, no subject-specific or viewpoint-specific limitations would be morally permissible even in such a setting. As endeavors to control minutely the patterns of public discourse and contemplation, subject-specific and viewpoint-specific prohibitions are sometimes more sharply at variance with the ideal of governmental self-restraint than are prohibitions that sweep indiscriminately.
Speaker-neutrality
A further type of governmental impartiality required under the principle of freedom of expression is speaker-neutrality. Although this kind of neutrality is closely related to neutrality of viewpoint, the two are not equivalent. For example, a particular person might be legally prohibited from speaking publicly on a certain issue, even though others who share that person’s position on the issue are legally permitted to speak publicly about it. Indeed, a particular person might be legally prohibited from speaking publicly about some matter, even though other people are legally permitted to appear on radio or television to articulate that very person’s pronouncements on the matter. Just such a situation obtained in the United Kingdom from 1988 to 1994, when broadcasters throughout the country were legally banned from airing the voices of members of Northern Irish terrorist groups—and members of the political party Sinn Féin—on radio or television (Donohue 2008, 293–4). British broadcasters during that period were legally permitted to employ actors to read out the words that had been uttered by the members of those groups in interviews or in other public statements. Consequently, everything said by the terrorists and their political comrades could legally be aired on radio or television in the United Kingdom, provided that the voices of the terrorists or their comrades were replaced with actors’ voices. In such a situation, there was no transgression of content-neutrality by the British system of governance, but there was an obvious transgression of speaker-neutrality.
A point to be emphasized here about the requirement of speaker-neutrality is that it forbids not only the identity-based disadvantaging of a speaker but also any identity-based preferential treatment. This point has sometimes been overlooked even by eminent champions of the values of liberal democracy. For instance, when writing about the flagrantly unjust U.S. Supreme Court case in which the Court upheld the conviction of Eugene Debs for speaking against the role of the United States in the First World War, John Rawls lamented that the Court was “little troubled by the constitutional question raised in Debs, even though the case involves a leader of a political party, already four times its candidate for the presidency” (1993, 351). Rawls was right to be indignant about the Supreme Court’s judgment in the Debs case, but he erred in suggesting that the stature of Debs as a politician was a factor which militated against that judgment. The stature of Debs as a politician was irrelevant; no preferential treatment of his utterances was warranted on that ground. Decisive instead was the fact that he with his orations did not deliberately give rise to any clear danger of serious and imminent misconduct. His speeches did not constitute any participation by him in such misconduct, as they amounted to advocacy rather than to incitement. This communication-neutral ground for acquitting Debs is also straightforwardly speaker-neutral. It applies to his orations the same standard—the Brandenburg Court’s version of the clear-and-present-danger standard—that is applicable to the orations of any other speaker.
Neutrality at the level of standing laws and neutrality in application
If and only if a legal mandate partakes of each type of neutrality that has just been adumbrated, it is in conformity with the requirements laid down by the principle of freedom of expression. Though some types or instances of communicative conduct might be prohibited by a mandate that is in conformity with that principle, they are not prohibited because of their distinctively communicative character or content or because of the identities of the people who engage in them. Instead of being prohibited qua acts of communication, those types or instances of conduct are prohibited because they are constitutive of some communication-independent misdeeds. Communication-neutrality and subject-neutrality and viewpoint-neutrality and speaker-neutrality, then, are the hallmarks of laws enacted by a system of governance that exerts the self-restraint involved in the realization of freedom of expression.
Of course—as I have emphasized in Freedom of Expression as Self-Restraint—the foregoing types of neutrality are crucial not only in connection with the terms and purposes of the laws that have been enacted by a system of governance, but also in connection with the processes whereby those laws are implemented through the decisions and actions of administrative and adjudicative officials. Even if a statute or some other law is itself impeccably in compliance with the principle of freedom of expression, it might be applied selectively by administrative officials or adjudicative officials in contravention of that principle. Consider, for example, the American case of Davis v Norman, which arose from some events in the state of Arkansas in the mid-1970s (Kramer 2021, 26–9, 40–1, 52–3). After Keith Davis had been killed in a collision while he was driving a truck at high speed to escape from a pursuing police car, his father—who blamed the police for the death—placed the wrecked truck on the lawn of his house as a way of protesting against what the father perceived as an abuse of the police’s authority. Under the terms of a local ordinance that prohibited the unenclosed storage of ruined motor vehicles, the father was served with a notice to remove the wrecked truck from his lawn. He explained to the local authorities that he was displaying the truck on his property as a statement of protest against the high-handedness of the police, and he submitted that his display was an instance of symbolic speech for which he could claim the protection of the First Amendment. His appeals against the seizure and removal of the vehicle from his lawn were dismissed by the federal courts that ultimately ruled on the matter.
