Hostname: page-component-cd9895bd7-dzt6s Total loading time: 0 Render date: 2024-12-26T20:45:50.836Z Has data issue: false hasContentIssue false

A DEFENSE OF MILL’S ARGUMENT FOR THE “PRACTICAL INSEPARABILITY” OF THE LIBERTIES OF CONSCIENCE (AND THE ABSOLUTISM IT ENTAILS)

Published online by Cambridge University Press:  04 May 2021

Daniel Jacobson*
Affiliation:
Philosophy, University of Colorado, Boulder, USA
Rights & Permissions [Opens in a new window]

Abstract

Mill advocated an unqualified defense of the liberty of conscience in the most comprehensive sense, which he understood to include not just the freedom to hold but also to express any opinion or sentiment. Yet considerable dispute persists about the nature of Mill’s argument for freedom of expression and whether his premises can support so strong a conclusion. Two prominent interpretations of Mill that threaten to undermine his uncompromising defense of free speech are considered and refuted. A better interpretation can be founded on Mill’s claim that the liberties of conscience are inseparable in practice. This claim can be defended with modern psychological insight about the nature of cognitive bias, and epistemological insight about why justification of creedal beliefs requires the universal toleration of opinion, insights which are largely anticipated by Mill. This argument is especially vital because it highlights the divide between classical liberalism and progressivism that has become a flashpoint in the current political debate over free speech.

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in anymedium, provided the original work is properly cited.
Copyright
© Social Philosophy & Policy Foundation 2021

John Stuart Mill defended the liberty of conscience in the most comprehensive sense, which he understood to include not just the freedom to hold but also to express any opinion.Footnote 1 Yet considerable dispute persists about the nature of Mill’s argument and whether his premises can support such a strong doctrine. Once society has reached the level of development where the rights associated with the principle of liberty obtain, Mill claims: “there ought to exist the fullest liberty of professing and discussing, as a matter of ethical conviction, any doctrine” (On Liberty, XVIII: 228fn).Footnote 2 Mill characterizes these rights as absolute and unqualified, and my interpretation respects this characterization of his defense of freedom of speech. Despite his repeated statements to this effect, however, my conclusion is controversial. Although some others defend an absolutist reading, it has not been commonly accepted by Mill scholars and political philosophers for two reasons: because of worries about harmful speech; and because its advocates struggle to give a comprehensive account of his position that coheres with the utilitarian foundation Mill requires of speech (and other) rights.

I will offer a novel account of Mill’s argument that justifies his absolutism, at least by his own lights, and illustrates why this view remains vital to the current debate over freedom of speech. Two aspects of this interpretation are especially salient both for understanding the argument and for appreciating its continued relevance. The first point is exegetical. Mill claims that the liberties of conscience—freedom of thought and freedom of speech—are practically inseparable, by which he means that they are inseparable in practice. But why think so? We are at least sometimes capable of holding an opinion without expressing it, and it is possible for society to permit belief of all doctrines while restricting the expression of some. The second point is empirical. Mill gives an indirect utilitarian argument for free speech, the premises of which rest on psychological claims that I cannot adequately defend here. I will instead ask what would have to be true about humanity, in order for the utilitarian foundations of his uncompromising defense of freedom of speech to hold for creatures like us, in societies like ours. Both of these points require me to begin by defending the principled and absolutist interpretation of his argument against its rivals.

I. Mills Absolutist Defense of Freedom of Speech

Two prominent and influential interpretations of Mill, which I term the pragmatic and the qualified interpretations, make his absolutist rhetoric out to be less meaningful than it initially appears.Footnote 3 These rival views understand the principle of liberty, the express purpose of which is to distinguish those aspects of life over which the individual is sovereign from those legitimately subject to social control, to protect only action that is, roughly, harmless to others.Footnote 4 This is rough because Mill expressly notes that talk of what affects or harms others must always be understood as referring to non-consenting others who are sound-minded adults; and that some degree of danger must be counted as harm, even if it does not result in injury, in order to include lucky drunk driving and the like. For the sake of simplicity, by “harmful” action I will mean action sufficiently dangerous to non-consenting others, and analogously for “harmless” action, eliding the details except when needed for clarity. The crucial point is that since the expression of opinion and sentiment cannot be guaranteed to be harmless, any such reading of Mill’s fundamental political principle implies that harmful speech lies outside the sphere of liberty immune from coercive interference.Footnote 5

The rival interpretations compromise Mill’s defense of speech either by holding that it is based on expedience rather than rights (which is the pragmatic interpretation), or else by taking it to protect only harmless opinions (as does the qualified interpretation). Mill has no principled objection to censorship in either case, and perhaps no objection to the censorship of harmful opinions at all, despite his claims to the contrary. Hence, the rival interpretations agree that Mill acknowledges no right to freedom of speech.

According to the pragmatic interpretation, Mill defends free speech as a matter of policy rather than principle.Footnote 6 Interference with an agent’s freedom of speech is no violation of her rights. Although the expression of opinion falls under the jurisdiction of society, in this view, the costs of censorship inevitably outweigh its benefits. Hence, censorship always turns out to be bad policy. Although this universal generalization is a strong claim, it is not a principled conclusion. Mill’s argument for free speech is thus analogous to his argument for free trade. Since Mill recognizes no right to economic liberty, restrictions on trade are merely misguided rather than unjust.Footnote 7 The pragmatic interpretation claims similarly that the expression of opinion falls outside the sphere of liberty precisely because some opinions are harmful. This implies that novel proposals to censor specific opinions should be subject to the utilitarian calculus—even though Mill blatantly refuses to do so, instead arguing peremptorily for the right to express any doctrine. But this interpretation fundamentally misunderstands Mill’s conception of liberty rights.

According to the qualified interpretation, by contrast, Mill does not defend free speech, despite his claims to that effect.Footnote 8 He instead distinguishes merely offensive opinions from pernicious ones, which have bad persuasive effects; or he differentiates between especially harmful speech, which diminishes a significant good such as autonomy, and speech that harms only in less important ways. Mill would support the censorship of truly dangerous opinion and sentiment, in this view; he does not defend freedom of speech but something more like the freedom of harmless speech. Since he expressly claims that “human beings should be free to form opinions, and to express their opinions without reserve” (Liberty, 260), the qualified interpretation rests heavily on a limiting case that Mill discusses immediately after this pronouncement: the corn dealer example. But this interpretation rests on a misunderstanding of that example, the point of which is not to make an exception for harmful speech but to draw a distinction crucial for any tenable conception of free speech. The proper characterization of freedom of speech allows Mill to mean what he says in its defense.

Discussion of Mill’s defense of free speech tends to focus on On Liberty, where the doctrine gets its most extensive development. Although I focus there as well, I also draw on evidence from throughout Mill’s work to show that his argument has been misunderstood, in part because interpreters fail to appreciate the idiosyncratic nature of his utilitarianism. Certain crucial but frequently overlooked aspects of Mill’s moral philosophy explain how he can defend speech rights as absolute in this straightforward sense: No opinion can legitimately be censored on grounds of its falsity, immorality, or even its harmfulness. These are precisely the grounds on which censorship has been and is currently defended.

