Part I: Introduction
Freedom of speech is among the most cherished constitutional rights in liberal democracies. It is entrenched in most contemporary constitutions as well as in international human rights treaties. It is often classified as a “first generation right”—a right protecting individuals from interference by the state. It is understood to be foundational to liberal polities either in the sense that it is a precondition to the existence of a liberal polity and/or that it is tightly related to liberal values such as autonomy, dignity and liberty. At the same time, the scope of what constitutes speech, what speech ought to be protected, the weight or the value attributed to the protection of speech vis-à-vis other rights or policy concerns, and the reasons underlying its protection are highly controversial. These controversies have important political and legal implications and they are reflected in the differential protection granted to speech in different jurisdictions.
The primary philosophical challenge is to explain why (and whether) speech ought to be protected more (or differently) than non-speech activities. When we protect speech we privilege speech relative to non-speech activities. The normative debate concerning the justifications for protecting speech also sheds light on what counts as speech. Only communicative action that at least potentially promotes the values underlying the protection of speech counts as speech. Consequently, identifying the values underlying the protection of speech also influences what activities count as “speech.” Thus, at least in legal discourse, the question of what counts as speech and what counts as protected speech are often interrelated. To address the normative question of why speech is protected, as well as to identify what counts as speech, we examine below in Part II four major theories purporting to justify the protection of speech: the marketplace of ideas, the autonomy-based theory, the self-realization justification and the democratic justification. We establish that none of these theories alone can justify the protection of speech as is currently practiced in contemporary liberal polities.
In recent years many feminists and ethnic or religious minorities have challenged the protection of certain forms of speech, e.g., racist speech or pornography. More particularly it was claimed that certain forms of speech either conflict with other rights, e.g., equality, or even may deprive minorities of the capacity or ability to exercise effectively their own right to free speech (the s’ilencing argument). It is the task of Part III to explore some of the minoritarian challenges to what constitutes speech and what constitutes protected speech.
Some radical critics of liberalism challenge the importance and significance of freedom of speech as such. They maintain that freedom of speech masks large-scale silencing and repression. “Repressive tolerance,” as it is sometimes labeled, is a radical position that rejects central, traditional, liberal political rights in the name of values such as autonomy and equality and regards the liberal protection of speech (independently of its content or merit) as a repressive mechanism designed to strangle rather than facilitate genuine public deliberation. These challenges will be examined in Part IV. We conclude by pointing out that the three positions described below (liberalism, minoritarian critics of liberalism and radical critics of liberalism) share similar assumptions and values. To the extent that minoritarian and radical critics of liberalism advocate restrictions of speech, they do so by invoking the very same values advocated by liberals: autonomy, dignity and equality.
Part II: Rationales for Protecting Speech
The scope of what the right to free speech includes is of course controversial. Often the scope of what constitutes “speech” is influenced by normative considerations. Yet, it is evident that the term “speech” is much too narrow to describe all the activities that are traditionally covered by the right to free speech. As Schauer noted: “What is ‘speech’ in ordinary usage is not necessarily what is ‘speech’ for purposes of the concept of free speech” (Schauer 1982: 13). Waving a flag, wearing a button with political symbols and producing a movie are also protected by the right to free speech. In contrast, there are activities that are clearly speech (in the ordinary sense of the word) that are not protected by the right to free speech, such as hiring somebody to commit murder. Often the right to free speech protects communicative activity—namely activity that conveys ideas, expresses emotions or sentiments, or conveys or evinces attitudes. Yet not all communicative activity is protected; physically attacking a person as an expression of hatred is not covered by the right to free speech even if it is a communicative activity.
The standard view is that free speech differs from a principle of general liberty. The protections granted to speech are far greater than the protection of activities that are not classified as speech (Schauer 1982: 7–8; Greenawalt 1989a: 120). This does not imply that the protection of speech is absolute; most advocates of protecting speech concede that urgent considerations often override the concern for protecting speech. Yet concerns which justify limitations on speech ought to be more urgent, more weighty or different in kind than concerns justifying the limitations of most other liberties. The question is, therefore, why speech should have greater protection compared with other types of activities that may be just as important and value enhancing as non-speech activities.