Although the terms and purposes of the ordinance that forbade the storage of demolished automobiles in the open air were unexceptionably communication-neutral and content-neutral and speaker-neutral, there are quite solid grounds for suspecting that the ordinance was applied selectively to the conduct of Mr Davis in a manner that was neither communication-neutral nor content-neutral. Among the people in the local community who understood the message imparted by the display of the ruined truck were the police. They were undoubtedly displeased by the harsh criticism of them which Mr Davis communicated through his positioning of the truck on his lawn. Hence, in the absence of any further evidence on the question, we have ample grounds for wondering whether the ordinance was being applied selectively against him to squelch that criticism. If the effectuation of the ordinance was indeed selective in that fashion, then the issuance of a removal-directive to Mr Davis transgressed the principle of freedom of expression (and the First Amendment) even though the ordinance itself was consistent with that principle. In any event, whatever may be the accuracy or inaccuracy of these surmises about the actions of the police in that Arkansas municipality, the general point here is that the foregoing types of neutrality are required (under the principle of freedom of expression) in processes of administration and adjudication as well as in the outputs of legislation. In relation to any law or policy that is adopted by a system of governance, the principle of freedom of expression is fulfilled only insofar as both the law or policy itself and the implementation of it are characterized by all of the aforementioned types of neutrality.
The matter of justification
In the fourth chapter of Freedom of Expression as Self-Restraint, I elaborate a full-scale justification of the moral principle of freedom of expression that has been outlined in this paper. Here, I can only touch upon that matter. As is perhaps evident, the several kinds of neutrality required under the principle of freedom of expression do indeed stand in need of justification. After all, a stance of wholesale communication-neutrality and content-neutrality and speaker-neutrality will carry some significant drawbacks in any number of instances. Let us here briefly return to the municipal ordinance which authorizes the selling of places for advertisements on the sides and backs of public buses but which excludes any advertisements that are expressive of certain viewpoints. If every such ordinance is disallowed, and if strict viewpoint-neutrality is insisted upon, the municipal government may sometimes find itself presented with applications for advertisements from odious groups such as the Ku Klux Klan or the Communist Party. Why should the local officials have to choose between not allowing any advertisements to be placed on buses and leaving open the possibility of occasional advertisements placed by such groups?
A satisfactory answer to this question, and a satisfactory answer to any cognate question about any of the other types of neutrality that are mandatory under the principle of freedom of expression, will amount to a thoroughly elaborated justification of that principle. In the chapter mentioned above, I have presented such a justification. In this article, in lieu of such a presentation, I will simply seek to convey a rough sense of the focus that is appropriate for vindicating the principle of freedom of expression. We should here proceed by reflecting afresh on what has already been said about the difference between the extent of a restriction and the selectivity of a restriction.
If the best justification for the principle of freedom of expression were to be centered on the provision of opportunities for addressors (speakers and writers and composers and artists and so forth) to impart their ideas and sentiments, we would be hard pressed to explain why selective restrictions on communications are quite frequently more problematic than wholesale restrictions. After all, the opportunities for expression available to addressors are curtailed by the latter restrictions far more extensively than by the former restrictions. We would face comparable difficulties if the best justification for the principle of freedom of expression were instead centered on ensuring that potential addressees (listeners and readers and viewers) enjoy ample access to the articulated ideas and feelings and hopes of other people. Selective restrictions impinge on such access less severely than do sweeping restrictions. Nor will we be able to account for the moral disparity between selective curbs and blanket curbs if we try to vindicate the principle of freedom of expression by concentrating on a societal desideratum such as the promotion of knowledge. Whereas the pursuit of such a desideratum would typically be badly hindered by any wide-ranging curbs on communicative activities, it could typically be furthered rather than hampered by some curbs that are suitably selective. Thus, like a focus on opportunities for self-expression or a focus on access to other people’s articulations of their ideas and sentiments, a focus on a societal desideratum such as the growth of knowledge cannot explain why selective restrictions on communications are sometimes especially objectionable.
To come up with a germane explanation, we need to train our attention instead on the relationship between a system of governance and the society over which that system presides. When such a system imposes selective constraints on communications, the control which it exerts over central aspects of people’s lives is particularly fine-grained. For reasons that can be broached only fleetingly here, the minuteness of the fine-grained control is both overweening and degrading. It is overweening because it assigns to governmental officials a decisive role in channeling the specificities of communicative interaction qua communicative interaction, 2 and it is degrading because it makes the success of a system of governance partly dependent on the non-occurrence of modes of expression which an ethically more robust system of governance could safely abide. Toleration is of course not equivalent to condonation; as I have emphasized in Freedom of Expression as Self-Restraint, a key aspect of the ethical robustness of a system of governance is the extent to which its operations avert any untoward effects that might ensue from the modes of expression which it leaves legally unrestricted. Consequently, a prime measure of the ethical health of a system of governance is its ability to sustain the self-restraint that is constitutive of freedom of expression. We can recognize as much when we ruminate on the types of neutrality (in the domain of communicative interaction) that are required under the principle of freedom of expression. Chapter 4 of Freedom of Expression as Self-Restraint has dilated on this matter at far greater length, and has thereby greatly clarified and deepened the highly compressed remarks in this paragraph. Here, my aim has simply been to provide a glimpse of a few strands of the theory developed there.