It is not merely of academic interest whether Mill’s defense of free speech is absolute or qualified, and if it is a matter of rights or expediency, because this interpretive dispute tracks an ongoing political debate. If freedom of speech does not lie within the protected sphere of liberty—which is to say, if it is not a moral right—then it is potentially open to exceptions he rejects. This point is especially important because the advocates of both rival interpretations typically suggest that, if Mill were in modern circumstances, he would favor censoring the opinions and sentiments that many academics consider to be beyond the pale of toleration.Footnote 9 But although these views may better suit the politics of contemporary academics, they founder as interpretations of Mill. Rather, his position marks the distinction between liberalism and progressivism, which has been obscured by the current, nearly synonymous usage of those terms. Whereas (classical) liberalism focuses on individual rights and personal responsibility, progressivism is collectivist and overtly skeptical of those individualistic notions, which leads it to promote egalitarian outcomes with less concern for procedural impediments.Footnote 10

The contrast between liberalism and progressivism is illustrated by the recent proposals to prohibit certain opinions for their harmfulness or immorality currently advocated by many progressives, including prominent scholars at the most prestigious American law schools.Footnote 11 Many progressive politicians now share the view that “hate speech” should be censored in the United States; some seem to think it already is. Former Vermont governor Howard Dean, who ran for president in 2004 on a progressive platform, has claimed—incorrectly but without retraction—that “Hate speech is not protected by the first amendment.”Footnote 12 And such transnational or post-national institutions as the United Nations and the European Union have endorsed the punishment of hate speech, Holocaust and climate change denial, and defamation of religion (however selectively this is enforced in practice).Footnote 13 Of the various targets of modern censorship, the most politically significant challenge surrounds those opinions classified as hate speech and sometimes even equated with violence.

Although this essay is primarily exegetical, I will show that the best interpretation of Mill’s argument has novel and underappreciated resources for defending freedom of speech against its current antagonists. In the following section (Section II), I will show that Mill’s defense of free speech is principled rather than merely pragmatic, contrary to the pragmatic interpretation. That is to say, it concerns the rights constitutive of a free society, which comprise those moral rules most crucial for human flourishing and progress. Then (in Section III), I will show that Mill defends the toleration of even harmful opinion, contrary to the qualified interpretation, which misconstrues the crux of the corn dealer example. This example belies rather than supports that interpretation. Finally (in Section IV), I argue that neither view can make sense of his claim that the liberty of thought “is impossible to separate [from] the cognate liberty of speaking and of writing” (Liberty, 227). Recent work in social epistemology supports Mill’s claim about the practical inseparability of the liberties of conscience, which is frequently obscured by a myopic reading of the arguments in Chapter Two of On Liberty.

Rival interpretations that suggest Mill would embrace such censorship, or that it can be supported on his premises, are misguided. These progressive challenges to freedom of speech would be anathema to Mill, its quintessentially liberal champion. Nevertheless, since his argument adverts to speech rights that are grounded in utility, it is open to opponents to claim that a better utilitarian argument would make exceptions for certain opinions, such as those classed as hate speech. Yet recent psychological work on cognitive bias and intellectual diversity show that Mill had prescient insights about human nature. These insights support his indirect utilitarian defense of freedom of speech against those who would undermine it by placing harmful and hateful opinion beyond the pale of toleration.

II. Speech and the Rights of Individuality: Against the Pragmatic Interpretation

The pragmatic interpretation of Mill allows him to defend freedom of speech, after a fashion, because it takes his absolutism to be unprincipled: it does not follow from the principle of liberty. In Mill’s terms, the issue is whether the expression of opinion is self-regarding action (which falls within the sphere of liberty protected from coercion) or social action (which is in principle subject to coercive interference). These are terms of art for Mill. Readers can be excused for confusion over his tendency to move casually between talk of what affects, harms, interests, and concerns others; but these are importantly different notions. I contend that the crucial distinction must be understood as differentiating what primarily concerns or chiefly interests the agent from what primarily concerns society—even though this is a disputed normative question rather than a seemingly determinate claim about harm. This reading is so superior to its alternatives, both theoretically and exegetically, that the gain more than compensates for its cost.

Over the course of this discussion, we will see that there are three reasons why self-regarding action must not be understood as action that only affects, concerns, or harms the agent (and consenting others). First, this gloss undermines the sphere of liberty at the heart of Millian liberalism, including self-sovereignty, which is so central to the principle of liberty that the paragraph introducing it concludes: “Over himself, over his own body and mind, the individual is sovereign” (Liberty, 224). Second, my gloss jibes with Mill’s most careful statements of the doctrine, such as this one: “To individuality should belong the part of life in which it is chiefly the individual that is interested; to society, the part which chiefly interests society” (Liberty, 276). This claim leaves a promissory note to be cashed, because it does not yet specify which acts fall into each category: what part of life belongs to individuality and what part belongs to society. When Mill does flesh out his view, he expressly notes that self-regarding action can be harmful (to non-consenting others); and he acknowledges the necessity of excluding some harms as reasons for compulsion, in order to secure a sphere of liberty. Third, the teleological structure of his ethics, which grounds rights in utility, does not require the distinction between self-regarding and social action to be reduced to any empirical distinction between the harmless and the harmful. The case of freedom of speech is illustrative on all three points.

The model for the pragmatic interpretation is given by Mill’s defense of the doctrine of free trade, understood (inter alia) as the claim that the quality and inexpensiveness of commodities are best secured by leaving producers and sellers free to set prices, and buyers free to purchase goods elsewhere. He claims nevertheless that trade falls under the jurisdiction of society, since he explicitly declares it to be a social rather than a self-regarding act.Footnote 14 Despite defending free trade, he holds that it “rests on grounds different from, though equally solid with, the principle of individual liberty” (Liberty, 293). His objections to interference with trade are grounded in expedience rather than rights. They rest on the inefficacy of price controls and similar restraints, as well as their bad unintended consequences. It is irrelevant for our purposes whether Mill was correct about this economic doctrine. His argument for free trade is by admission a matter of expediency—in explicit contrast to his rejection of any prohibition on the sale of goods that have a permissible use, which violates the rights of prospective consumers.Footnote 15 This appeal to expedience in defense of free trade dovetails with his denial that trade is self-regarding action, despite its taking place between consenting adults for their (expected) mutual benefit.

According to the pragmatic interpretation, Mill’s defense of free speech is analogous to his defense of free trade. If the expression of opinion is a social rather than a self-regarding act, then his defense of free speech is not a rights claim but a universal generalization that censorship is always bad policy. This conclusion invites the suggestion that some better program of censorship—one targeted at genuinely harmful and low-value opinions—will produce sufficiently good results as to outweigh the costs of intolerance. But the pragmatic interpretation crucially misconstrues both Mill’s conclusion and his argument. Although he sometimes casually refers to self-regarding action and character traits as those that do not affect or harm others, he repeatedly qualifies this notion, particularly in his most detailed discussion of the subject. There he expressly grants that some self-regarding actions can affect, and even harm, non-consenting others.

This point is crucial because the misidentification of self-regarding action with harmless action diminishes the sphere of individual liberty in ways inconsistent with Mill’s conviction, repeated throughout his work, that there is a substantial sphere of liberty: “a circle around every individual human being, which no government … ought to be permitted to overstep” (Principles of Political Economy, III: 938).Footnote 16 This claim that there is a substantial sphere of individual liberty is nothing less than the crux of Millian liberalism. Indeed, Mill retrospectively summarized the argument of On Liberty as a defense of “the doctrine of the rights of individuality” (Autobiography, I: 260). The gloss of the self-regarding as the harmless undermines this doctrine, because almost everything we do affects others without their consent. Mill rejects this gloss for just that reason.