The special protection of speech is part of a more general phenomenon characterizing many rights; rights protect certain forms of behavior, e.g., speech, religion, equality, etc. They provide, therefore, differential protection to different activities. One of the great challenges of a theory of rights in general and a theory justifying the protection of any particular right is to explain the reasons underlying the differential protection of activities, all of which seem to produce similar benefits and generate similar harms. In the context of speech, we can ask why should speech and non-speech activities which are equally autonomy enhancing (or, more generally, equally value enhancing) be protected differentially (Harel 2005)? We shall address this question at the end of this section.
The rest of this section is devoted to explaining the rationales justifying the protection of speech. The traditional justifications for protecting speech fall into two categories. Under the first category, speech is protected in order to shield an individual from restrictions even when such restrictions would be conducive to welfare. Dignity-based and autonomy-based concerns fall into this category; their advocates often maintain that speech ought to be protected even when its protection is detrimental to important societal values. Under the second category, protecting speech is justified because in the long run it is conducive to social welfare. The claim that protecting speech is conducive to the discovery of truth is an influential example of a justification of the second category; it is founded on the conjecture that protection of speech is conducive to the society as a whole. Note that this difference in the type of justification has important ramifications. The first camp (the deontological camp) is typically less willing to conduct “balancing” of speech concerns with other non-speech concerns. In contrast, the second camp (the consequentialist camp) regards the protection of speech as a means to promote social welfare. It follows that speech, which is not conducive to social welfare, deserves no protection (Nagel 1995: 86–89).
Let us examine, therefore, four main justifications used in the literature: the marketplace of ideas, the autonomy-based argument, the self-realization argument and the democracy argument. Note that these justifications are only a subset of the potential justifications for rights-based protection of speech and that the categorization provided here is not the only categorization provided in the literature. Furthermore, like many classifications, there are arguments that can fall into more than one category.
The Marketplace of Ideas
Under the conventional marketplace of ideas argument, the protection of speech is conducive in the long run to the discovery of truth. It is evident that this argument is not an individualistic argument based on either the rights of the speaker to speak his mind or the right of the audience to benefit from the speech. Instead, it is based on long-term societal benefits resulting from the protection of speech. Speech has an instrumental value in promoting truth and promoting truth is socially valuable. One of the earliest attempts to justify the protection of speech dates back to the seventeenth century. John Milton argued that:
[T]hough all the winds of doctrine were let loose to play upon the earth, so Truth be in the field, we do injuriously by licensing and prohibiting to misdoubt her strength. Let her and Falshood grapple; who ever knew Truth put to the worse, in a free and open encounter?
But why should truth and falsehood grapple with each other? Why should not we simply censor falsehood and thus guarantee the victory of truth? Why is the victory of truth guaranteed? A philosophically sophisticated version of the argument was developed by John Stuart Mill, who identified three distinct claims (Mill 1859). In Mill’s view: 1) if a censored opinion contains truth, its silencing is damaging as it lessens the probability that truth be revealed. In his view: “complete liberty of contradicting and disproving our opinion, is the very condition which justifies us in assuming its truth”; 2) if conflicting opinions each contain some truth, the clash between them is the only method of discovering what the truth is; and 3) even if the opinion has no truth in it, challenging the accepted position contributes to its vitality and decreases the chances that it degenerates into a prejudice or dogma. Mill famously contrasted “dead dogma” with “living truth” and he maintained that: “Truth gains more even by the errors of one who, with due study and preparation, thinks for himself, than by the true opinions of those who only hold them because they do not suffer themselves to think” (Mill 1859).
Mill’s view has been immensely influential; it has been endorsed by courts and by numerous legal theorists (Ingber 1984). In a dissenting opinion that later became orthodoxy in the United States, Justice Oliver Wendell Holmes Jr. argued that: “the best test of truth is the power of the thought to get itself accepted in the competition of the market” (Abrams v. U.S. (1919) 250 U.S. 616). The claim that robust discussion is conducive to the discovery of truth has a particular appeal in the Anglo-American world and it has been analogized to the traditional justification for the adversary system based on cross-examination (Schauer 1982: 16).
Ironically, the economists who first identified the virtues of the marketplace of commodities were quick to point out the paradoxical nature of this analogy. Ronald Coase asked the following question: if indeed the protection of freedom of speech is grounded in the marketplace of ideas, why is speech so much more protected than the marketplace of commodities? His answer was cynical: “The market for ideas is the market in which the intellectual conducts his trade. The explanation of the paradox is self-interest and self-esteem. Self-esteem leads the intellectuals to magnify the importance of their own market… But self-interest combines with self-esteem to ensure that, while others are regulated, regulation should not apply to them” (Coase 1974: 386; Coase 1977: 1). Coase also pointed out that perhaps the real motive for protection of freedom of speech is not the keen interest in truth but the fact that “the public is commonly more interested in the struggle between truth and falsehood than it is in truth itself” (Coase 1974: 391).
