DANIEL JACOBSON
There are three rival views of Mill’s defense of speech rights, which we might call the absolutist, the pragmatic, and the qualified interpretations. According to the absolutist interpretation, Mill defended an unqualified doctrine of freedom of speech, based on one of his fundamental principles: the principle of liberty. In this view, any interference with an agent’s freedom of speech – as that notion is properly understood – counts as a violation of his liberty rights. This reading stresses what Mill expressly and repeatedly says about what he calls the free society, which has developed sufficiently to sustain individual rights and representative government. Once society has progressed sufficiently to reached the level of development where the principle of liberty should be respected: “there ought to exist the fullest liberty of professing and discussing, as a matter of ethical conviction, any doctrine” (Liberty, XVIII: 228n).
According to the pragmatic interpretation too, Mill defends freedom of speech as claimed, but he does so in an unprincipled way. His defense does not rest on the principle of liberty, and interference with an agent’s freedom of speech is no violation of her rights. Rather, Mill defends free speech as he defends free trade: not as a matter of rights, let alone because it follows from the principle of liberty, but simply on pragmatic grounds.1 In this view, speech falls under the jurisdiction of society, in principle, but it turns out that restrictions on the expression of opinion and sentiment are never optimal. This is indeed how Mill defends the doctrine of free trade, which he grants “rests on grounds different from, though equally solid with, the principle of individual liberty” (Liberty, XVIII: 293). His objections to coercive interference with trade are practical rather than principled. They rest on the inefficacy of restraints on trade, which are wrong because they do not produce their intended effects and have bad unintended consequences as well.
According to the qualified interpretation, by contrast, Mill does not in fact defend the freedom to express any doctrine, his claims to the contrary notwithstanding. Instead he draws a distinction between harmful and merely offensive speech; or between especially harmful speech, which diminishes autonomy or some other higher good, and speech that harms only in less significant ways.2 In short, Mill does not defend freedom of speech at all, but something more like the freedom of harmless speech. Since this interpretation must set aside as exaggeration so much of what Mill says about free speech, it rests heavily on a putative exception he makes to the doctrine: the corn dealer example. Mill notes that the opinion that corn dealers are starvers of the poor must be tolerated – that doctrine too must be allowed to be professed and discussed – but its expression can justly be punished specifically when it is advocated to an angry crowd gathered outside a corn dealer’s house.3 The qualified interpretation understands the corn dealer example as just one of a wide variety of cases where the (significant) harmfulness of an opinion justifies its prohibition and the punishment of its expression, in Mill’s view.
In this chapter, I will argue that the absolutist interpretation is the only tenable reading of Mill’s position on freedom of speech. The pragmatic interpretation is misguided to claim that the expression of opinion falls beyond the pale of the principle of liberty; and hence that it is subject, in principle, to restriction in the free society.4 It comes to this view through a faulty reading of the principle of liberty as a harm principle, on which there is reason – though not always sufficient reason – to interfere with all and only actions harmful to non‐consenting others.5 And the qualified interpretation misunderstands the significance of the corn dealer example, which is not intended as an example of an opinion that can properly be punished for its harmfulness. Rather, it is meant to differentiate between the “profession and discussion” of opinions and sentiments, and the more performative things people do with words, which are not universally immunized by Mill or included in any tenable conception of free speech. The freedom of expression does not include the freedom to threaten, conspire, or even suborn (and hence commit) murder – all of which can be done by speaking.
Although this discussion will focus on freedom of speech, it has further implications. It bears on how Mill’s principle of liberty should be understood; on his distinction between self‐regarding and social actions; and, perhaps most significantly as a matter of Mill interpretation, on how his principle of liberty can be reconciled with his unorthodox but genuine commitment to utilitarianism. Moreover, the issue of freedom of speech is especially significant in the current political context, because the justification for the absolutist interpretation over its rivals hangs on distinctions crucial to the difference between classical liberalism and progressivism, which is often obscured by vagueness in the use of the term liberal. Mill’s classically liberal commitments support an unqualified defense of speech rights against the skepticism of modern progressives. This chapter does not attempt to argue for the empirical and evaluative commitments of classical liberalism, merely to show how fundamental they are to Mill’s moral and political philosophy, contrary to some modern reinterpretations of him as a proto‐progressive.