Immediately after he claims that there is a sphere of liberty properly immune to compulsion—which he initially identifies as comprising action that affects only the agent himself—he continues: “When I say only himself, I mean directly, and in the first instance: for whatever affects himself, may affect others through himself”; and “the objection which may be grounded on this contingency will received consideration in the sequel” (Liberty, 225).Footnote 17 By “in the sequel,” Mill means in what follows; and when he considers this objection, he writes:

I fully admit that the mischief which a person does to himself may seriously affect, both through their sympathies and their interests, those nearly connected with him, and in a minor degree, society at large. When, by conduct of this sort, a person is led to violate a distinct and assignable obligation to any other person or persons, the case is taken out of the self-regarding class, and becomes amenable to [punishment].Footnote 18 (Liberty, 281)

This qualification shows that Mill anticipates the objection that, since almost every action can potentially harm others, he cannot secure his central claim about the rights of individuality and the sphere of liberty: the idea that there are “large departments of human life from which [coercion] must be unreservedly and imperiously excluded” (Political Economy, III: 937). His response is that what makes a typically self-regarding type of action lose that status is not that it harms others but that it violates an obligation. If all acts harmful to non-consenting others were violations of their rights, then they would all thereby breach an obligation. Hence, this cannot be Mill’s view. He states explicitly that the conduct society may demand of an individual is not that she never harm others, even without their consent, but that she does not harm “certain interests which, either by express legal provision or by tacit understanding, ought to be considered as rights” (Liberty, 276).Footnote 19 The obvious implication of this statement, which Mill corroborates repeatedly, is that we do not have a right to protection against all harms. We can legitimately be subject not only to offense but, in some cases, to real harm to our interests.

Hence, in order to determine whether any given action is self-regarding, it must be specified what are one’s obligations and which of one’s interests ought to be protected as rights. This does not simply fall out of a simple causal claim about what can affect (or harm) no one but the agent himself. In the following section, I will illustrate how this conception of self-regarding action is consistent with a teleological metaethics, which grounds moral rules and moral rights in utility. For now it will suffice to note that, although he does not set out a complete theory of rights and obligations, Mill expressly characterizes the sphere of liberty as comprising self-sovereignty and the freedoms of expression and association. These rights of individuality rule out certain claims about what society can rightly compel. He thus unequivocally rejects the claim that private intoxication can be prohibited in order to create a moral ecology of sobriety.

So monstrous a principle is far more dangerous than any single interference with liberty; there is no violation of liberty which it would not justify; it acknowledges no right to any freedom whatever, except perhaps to that of holding opinions in secret, without ever disclosing them; for, the moment an opinion which I consider noxious passes any one’s lips, it invades all the ‘social rights’ attributed [to] me … (Liberty, 288).

Not only does Mill reject the proto-communitarian claim that society can interfere with the individual’s sovereignty over her own mind and body by enforcing sobriety, he specifically objects to the idea that society can compel individuals not to express noxious opinions. This strongly suggests that Mill considers the expression of opinion to count as self-regarding, even though it can affect others without their consent. This is exactly what he says, after introducing the principle of liberty in Chapter 1 of On Liberty. He then lays out how his argument will proceed:

It will be convenient for the argument, if, instead of at once entering upon the general thesis [of the sphere of liberty], we confine ourselves in the first instance to a single branch of it, on which the principle here stated is, if not fully, yet to a certain point, recognized by the current opinions. This one branch is the Liberty of Thought: from which it is impossible to separate the cognate liberty of speaking and of writing. (Liberty, 227)

He then dedicates Chapter 2 (“Of the Liberty of Thought and Discussion”), to arguing for this single aspect of liberty. The structure of Mill’s argument in On Liberty, which makes freedom of speech exemplary of the rights of individuality, thus belies the pragmatic interpretation.

Defenders of this view must claim that only the liberty of thought counts as part of the sphere of liberty, strictly speaking, and that the cognate liberties are included in this discussion by courtesy. This reading seems implausible on its face, since Chapter 2 is dedicated almost entirely to the freedom of speech, but its advocates offer two arguments for it. First, they claim (or more typically assume) that, because the public expression of opinion obviously affects others and can harm them without their consent—for instance by offense or persuasion—it must count as a social act. This cannot simply be assumed, however, when Mill grants that some harmful actions are nevertheless self-regarding. Second, there is an ambiguous passage in On Liberty frequently cited in defense of the pragmatic interpretation.Footnote 20 Mill writes that the expression of opinion “may seem to fall under a different principle, since it belongs to that part of the conduct of an individual which concerns other people; but, being almost of as much importance as the liberty of thought itself, and resting in great part on the same reasons, [the liberty of expression] is practically inseparable from it” (Liberty, 226; emphasis added).

Since the claim that expression concerns other people is offered as a reason why it might seem to fall under a different principle than thought, this passage is ambiguous. It is unclear whether this claim is made in Mill’s own voice, or if it is offered as the reason why freedom of speech misleadingly appears to fall under a different principle than freedom of thought. Because his writing is often superficially inconsistent—sometimes deliberately, for the sake of simplicity—it is a mistake to base any interpretation on a close reading of a single passage.Footnote 21 Taken in context, it is apparent that the intention of this paragraph is to describe the sphere of liberty. Consider how he describes that sphere in the Principles: “I apprehend that it ought to include all that part which concerns only the life, whether inward or outward, of the individual, and does not affect the interest of others, or affects them only through the moral influence of example” (Political Economy, 938; emphasis added). Here too, Mill first casually refers to what does not affect or harm others, and then qualifies this simple notion by excluding certain harms—in this case, harm from bad example.

In On Liberty, Mill not only specifies the region of human liberty as including freedom of speech, he begins Chapter Two by claiming that even when there is popular support for the intolerance of some opinion: “I deny the right of the people to exercise such coercion, either by themselves or by their government. The power itself is illegitimate” (Liberty, 229). This claim is explicitly about the expression of unpopular opinion. It is antithetical to his pragmatic defense of free trade, which grants that trade primarily concerns society but maintains that its regulation, though legitimate in principle, should be eschewed as counterproductive.

Indeed, a comparison of Mill’s defense of free speech with his defense of free trade undermines the analogy between them at the heart of the pragmatic interpretation. With respect to trade, he notes that competition involves opposing interests, and that the winners of a fair competition will inevitably harm the interests of the losers.