Even if one ultimately rejects Coase’s cynical explanation for the special protection of speech on grounds of self-interest and self-esteem of intellectuals, the differential treatment of speech on the one hand and commodities on the other hand is a challenge to the advocates of this argument. The marketplace of ideas clearly increases the chances of “true ideas” to appear in the market of ideas, but it also increases the spread of false ideas (Schauer 1982: 28; Ingber 1984: 7). The famous assertion that false speech should be countered by “more speech, not enforced silence” is based on an empirical generalization that truth has greater chances of winning in the marketplace of ideas. Mill himself concedes that truth does not prevail if persecuted and says that: “It is a piece of idle sentimentality that truth, merely as truth, has any inherent power denied to error of prevailing against the dungeon and the stake” (1859). Yet the same argument ought to be extended to the case in which truth is not persecuted; the premise that freedom of speech leads on the whole to the discovery of truth is not more than an unsubstantiated conjecture (Ingber 1984; Schauer 1982: 28; Barendt 2005: 9). It has also been argued correctly that the so-called marketplace of ideas is not open to everyone who wants to communicate her ideas (Barendt 2005: 12). Disparities of power and money may have destructive influence on the robustness of public discourse. Furthermore, even if protecting speech is conducive to truth, it is not clear that truth is always or even typically desirable. Often the regulation of speech is based on arguments of public policy that have nothing to do with truth or falsehood (Schauer 1982: 22–23). A true expression such as publication of a criminal record of a defendant may undermine her prospects for a fair trial. True expressions may violate privacy, undermine stability and lead to violence. Some theorists also have pointed out that the marketplace of ideas argument presupposes a process of rational thinking and, consequently, the less rational individuals are, the less forceful the theory of the marketplace is (Schauer 1982: 30; Ingber 1984: 15; Baker 1982: 14). Most skeptical of all have been those who have argued against the claim that truth is objective and maintain that truth is being created rather than discovered (Baker 1989: 13).
Yet despite these difficulties there are less ambitious arguments based on the marketplace of ideas that have merit. Mill was right that challenging even true ideas increases the vitality and potency of these ideas. Furthermore, as Mill himself noticed, challenging even true ideas is conducive to the very critical capacity to scrutinize ideas and, therefore, in the long run, may be as conducive to truth as it is conducive to the development of the critical deliberative tools necessary to evaluate truth and falsehood. Under this understanding, free speech is conducive to truth because it is conducive to the development of intellectual virtues that, in turn, are instrumental to truth. Furthermore, the claim that free speech is conducive to intellectual virtues may support the protection of speech even if the development of such virtues is not conducive to truth, as it seems that the perfection of intellectual virtues is valuable independently of whether such perfection is conducive to truth.
The Autonomy-Based Argument
The autonomy-based argument is only one of several characteristically liberal arguments made in favor of protecting speech. These arguments include arguments from dignity, diversity and autonomy (Schauer 1982: 60–72). It is to be emphasized that the autonomybased argument is in effect a family of arguments. One theorist identified six distinctive meanings of autonomy that are used in defending the right to free speech (Briston 1998). We will examine here only two distinct autonomy-based justifications: one based on a negative conception of autonomy and another based on a positive conception of autonomy. Under the negative concept of autonomy, autonomy is designed to protect individuals from outside control of the state and to maintain a personal space for the individuals; in contrast the positive conception is designed to guarantee the actual exercise of autonomy. It is not merely the protection from outside or external interference which counts as autonomy enhancing but the actual, active exercise of one’s deliberative powers.