The central claim of the pragmatic interpretation is that Mill does not believe in a right to freedom of speech, despite defending the doctrine on pragmatic grounds. This reading must hold, in particular, that the principle of liberty does not entail any such right. The model for this argument comes from Mill’s defense of free trade – understood as the doctrine that sellers should be allowed to set prices as they choose, and buyers free to purchase goods elsewhere – which he expressly grants not to rest on the principle of liberty. His objection to price controls, tariffs, and the like are practical: the quality and price of commodities are best ensured by the state forgoing such regulation of trade. It makes no difference for present purposes whether Mill was correct on this matter of economic policy. The crucial point is that Mill holds that trade properly falls under the jurisdiction of society.6 That is why he claims that it would be no violation of the liberty rights of the seller to ban the sale of any particular good, though it would be a violation of the liberty of the buyer (Liberty, XVIII: 293). The prohibition of alcohol is a violation of the rights of would‐be buyers, for instance, because it contradicts the slogan with which Mill sums up the principle of liberty: “Over himself, over his own body and mind, the individual is sovereign” (Liberty, XVIII: 224).
Since the pragmatic interpretation agrees with the absolutist interpretation about Mill’s conclusion – that he ultimately defends freedom of speech – it is not belied by his repeated claims that the liberty to express one’s opinion and sentiment should be exceptionless. But it has a great deal of trouble making sense of the peremptoriness of Mill’s arguments for this conclusion. To take just the most obvious example, Mill writes:
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, XVIII: 229)
It is not trivial to square this argument with Mill’s commitment to utilitarianism, since that theory seems committed to taking the numbers so seriously, but there are some indirect forms of the theory that can accommodate it – given certain empirical assumptions. Consider rule utilitarianism, on which an action is wrong if it violates the best set of moral rules, where those are determined by the utility of their acceptance, including the costs of non‐compliance. If a rule mandating freedom of speech is one of the best moral rules, then it would follow that it is wrong to silence even the lone dissenter. But this will be the case only if alternative approaches have worse outcomes, in particular rules that make exceptions for putatively low‐value speech, as progressives are wont to do with “hate speech” and skepticism about climate change, both of which have been recently targeted for censorship by prestigious transnational institutions (such as the European Union and the United Nations).7
Hence the peremptory nature of Mill’s argument challenges any merely pragmatic defense of free speech, which must remain sensitive to alternative policies and changing circumstances, since the doctrine is a matter of policy rather than principle. Moreover, Mill’s discussion of the topic is couched in claims about rights. In the passage previously quoted, he adds: “I deny the right of the people to exercise such coercion [over speech], either by themselves or by their government” (Liberty, XVIII: 229). This contrasts sharply with his treatment of free trade; and it is exactly what he should not say, according to the pragmatic interpretation. Although it would be a mistake to restrain speech, in this view, it falls within the jurisdiction of society to do so.
Supporters of the pragmatic interpretation rely primarily on the fact that speech (i.e., the expression of opinion and sentiment) can harm others without their consent. If one reads the principle of liberty as a harm principle, which protects only action harmless to non‐consenting others, then this puts speech outside the sphere of liberty protected by that principle. In Mill’s terms, the potential harmfulness of speech renders it a social rather than a self‐regarding act. These are terms of art for Mill, since an action is not social simply in virtue of involving other people; acts that involve others consensually count as self‐regarding. Thus Mill’s claim that trade is a social act simply recapitulates his claim that it is not protected by the principle of liberty. Were the principle of liberty a harm principle, then the only self‐regarding acts would be (roughly) those that are harmless to non‐consenting others. This is rough because some level of danger must be counted as harm even if it does not issue in injury, in order for any harm principle to be tenable; otherwise lucky drunk driving would not be punishable. Hence the fate of the pragmatic interpretation rests on whether the principle of liberty is a harm principle, and self‐regarding action – those actions that the principle makes immune to interference – is tantamount to such “harmless” action.8
It must be admitted that this way of understanding the principle of liberty is the most natural one, if the two sentences that seem to give its official statement are taken by themselves, out of context. They read:
That principle is, that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self‐protection. That the only purpose for which a power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others.