But it is, by common admission, better for the general interest of mankind, that persons should pursue their objects undeterred by this sort of consequences. In other words, society admits no right, either legal or moral, in the disappointed competitors, to immunity from this kind of suffering. (Liberty, 292–93)

Notice that Mill does not endorse weighing the costs and benefits of interference with competition, so as to determine when the suffering of the disappointed outweighs the happiness of the successful.Footnote 22 Rather, because competition is crucial to progress, according to a tenet of classical liberalism that Mill expressly endorses, there is no right to protection from this kind of suffering.Footnote 23

Orthodox forms of utilitarianism cannot rule out consideration of competitive harm, sympathetic harm, the harm of bad example, or even those harms that arise from what an individual does to himself or with consenting others. Nor can they respect rights, as Mill clearly does—so clearly and so consistently as to belie the most common interpretations of his moral and political philosophy. In the following section, we will see that his theory of rights is consistent between On Liberty and Utilitarianism, in particular. And we’ve already seen that he excludes certain kinds of harm as not providing reason to interfere with liberty. I have not argued for this view or even fully explicated it, and it is open to an orthodox utilitarian or modern progressive to reject any such sphere of liberty—even simply on the grounds that its defense requires excluding certain genuine harms as reasons for social compulsion. My aim is rather to illustrate two crucial points. First, this doctrine of exclusion coheres with Mill’s defense of a substantial sphere of liberty that includes freedom of speech. Second, the grounds for this exclusion are themselves utilitarian, though indirect. Mill holds that it is better for the general interests of humanity not to treat such harm as justifying interference with individual liberty.

Since Mill denies that we have any right to protection from sympathetic harm over what people do, it is no surprise that he holds the same view of offense over what people say. He acknowledges that many people have antipathy toward conduct and opinion they find offensive, which is often grounded in deeply held religious and moral conviction. Nevertheless, “there is no parity between the feeling of a person for his own opinion, and the feeling of another who is offended at his holding of it; no more than between the desire of a thief to take a purse, and the desire of the rightful owner to keep it” (Liberty, 283). This lack of parity is not a matter of the quantity of net happiness, since in either case that could go either way. It is a matter of rights. Similarly, since Mill excludes the harm of bad example from other people’s actions, we should expect him to hold the same thing about the harm of persuasion from their professions of opinion. The crucial point for present purposes is that these are exactly the harms most likely to be caused by the expression of opinion and sentiment.

Whereas the pragmatic interpretation implies that regulations on the expression of opinion should be subject to direct utilitarian calculation, Mill’s positive argument too appeals to rights rather than expedience. He claims:

If all mankind minus one, were of one opinion, and only one person were of the contrary opinion, mankind would be no more justified in silencing that one person, than he, if he had the power, would be justified in silencing mankind. (Liberty, 229)

The peremptoriness of this claim contradicts the pragmatic interpretation, which makes his defense of free speech into a universal generalization about the costs and benefits of intolerance. Mill’s argument seems to resemble a natural rights approach, on which the right to freedom of expression does not rest on considerations of utility. But neither approach can accommodate the argument, since orthodox utilitarianism is inconsistent with his respect for rights, and the natural rights view is incompatible with their teleological grounding. Mill’s denial of any right to protection from harm by offense and persuasion allows him to argue for the toleration of all opinions. And only the claim of a moral right to freedom of speech implies that mankind would be no more justified in silencing one person than she would be in silencing mankind.

In the final section of the essay, I will explain why the idiosyncratic nature of Mill’s utilitarianism supports freedom of speech over more qualified alternatives. In his view, the expression of opinion falls within the rights of individuality despite affecting others without their consent, sometimes harmfully. This conclusion is exactly what is at issue between classical liberals and progressives in the current debate over free speech. Thus far my aim has been to demonstrate his commitment to freedom of speech as a matter of rights, contrary to the pragmatic interpretation. It still must be explained why Mill holds that there is no right to protection from certain kinds of harm, specifically those must likely to issue from the expression of opinion. The explanation will illuminate his reasons for considering the right to free expression of opinion to be “practically inseparable” from the right to free thought, and it will provide novel resources in support of his absolutism against more qualified doctrines that make exceptions for hateful and harmful opinions.

III. Speech Rights and Harmful Speech: Against the Qualified Interpretation

Despite the prominence of rights claims in Mill’s moral and political philosophy, familiar worries about the ability of utilitarianism to respect rights make it worthwhile to illustrate the coherence between the account of rights given in On Liberty and Utilitarianism. It is noteworthy that he offers the same account of moral rights in both works (and elsewhere). There are two aspects of this account that might seem to be in tension. On one hand, rights are founded in utility; on the other, rights trump utility. Both aspects of Mill’s theory of rights are crucial, and the tension between them is superficial. Since rights are founded in utility for Mill, self-regarding action can be characterized in terms of those harms against which we have a right to protection, consistent with his teleological metaethics. We need not understand self-regarding action as action that is harmless to non-consenting others, and doing so would undermine Millian liberalism.

Put most simply, rule utilitarianism can accommodate all these claims. My reading of Mill as a rule utilitarian is not uncontroversial, and it cannot be fully defended here.Footnote 24 It must suffice to show that this view—and only this view—makes sense of what he says both about rights in general and freedom of speech in particular.Footnote 25 Since even the most sophisticated, multilevel versions of act utilitarianism treat moral rules as heuristics, that theory is inconsistent with the claim that moral rules issue in genuine obligations, let alone in rights that override considerations of utility.Footnote 26 It cannot accommodate the peremptoriness of Mill’s defense of free speech, on which society would be no more justified in silencing one individual than she would be in silencing all of society. For act utilitarianism, everything rests on an accounting of the costs and benefits, which differ drastically between these cases.

According to rule utilitarianism, the deontological status of an action is determined by those moral rules whose acceptance is most conducive to human flourishing, considered on the whole and in the long run. Rule utilitarianism can ground rights in utility (axiologically), despite holding that they trump considerations of utility (deontologically).Footnote 27 This is exactly what Mill claims. He accepts the principle of utility, understood as the axiological claim that all and only happiness is intrinsically good, as the ultimate foundation of moral metaphysics.Footnote 28 Nevertheless, he insists that secondary principles—moral rules of right and wrong action—generate binding obligations. Moreover, the most important of these secondary principles comprise “certain classes of moral rules, which concern the essentials of human well-being more nearly, and are therefore of more absolute obligation, than any other rules for the guidance of life,” because they establish “a right residing in an individual” (Utilitarianism, 255).

This is precisely what Mill claims about freedom of speech: that it is essential to human well-being. His summation of the argument of Chapter 2 of On Liberty concludes: “We have now recognized the necessity to the mental well-being of mankind (on which all their other well-being depends) of freedom of opinion, and freedom of the expression of opinion” (Liberty, 257-58). Mill goes so far as to declare it “irrelevant and out of place” to discuss whether the advocacy of tyrannicide is immoral, because his defense of speech includes all opinions, including immoral ones (Liberty, 228fn). It is thus absolute in the sense of being unqualified by the value of the opinion: it applies not only to false and immoral doctrines, but also to hateful sentiments and harmful opinions.Footnote 29 “However positive any one’s persuasion may be, not only of the falsity but of the pernicious consequences” of an opinion, its discussion and advocacy must be tolerated (Liberty, 234). Mill’s explicit advocacy of the freedom to express any opinion provides a strong case against the qualified interpretation. Yet it must be admitted that he tends to make broad pronouncements, even exaggerations, which he later qualifies.