Under the justification based on a negative conception of autonomy, speech is a private domain—an area that is under the exclusive control of the individual. Often this argument is based on the view that protection of rights has intrinsic rather than instrumental value (Nagel 1995: 86). The right to free speech is, under this view, a fundamental and nonderivative element of morality. Thomas Scanlon developed the most sophisticated version of this argument. In Scanlon’s view: “To regard himself as autonomous a person must see himself as sovereign in deciding what to believe and in weighing competing reasons for action… An autonomous person cannot accept without independent consideration the judgment of others as to what he should believe or what he should do” (Scanlon 1972: 215–16). A similar reference to sovereignty of the person is made by Nagel, who argues that: “The sovereignty of each person’s reason over his own beliefs and values requires that he be permitted to express them… it also requires that he not be protected against exposure to views or arguments that might influence him in ways others deem pernicious” (Nagel 1995: 96). In both cases what is at stake is the view of persons as autonomous agents, and autonomy is identified with what can be labeled sovereignty—immunity from certain forms of external control. An individual is sovereign if she does not accept without questioning the judgments of others. Scanlon rejects the instrumental arguments for protecting speech (based on societal well-being) and adopts instead a noninstrumental, non-consequentialist, highly individualistic approach. If speech is to be protected solely on the basis of controversial, empirical claims about its long-term positive effects or the negative effects of restrictions, then the free speech principle is vulnerable to empirical challenges and subject to contingent factors concerning its expected societal effects (Scanlon 1972: 205).
Scanlon starts his discussion with a very important and seemingly paradoxical observation. The right to free speech does not target restrictions on speech as such but only, or primarily, restrictions that are grounded in certain justifications for these restrictions. Particularly pernicious are restrictions that are based on the risk that the views communicated were to be believed or acted upon (Scanlon 1972: 209). Compare for instance restricting the dissemination of information as to how to make nerve gas in the kitchen to a restriction of a particularly effective piece of political propaganda that may lead to schism and bloody civil war. Both restrictions are designed to promote the public good and the harms resulting from the speech in both cases could be similar. Yet it is the second restriction that seems problematic. This is because the speech restricted in the second case is one that promulgates what the speaker takes to be good reasons for action, while in the first case it merely provides individuals with the means to do what they want to do anyhow (Scanlon 1972: 211–12). Similarly, if someone urges me to commit violence the choice to commit violence is ultimately mine: “The contribution to the genesis of [the] action made by the act of expression is, so to speak, superseded by the agent’s own judgment” (Scanlon 1972: 212).
These observations led Scanlon to articulate what he labels the Millian principle that consists of two sub-principles. Restrictions of speech could not be justified on the grounds that a) the speech harms those who come, as a result of it, to have false beliefs; or b) the speech harms people as a result of harmful acts committed as a result of the fact that the act of expression leads agents to believe (or increased their tendency to believe) these acts are worth performing. Scanlon also shows that his analysis can account for famous exceptions to the protection of speech, e.g., shouting “fire” in a crowded theater.
Scanlon himself recognized some of the limitations of the Millian principle. First, the Millian principle is under-inclusive; restricting all speech for the sake of saving resources necessary for protecting speakers is consistent with the Millian principle as the restriction is justified on grounds that are not recognized by the Millian principle as justifying protection of speech. Second, the Millian principle is over-restrictive as there are clearly justified restrictions that violate the Millian principle, such as restrictions on speech that create “clear and present danger” (Schenck v. United States (1919) 249 U.S. 47). For instance, a speaker who urges his audience to condemn all homosexuals as they are morally depraved may under certain circumstances create clear and present danger (because the audience may act upon the belief that homosexuals are evil). Hence, such speech may fall within the scope of the second part of the Millian principle: the speech harms people as a result of harmful acts (of violence) committed as a result of the fact that the act of expression leads agents to form beliefs (concerning the moral depravity of homosexuals). As to the problem of under-inclusiveness, Scanlon maintains that the protection of speech may be grounded in different justifications; he does not argue therefore that the Millian principle is the exclusive justification for the protection of speech. As to the problem of over-inclusiveness, Scanlon concedes that the Millian principle is not an absolute principle; it may be defeated by conflicting considerations.
Yet opponents of the Millian principle exposed additional theoretical weaknesses.
For example, why could not autonomous persons agree “to foreclose some inputs for them which they would ideally like to have in some situations” (Greenawalt 1989a: 115)? If people are autonomous should not they also be free to bar the dissemination of ideas that they know (or judge) are dangerous (Barendt 2005: 17; Briston 1998: 329)? Under this view, our moral dignity dictates to respect not only our judgments concerning what is right or wrong but also our judgments concerning what the best circumstances for deliberating about right and wrong are. Why should our second-order judgments (concerning what the best preconditions for deliberation on right and wrong are) deserve less protection than our first-order judgments (concerning what is right and wrong)?