(Liberty, XVIII: 223)
Mill then makes explicit that this rules out paternalism – the coercion of people for their own good – when it comes to sound‐minded adults, and he sums up the doctrine with the slogan already quoted about individual sovereignty over mind and body.
This passage poses several obvious exegetical problems. The two consecutive statements (of what Mill calls “one very simple principle”) do not seem synonymous, since the concept of an exercise of power is more expansive than just interference with liberty of action. Indeed, Mill goes on to argue that certain positive compulsions are legitimate, such as taxation, even though he never claims that taxation can only be used to prevent harm, as opposed to promoting the good.9 Even if taxation does not interfere with the liberty of the taxpayer, it certainly counts as an exercise of power over him. But I propose to ignore the problem about positive compulsions and set aside worries about how to understand harm. These are nuances as compared to the most significant point, which has been extremely influential since noted by contemporary utilitarian critics of Mill such as James Fitzjames Stephen. In short, the objection is that (almost) everything we do potentially affects others, even harmfully, so no significant sphere of liberty is secured by Mill’s principle – and certainly not freedom of speech.
If we do not try to understand the seemingly official statement of the principle of liberty in isolation, however, it will become clear that Mill cannot intend it to be a harm principle. If the principle of liberty were a harm principle, then it would protect freedom of speech only if the expression of opinion and sentiment could somehow be guaranteed to be harmless. But there is no evidence that Mill adopted this naive theory of harm – which we might call the stick‐and‐stones theory, after the nursery rhyme that concludes “but words will never harm me” – and conclusive evidence that he does not. When Mill identifies individual rights as having a “more binding obligation” than do ordinary maxims because they are more vital to human well‐being, he expressly refers to those “moral rules which forbid mankind to hurt one another (in which we must never forget to include wrongful interference with each other’s freedom)” (Utilitarianism, X: 255). Thus Mill rejects the sticks‐and‐stones theory by taking wrongful violations of liberty to count as harms; moreover, such violations count as harms against which one has the right to protection in the free society.
Since Mill rejects this untenable theory of harm, any interpretation of the principle of liberty that reads it as a harm principle, which immunizes only those actions harmless to non‐consenting others, seems forced to conclude that speech lies beyond the pale of the principle of liberty. This is precisely the argument for the pragmatic interpretation. If the principle of liberty applies only to harmless actions, and the expression of opinion and sentiment can harm non‐consenting others, then that principle cannot immunize speech. Mill’s defense of free speech therefore must be based on pragmatic grounds, like his defense of free trade. A first sign of trouble for this reading, however, is that Mill claims otherwise: that his defense of freedom of speech is part of the general defense of liberty. After having introduced the principle of liberty in Chapter 1 of On Liberty, he begins its defense in Chapter 2, which is titled “Of the Liberty of Thought and Discussion.” In explaining his argumentative strategy, Mill writes:
It will be convenient for the argument, if, instead of at once entering upon the general thesis [that is, the principle 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, XVIII: 227)
Those defenders of the pragmatic interpretation who notice this point puzzle over why Mill would say that the liberty of speaking and writing is impossible to separate from the liberty of thought. After all, we are not always incapable of thinking something without saying it. But the first question is not whether Mill is correct in his claim that speech is, in practice, impossible to separate from thought; it is whether this is Mill’s view. The fact that he discusses thought and discussion together in Chapter 2 of On Liberty, specifically as constituting a single element of the sphere of liberty, makes this hard to deny. A clue to the puzzle can be found in the epistemological arguments of that chapter, where he claims that justification is fundamentally social. Mill believes that exposure to opposing points of view is a prerequisite for knowledge, because in order to be justified in one’s beliefs, one must know the best opposing arguments, which can only be adequately propounded by their advocates. So although we can sometimes keep our thoughts to ourselves, we cannot have knowledge by ourselves.