The qualified interpretation relies crucially on an example Mill gives of speech that is not immune from coercive interference: the corn dealer case. If this case is an exception based on the harmfulness of the opinion expressed, then Mill’s absolutist rhetoric should be taken as another exaggeration. But if this example is instead intended to illustrate how to construe freedom of speech such that it can be defended without any such qualification, then the corn dealer case provides decisive support for the absolutist reading. Here is what Mill says:

No one pretends that actions should be as free as opinions. On the contrary, even opinions lose their immunity, when the circumstances in which they are expressed are such as to constitute their expression a positive instigation to some mischievous act. An opinion that corn dealers are starvers of the poor, or that private property is robbery, ought to be unmolested when simply circulated through the press, but may justly incur punishment when delivered orally to an excited mob assembled before the house of a corn-dealer, or when handed about among the same mob in the form of a placard. (Liberty, 260)

Notice that the expression of these opinions loses its immunity to interference only in a specific context where it constitutes incitement to riot. Moreover, although the “mischievous act” instigated by this speech is harmful, it cannot be the harmfulness of the opinion itself that removes this token act of expression from the self-regarding class. The opinion that corn dealers are starvers of the poor harms the interests of corn dealers, insofar as people are persuaded by it, but Mill insists on its toleration in ordinary discursive contexts. The point of the example is that its profession can be prohibited and punished only when the context renders it a thinly veiled instigation to violence against this corn dealer.Footnote 30 Consider too the other opinion mentioned in this passage. Mill believed Proudhon’s dictum, “Property is theft,” to be harmful to both the rich and the poor; and that society would be made worse off if people were persuaded by it.Footnote 31 He rejects silencing it nevertheless, because even harmful opinions should be tolerated (and criticized).

The distinction between ordinary expressions of opinion and performative utterances such as incitement to riot clearly underlies his argument.Footnote 32 Freedom of speech concerns the profession and discussion of opinion, and it requires the tolerance of all doctrines; but it does not apply to actions simply because they are performed with words. No free speech absolutist defends the freedom of speech acts. It would be absurd to claim that the best moral rules immunize all speech acts, including fraud and conspiracy—let alone the clichéd example of shouting “Fire!” in a theater (maliciously, in order to cause panic). The corn dealer example is designed to set aside such specious exceptions, which do not qualify the actual doctrine.

The distinction between incitement and persuasion is clearer in standard cases, where the speaker’s intentions are not obscured. Mill deliberately poses the most philosophically challenging case in his discussion of the corn dealer. It is challenging because incitement can have the same consequences as persuasion, and the specific case of incitement he imagines is performed by assertion. Incitement is usually not so subtle but more like, “Kill the corn dealer!” This case shows that some grammatical assertions are nevertheless performative utterances—and therefore beyond the pale of free speech immunity—just as some grammatical imperatives are best understood as expressions of opinion, such as “Workers of the World, Unite!” when written in a manifesto.Footnote 33 The point is not that there are no hard cases, which might even require the courts to adjudicate, but that common sense often suffices to draw the crucial distinction even when the grammatical form of the utterance is misleading. Hence, the crucial distinction for any tenable doctrine of freedom of speech, which puts certain actions performed by speaking (or writing) outside the purview of the sphere of liberty, is not between harmless and harmful opinions but between the expression of opinion and other illocutionary actions.Footnote 34

Mill places the expression of opinion into the self-regarding category, protected by right, except in specific contexts where it constitutes a wrongful performative speech act. The expression of opinion is thus treated analogously with other typically self-regarding action. The corn dealer’s rights are violated by incitement to murder but not by the advocacy of bad economic doctrine or pernicious ideology, even when it has similar results.Footnote 35 Someone who is convinced by Proudhon, and decides to liberate your pocketbook, can be punished for the crime; but Proudhon cannot be punished for promulgating the opinion, even though his intention is to persuade. He should be criticized for professing pernicious opinions but not compelled to have or express better ones. Hence, Mill’s defense of freedom of speech is absolute in exactly the relevant sense: it is a defense of the profession and discussion of all opinions, regardless of their value. Whatever its substantive merits as a matter of policy, the qualified interpretation favored by progressives founders as an account of Mill’s classically liberal position.

IV. The Practical Inseparability of the Liberties of Conscience

Although Mill did not countenance any qualifications on free speech based on the value of the opinion being expressed, it can still be claimed that a rule utilitarian argument supports more limited speech rights than Mill thought. It might justify an exception for hate speech, for instance. This claim cannot be made compelling as an interpretation of Mill, since he excludes as reasons to censor an opinion precisely the harms most plausibly attributed to its expression. Nevertheless, progressives can fairly complain that, without further argument, such exclusion begs the question against their objections to freedom of speech. These focus either on the emotional harm suffered by vulnerable groups confronted by hate speech; the physical manifestation of this psychological trauma, which is claimed to render such speech akin to (or literally) violence; or the pernicious consequences of opinions whose toleration supposedly compromises the social equality of marginalized groups.Footnote 36 Why then doesn’t the best set of moral rules enshrine more-limited speech rights, which permit intolerance of hateful and otherwise harmful opinions?

Since the aims of this essay are largely exegetical, I cannot aspire to defend the liberal view adequately against these challenges. Nevertheless, I contend that Mill’s argument has underappreciated resources that are directly relevant to the current debate over free speech. His argument for the practical inseparability of the liberties of conscience anticipates psychological and philosophical insights that support the liberal position over its progressive alternatives. Because I cannot demonstrate the accuracy of the picture of moral psychology that underwrites Mill’s argument, I will instead sketch what would have to be the case about humanity in order for his conclusion to be best supported by rule utilitarian considerations about the conditions under which mental well-being best flourishes.

The rights of individuality center around sovereignty over one’s own body and mind. Yet there is a crucial respect in which our minds are not our own, despite the fact that the liberty of thought precludes our being compelled to believe (or even avow) any doctrine. The liberties of conscience are inseparable because justification, and hence knowledge, is social. Mill held that the justification of our beliefs—especially about creedal matters, which most arouse emotion and motivate intolerance—require social conditions that cultivate the best opposing arguments. This is true even when our beliefs rely primarily on testimony, because expert testimony is subject to the same epistemological hazards, and its reliability therefore depends upon the existence of institutions that tolerate rather than silence dissent.Footnote 37 Hence, we lack knowledge unless we (or the experts on whom we rely) can refute the strongest arguments against what we believe. This is the crucial respect in which knowledge is social: it requires not merely the toleration of dissenting opinions but the cultivation of diversity of opinion. The second chapter of On Liberty is dedicated to epistemological arguments for freedom of speech, the details of which are often criticized in picayune ways that overlook how the crux of his argument has been vindicated by subsequent work in social epistemology and psychology.Footnote 38

Mill’s claim that the liberties of conscience are inseparable in practice amounts to the thesis that many reasons freedom of thought is a prerequisite for mental well-being also apply to freedom of speech. That is not to say that it is impossible for society to tolerate thought but not expression, or that there are no additional reasons that it decreases well-being to compel belief (or its avowal through oaths). The claim is rather that in order for the liberty of thought to benefit human flourishing in the most important respects—such as to advance knowledge, provide justification for our opinions, and produce active conviction rather than “dead dogma”—our beliefs must be subject to the sort of scrutiny that is impossible when the profession of unpopular opinion is legally or socially prohibited. We might call these epistemological aims the advancement of knowledge in the broadest sense.

In this respect, Mill anticipates what Allen Buchanan characterizes as the “central message of social epistemology”: the insight that knowing is “largely a collective enterprise.”Footnote 39 This insight helps explain why public expressions of opinion ordinarily fall into the sphere of liberty protected in principle from coercive interference. Most social epistemology focuses on beliefs based on the testimony of others, especially experts and witnesses, whereas Mill frames his argument in more traditional internalist terms. Nevertheless, his argument coheres with the central insight for two reasons. First, the reliability of the experts on whom we rely requires social conditions that do not suppress opinions that dissent from prevailing orthodoxy. Second, creedal opinions differ from empirical beliefs in a crucial respect. Even if we grant the dubious notion of moral expertise, creedal beliefs accepted on authority are especially prone to becoming dead dogma rather than conviction capable of sustaining and motivating our central projects. If the advancement of knowledge requires social conditions that can only be provided through freedom of speech, then the liberty to profess and discuss any opinion is, in practice, inseparable from the liberty of thought.

Although Millian liberalism advocates the toleration of all opinions, it is not neutral about the moral ecology. It favors conditions of intellectual diversity that many consider at most second best, inferior to the silencing of viewpoints we despise. Nevertheless, the attempt to shield people from offense and error is misguided, in his view, because “only through diversity of opinion is there, in the existing state of human intellect, a chance of fair play to all sides of the truth” (Liberty, 254).Footnote 40 What would have to be the case about human nature for this claim to be true? In short, we would have to be profoundly conformist and prone to forms of cognitive bias that make our beliefs unjustified unless the social conditions are such as to protect unpopular opinion. This hypothesis is not so unlikely.Footnote 41

Mill held that even educated people suffer from what is known as (informational and normative) conformity and confirmation bias: the tendencies to adopt the views held by one’s peers, and to be more sensitive to evidence that confirms rather than refutes one’s antecedent beliefs. Indeed, people even resist exposure to contrary opinions, especially on creedal matters of religion, morality, and politics. This phenomenon is referred to as epistemic closure. Moreover, when we succeed in surrounding ourselves with like-minded people and sources of information, we are subject to group polarization: the tendency of a group of the like-minded to become more radical in their views. The result of these mutually exacerbating cognitive biases is that, as Mill observed, “the weakest part of what everybody says in defence of his opinion, is what he intends as a reply to antagonists” (Liberty, 251). His great insight here is that it is not just intellectual dishonesty that causes people to argue against straw men, but ignorance of the best opposing arguments—and that the more intellectual conformity exists, the more this condition obtains.

When we are not acquainted with the strongest opposing arguments—which is exactly the condition produced by intellectual conformity—then we lack a familiar and important form of justification. Mill thought that even educated people suffer from this condition, due to their lack of exposure to what he termed adverse discussion: confrontation with opposing arguments. Yet people find adverse discussion unpleasant and often choose to avoid it. In this condition,

Their conclusion may be true, but it might be false for anything they know: they have never thrown themselves into the mental position of those who think differently from them, and considered what such persons may have to say; and consequently they do not, in any proper sense of the word, know the doctrine which they themselves profess. (Liberty, 245)

The more averse we are to considering opposing viewpoints, the more we will be prone to epistemic closure and confirmation bias. And when we form and maintain these beliefs on the basis of nonrational factors, such as identification with an in-group whose dogmas are adopted uncritically, these complementary forms of cognitive bias undermine the justification of our beliefs. Moreover, insofar as we are prone to group polarization, the lack of diversity of opinion will lead to unjustified extremism. This condition is most coercive when it issues in what Mill calls intemperate discussion, which stigmatizes those who hold contrary opinions as bad and immoral—as people who should not be refuted but, as the current phrase has it, “cancelled.”

I will leave it to the reader to consider the extent to which these phenomena, all of which are clearly anticipated by Mill, have been borne out by recent developments. But I will add an additional point about the consequences of attempts to ban specific opinions on the grounds of their harmfulness. To ban a category of opinion classed as hate speech, white supremacy, or anything to which the tendentious suffix “-phobia” is attached, creates an incentive for those with the power to silence speech to conflate creedal opinions they despise with whatever category of speech can be silenced. This is not an abstract or hypothetical point but an observation of what is already happening. The defense of freedom of speech itself is increasingly claimed to be a cover for white supremacy, an idea that has passed from critical race theory into popular culture.Footnote 42 Moreover, the creation of classes of opinion whose expression gets deemed violence creates an incentive to be traumatized—to claim that exposure to offensive opinions triggers emotional harm such as, at the extreme, post-traumatic stress syndrome. This incentive applies even to those who are sincere, not simply exercising their power to silence opinions they abhor. These tendencies are mutually exacerbating, since the aversion to give a fair hearing to opposing views leads people to construct less reasonable, more abhorrent antagonists.

Mill’s convictions about conformism, cognitive bias, and the necessity of cultivating diversity of opinion render the indirect utilitarian argument supporting his defense of freedom of speech (and the rights of individuality in general) considerably more plausible. Whether or not one is convinced by Mill’s argument, this constitutes decisive evidence in favor of my interpretation, which takes his defense of freedom of expression to be as absolute as he expressly and repeatedly claims. The right to free speech can be seen as grounded in utility in the broadest sense, given Mill’s repeated claims that the development by an individual of her own mind is an essential component of her well-being, and that the cultivation of diversity of opinion is a prerequisite for knowledge and the advancement of society. Although Mill grants that the profession and discussion of opinions can do harm, nevertheless freedom of expression is central to the sphere of liberty protected from coercive intervention. This conviction reflects not the weakness of speech but its paramount importance to human flourishing.

Footnotes

Funding for this project was provided by the Institute for Humane Studies, the Murphy Institute at Tulane University, the Liberty Fund, and a University of Michigan Humanities Grant. I am also grateful for astute comments from several audiences, as well as the editors and an anonymous referee for this journal.

References

1 This applies to what Mill calls the free society, not to every stage of civilization. And it specifically concerns the expression of opinion, not all speech acts. Both of these points will be discussed in more detail presently.

2 All references to Mill are to the Collected Works of John Stuart Mill, ed. J. M. Robson (Toronto: University of Toronto Press, 1981) and given initially as (title, volume, page) and subsequently as (short title, page).

3 While I cannot hope to give an exhaustive catalogue, most scholars who address Mill’s political philosophy adopt (at least) one of the two readings I oppose, or else conclude that Mill was irremediably inconsistent. These two interpretations are not mutually exclusive, and they seem inevitable when one reads the principle of liberty as a harm principle. Some important representatives of what I’m calling the pragmatic interpretation include Crisp, Roger, Mill on Utilitarianism (London: Routledge, 1997)Google Scholar; Grey, John, Mill on Liberty: A Defence, 2d ed. (London: Routledge, 1996)Google Scholar; Skorupski, John, John Stuart Mill (London: Routledge, 1989)Google Scholar, 373 et seq.; Turner, Piers Norris, “‘Harm’ and Mill’s Harm Principle,Ethics 12 (2014): 299326 CrossRefGoogle Scholar; and Riley, Jonathan, “J. S. Mill’s Doctrine of Freedom of Expression,Utilitas 17 (2005): 147–79CrossRefGoogle Scholar. Advocates of the qualified interpretation include Fred Berger, Happiness, Justice and Freedom: The Moral and Political Philosophy of John Stuart Mill (Berkeley: University of California Press, 1984); and Dyzenhaus, David, “John Stuart Mill and the Harm of Pornography,Ethics 102 (1992): 534–51CrossRefGoogle Scholar; and it receives its most compelling formulation in David Brink, Mill’s Progressive Principles (Oxford: Oxford University Press, 2013).

4 The fundamental misunderstanding of the principle of liberty as a harm principle is simply untenable, as both the weak and strong versions attributed to Mill cannot do justice to his view defended in On Liberty. See Daniel Jacobson, “Mill Does Not Have a Harm Principle” (MS).

5 Mill is sometimes taken to hold what might be called the sticks-and-stones theory of harm (after the nursery rhyme that ends “but words will never harm me”), which diminishes the importance of harm as a moral concept, since everyone prefers minor physical damage to severe emotional distress. But there is conclusive evidence that he does not hold this simplistic view. When he considers the most vital moral rules, “which forbid mankind to hurt one another,” he continues: “in which we must never forget to include wrongful interference with each other’s freedom” (Utilitarianism, X, 255).

6 Thus Riley argues that Mill’s defense of free speech does not rest on the principle of liberty: “Even though expression is other-regarding conduct and thus legitimately subject to social regulation, however, Mill consistently defends a broad policy of laissez-faire with respect to expression, analogous to the policy of laissez-faire which he defends with respect to economic exchange of trade.” Riley, “Mill’s Doctrine,” 149.

7 In Mill’s view, “Justice implies something which it is not only right to do, and wrong not to do, but which some individual person can claim from us as his moral right” (Utilitarianism, 247). While this account of justice is controversial, the important point for our purposes is simply the rights claim, or lack thereof.

8 Brink thus claims Mill “is clearly not a free speech absolutist, as his application of the harm principle to the corn dealer case illustrates.” Brink, Mill’s Principles, 171. By Mill’s “harm principle,” he means the principle of liberty. See note 4.

9 In regard to hate speech regulation, Brink claims: “Insofar as such regulations [on speech] target genuine harm, and not mere offense, they would be defensible according to Millian principles” (Brink, Mill’s Principles, 164).

10 Compare the procedural conception of justice in Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974) and F. A. Hayek, Law, Legislation, and Liberty: A New Statement of the Liberal Principles of Justice and Political Economy (Abingdon: Routledge and Kegan Paul, 1982) to the outcome-based conception of justice in Rawls and egalitarianism generally. See Scheffler, Samuel, “Reactive Attitudes and Liberalism,” reprinted in Boundaries and Allegiances: Problems of Justice and Responsibility in Liberal Thought (Oxford: Oxford University Press, 2001)Google Scholar for a similar claim, although he draws the distinction as between Rawlsian and utilitarian liberals (who are progressives in my terminology), on one hand, and “the older tradition of Lockean liberalism” (my classical liberals) on the other. But Scheffler’s terminology misconstrues Mill’s view, the strongest form of classical liberalism, by implying that it is incompatible with a utilitarian foundation. Part of my aim here is to demonstrate the falsity of this assumption.

11 See Jeremy Waldron, The Harm in Hate Speech (Cambridge, MA: Harvard University Press, 2012); Wayne Sumner, The Hateful and the Obscene (Toronto: University of Toronto Press, 2004); and David Brink, “Millian Principles, Freedom of Expression, and Hate Speech,” Legal Theory 7 (2001): 119–57 for philosophical defenses of regulations on moral and political speech. The peculiarity of American absolutism about speech is a central theme in the work of Frederick Schauer; see e.g., Frederick Schauer, “The Exceptional First Amendment,” in American Exceptionalism and Human Rights, ed. Michael Ignatieff (Princeton, NJ: Princeton University Press, 2005). Owen Fiss writes that the state, “may even have to silence the voices of some in order to hear the voices of the others.” Owen Fiss, The Irony of Free Speech (Cambridge, MA: Harvard University Press, 1996), 4. Robert Post celebrates the fact that, “Liberated from traditional inhibitions against official suppression of speech, the left has mobilized to pursue a rich variety of political agendas.” Robert Post, ed., Censorship and Silencing (Los Angeles, CA: The Getty Research Institute, 1998), 2. These examples are drawn from renowned First Amendment scholars, not the more radical advocates of critical legal and racial studies.

12 See <http://www.politifact.com/truth-o-meter/statements/2017/apr/21/howard-dean/howard-deans-wrong-tweet-constitution-doesnt-prote/>. A 2017 Brookings Institute survey found that 44 percent of U.S. college students agree with this claim, though it is false as a matter of constitutional law. Moreover, the survey found that a majority of students, including 62 percent of Democrats, favor shouting down campus speakers they deem offensive. <https://www.brookings.edu/blog/fixgov/2017/09/18/views-among-college-students-regarding-the-first-amendment-results-from-a-new-survey/>.

13 For example, former United Nations Secretary-General Ban Ki-moon stated that, as much as he supports free speech: “When some people use this freedom of speech to provoke or humiliate some others’ values and beliefs, then this cannot be tolerated.” Secretary-General Ban Ki-moon, Press Conference at United Nations Headquarters, New York, September 19, 2012 <http://www.un.org/press/en/2012/sgsm14518.doc.htm>.

14 This claim also requires argument, since one might think that capitalist acts between consenting adults (in Robert Nozick’s phrase) should be part of the sphere of liberty. That is the libertarian view. But Mill’s classical liberalism deviates from libertarianism in this fundamental respect, as well as on taxation, notwithstanding its more pronounced differences with progressivism on these and other issues.

15 When it comes to interference with trade intended to make some commodity difficult or impossible to obtain, Mill writes: “These interferences are objectionable, not as infringements on the liberty of the producer or seller, but on that of the buyer” (Liberty, 293).

16 This quotation refers specifically to governmental coercion because it is taken from the “Limits of Laissez-Faire” section of the Principles, but Mill’s argument in On Liberty extends to social coercion as well.

17 One might try salvaging a non-normative gloss of the self-regarding using the distinction between direct and indirect effects of action, but that will not quite do. Mill uses this distinction to support the normative claim that the interposition of another person’s agency mitigates the harmful effects of an action. But he does not hold that commercial acts count as self-regarding when they affect non-consenting others only indirectly. Most important, the same problem arises for Mill in any case: he must justify the (superficially anti-utilitarian) claim that some class of harms do not count as reasons to interfere with individual liberty—despite being foreseeable or even intended.

18 Mill actually ends this sentence with the phrase: “moral disapprobation in the proper sense of the term.” But it is clear from the passage as a whole that he is talking about what can properly be punished, consistent with his view that disapprobation can be a form of punishment (Utilitarianism, 246).

19 Mill adds that society may demand “each person’s bearing his share … of the labours and sacrifices incurred for defending the society or its members from injury and molestation” (Liberty, 276). These too are obligations, though they do not correspond to anyone’s rights.

20 Riley for instance seems to think that the following passage unambiguously places the expression of opinion in the social class, and that no additional argument for this claim is needed (Riley, “Mill’s Doctrine,” 147 et seq.).

21 Mill criticized Bentham for the opposite fault: “He could not bear, for the sake of clearness and the reader’s ease, to say, as ordinary men are content to do, a little more than the truth in one sentence, and correct it in the next. The whole of the qualifying remarks which he intended to make, he insisted upon imbedding as parentheses in the very middle of the sentence itself” (“Bentham,” X, 114). The question for Mill is not whether an individual’s expression of opinion concerns or affects non-consenting others, but whether it primarily, or in the first place, concerns them. Mill clearly intends to make those qualifying remarks, though he does not always embed them in a given sentence.

22 Although the harms of fair competition are not among those against which we have a right to protection, this does not mean that we have a right to free trade or economic freedom in Mill’s view. This hangs on whether a rule utilitarian argument putting economic liberty on a par with political liberty would be most conductive to well-being. One can deny that while still holding that price controls are counterproductive, for instance, because they create shortages that most harmfully affect those who the controls were meant to help. Thanks to an anonymous referee for pressing me to explain this point further.

23 “The spirit of commerce and industry is one of the greatest instruments not only of civilization in the narrowest, but of improvement and culture in the widest sense” (“De Tocqueville on Democracy in America” [2], XVIII, 197.)

24 There are various forms of rule utilitarianism, and forms of indirect utilitarianism other than rule utilitarianism, but it would take us too far afield to consider which best resembles Mill’s view. The crucial point is that his commitment to rights and obligations arising from moral rules is genuine, and is inconsistent with any form of act utilitarianism, however sophisticated.

25 In fact, the decisive evidence that Mill held an indirect form of utilitarianism rather than act utilitarianism is precisely that he respects moral rights throughout his work. This point is buttressed by the fact that he insists on supererogatory action and that moral rules ground genuine obligations. No close reading of one passage—especially not the “proportionality statement” of the Greatest Happiness Principle can override these considerations. See Daniel Jacobson, “J. S. Mill and the Diversity of Utilitarianism,” Philosophers’ Imprint 3 (2003) <http://www.philosophersimprint.org/003002/>.

26 There are passages in Chapter 2 of Utilitarianism that are misleading on these points (though Chapter 5 vindicates them). These passages must be understood by appreciating Mill’s purpose in that chapter, which is to defend the common creed of utilitarianism, not to specify his own version of it. See Jacobson, “Diversity of Utilitarianism.”

27 Although Mill famously declares that utility is “the ultimate appeal on all ethical questions,” including the justification of rights, he continues by saying that this must be understood as “utility in the largest sense, grounded on the permanent interests of man as a progressive being” (Liberty, 224): that is, one capable of self-improvement.

28 See D. G. Brown, “What is Mill’s Principle of Utility?” Canadian Journal of Philosophy 3 (1973): 1–12. This axiological claim is what Mill defends as the Principle of Utility in Chapter 4 of Utilitarianism, not a moral principle about right and wrong (such as the Greatest Happiness Principle).

29 There are of course other senses of “absolute,” including the absurd doctrine that any act of speech is immune to coercive interference—a doctrine that no defender of free speech has ever advocated. As we will see, Mill’s argument does not conflict with restrictions that are based on the circumstances of the speech (such as in a monastery or library) rather than the content of the opinion. And it does not extend to all speech acts but focuses on the profession and discussion of opinion, “as a matter of ethical opinion.” There are legitimate challenges concerning how to apply this doctrine to such issues as libel and non-propositional artistic expression, which are not what is at issue between the various interpretations and cannot be broached here.

30 This is close but not equivalent to what Mill says about the doctrine of tyrannicide. Its advocacy should be tolerated except in contexts where it constitutes an instigation to a specific murder, and even then “only if an overt act has followed, and at least a probable connexion can be established between the act and the instigation” (Liberty, 228fn).

31 “We suppose the majority [of the poor] sufficiently intelligent to be aware that it is not to their advantage to weaken the security of property” (Representative Government, 442).

32 Although the language of speech acts postdates Mill, he obviously draws an intuitive distinction between the profession of opinion and other things one can do with words. Though J. L. Austin would later claim that the distinction between constantive and performative speech acts is insufficiently fine-grained, he did not reject the idea of performative utterance but refined it with his theory of speech acts and illocutionary action. Nevertheless, the intuitive distinction—which after all belongs to ordinary language—must be captured by the more complex theory. Whatever one calls an expressive act, something like profession must count as a type of illocutionary action. It passes the classic test: “I hereby profess P, thereby professing P.” See Austin, J. L., How to Do Things With Words (Oxford: Clarendon Press, 1962)Google Scholar.

33 Some genuine imperatives are self-regarding actions on other grounds than qua expressions of opinion.

34 As common sense suffices to identify most performative utterances, it can also detect action intended to silence specific opinions, such as bans on pamphleting outside military recruitment offices or abortion clinics. This is not to deny the existence of hard cases but to insist on the prevalence of easy ones.

35 Mill’s implicit theory of incitement by assertion, illustrated by the corn dealer case, is best understood as trading on the fact that the agency of another person intervenes in the case of persuasion. It can perhaps be justified by an insight about mob psychology and the fragility of social norms that are commonly violated. In this view, mobs function not as a collection of agents but more like a natural or animal force. Although I think Mill gets the corn dealer example right, he could instead immunize the assertion but not the imperative. The main point is that an assertion might, because it is persuasive, have the same perlocutionary effect as incitement without losing immunity.

36 The first of these arguments is familiar; for the latter two see Barrett, L. F., “When is Speech Violence?New York Times (July 14, 2017)Google Scholar and Waldron, Harm in Hate Speech, respectively.

37 Mill’s arguments are made in terms of epistemological internalism: he is concerned with the evidence we have for our beliefs. Although I will discuss the issues in those terms, I believe that they can be translated into externalist terms about the reliability of the sources on which our beliefs rely, such as expert testimony.

38 For an overview, see Sunstein, Cass, Conformity: The Power of Social Influences (New York: New York University Press, 2019)Google Scholar. For some seminal work on these subjects see: Kelman, H. C., “Compliance, Identification, and Internalization: Three Processes of Attitude Change,Journal of Conflict Resolution 2 (1958): 5160 CrossRefGoogle Scholar, on conformity; Myers, D. G. and Lamm, H., “The Group Polarization Phenomenon,Psychological Bulletin 83 (1976): 602627 CrossRefGoogle Scholar, on group polarization. Peter Wason, “On the Failure to Eliminate Hypotheses in a Conceptual Task,” Quarterly Journal of Experimental Psychology 12 (1860): 129–140, on confirmation bias.

39 Allen Buchanan, “Political Liberalism and Social Epistemology,” Philosophy and Public Affairs 32, no. 2 (2004): 98. As I go on to note, this is an admittedly different use of social epistemology than most social epistemologists have had in mind, and Buchanan may not agree with my application of it to Mill’s argument.

40 Note that the claim here is not that freedom of speech is sufficient for the advancement of knowledge but that it is a necessary condition. There will be other necessary conditions, including a ban on forms of silencing that are less obviously intolerant than censorship and which might be neutral on their face, though not in effect. I am grateful to an anonymous referee on this point.

41 It would be an extremely worthwhile project in moral epistemology to catalog the various forms of cognitive bias and intellectual prejudice that support Mill’s hypothesis. But that would be the work of another paper. Here I aspire only to point in the direction of this evidence, not to consider it adequately.

42 Classical liberalism is claimed to be a stalking horse for racism: “the favorite term of alt-righters who mistakenly think they’re cleverly fooling you.” This was popular feminist Amanda Marcotte’s comment on New York Times editorial writer Bari Weiss (ironically a Jewish lesbian, not a favored group among the alt-right). See Jon Levine <https://www.yahoo.com/entertainment/york-times-editor-savaged-fascists-column-bari-weiss-135843444.html>.