But how can Mill think that freedom of speech follows from the principle of liberty without holding that the expression of opinion is harmless? The answer is that the principle of liberty is not a harm principle, and self‐regarding action – that is, action protected by the principle of liberty as a matter of rights – is not equivalent to harmless action. In the passage where Mill discusses what he means by self‐regarding action, he explains:
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].10
(Liberty, XVIII: 281).
The crucial point is that what makes a typically self‐regarding action lose that status is not that it harms others, or even that it harms their interests, but that it violates an obligation. The example Mill uses to illustrate the point is drunkenness, which is ordinarily a self‐regarding vice, protected by the principle of liberty. Although my drunkenness may harm those who care about me and affect the interests of those who might have benefited by my sobriety, it can only violate the rights of those to whom I have an obligation that I cannot meet due to my being a drunk, such as my family and creditors. The conduct that society may demand of individuals is not that one never harm others, even without their consent, but that one does not harm “certain interests which, either by express legal provision or by tacit understanding, ought to be considered as rights” (Liberty, XVIII: 276).11 Among the harms to which one is not entitled to protection from society, Mill includes harm from persuasion, harm by example, and offense. We will need to explain why Mill does not consider these (and certain other) harms to count as reasons for compulsion, but the first point to note is that these are exactly the harms most likely to be caused by the expression of opinion and sentiment.
Hence the central argument for the pragmatic interpretation – that because speech can do harm, it cannot be a self‐regarding act protected by the principle of liberty – is belied by the fact that Mill expressly grants that some self‐regarding action is harmful. What makes an act that would otherwise be self‐regarding fall out of that class, and lose its immunity, is not that it does harm but that it violates an obligation. This doctrine implies that not all harms are violations of rights, since rights ground obligations of non‐interference; and that is just what Mill avows. Moreover, the harms that can be attributed to the profession of specific opinions and sentiments, such as those involving persuasion and offense, are not such violations. Hence the potential harmfulness of the expression of opinion and sentiment does not make speech a social act, subject in principle to regulation by society. This belies the pragmatic interpretation, but it also raises an obvious question. What is the principle of liberty if it is not a harm principle, and what kinds of action does it protect from interference?
We can understand the place of the defense of freedom of speech in Mill’s philosophy only by recognizing that the purpose of the principle of liberty is to circumscribe a sphere of liberty where the individual is sovereign and cannot rightfully be subject to coercion. While this accords with Mill’s summation of the principle in On Liberty, the doctrine is more clearly expressed in his Principles of Political Economy, where he writes:
[T]here is a circle around every individual human being, which no government … ought to be permitted to overstep. That there is, or ought to be, some space in human existence thus entrenched around, and sacred from authoritative intervention, no one who professes the smallest regard for human freedom and dignity will call into question: the point to be determined is, where the limit should be placed; how large a province of human life this reserved territory should include. 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.
(Principles, III: 938; emphasis added)12
Mill declares that there is a significant region of human liberty, containing “large departments of human life from which [coercive interference] must be unreservedly and imperiously excluded” (Principles, III: 937).
The fact that the harms characteristic of the expression of opinion and sentiment do not count as violations of rights allows Mill to be peremptory in his argument for freedom of speech, and specifically to claim that even harmful and immoral speech counts as self‐regarding action. While this claim still has to be defended, the textual evidence already given suffices to belie the pragmatic interpretation. First, Mill expressly denies its central claim by stating that freedom of speech is one element of the sphere of liberty. Second, Mill’s conception of self‐regarding action includes certain acts harmful to non‐consenting others: those that harm in ways against which we are not protected by right. If the principle of liberty is understood not as a harm principle but instead as circumscribing a sphere of liberty, which expressly includes the liberty of thought and discussion, then that helps explain the peremptoriness of his argument.
Mill repeatedly states that all opinions and sentiments should be tolerated – that is, permitted expression though not exempted from criticism – including immoral and even harmful opinions. “However positive any one’s persuasion may be, not only of the falsity but of the pernicious consequences – not only of the pernicious consequences, but … the immorality and impiety of an opinion” (Liberty, XVIII: 234), none have the right to prevent the opinion from being heard. Hence he states that it is irrelevant to his argument whether the doctrine of “tyrannicide” is immoral. Its advocacy should be tolerated in any case, 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, XVIII: 228n).
The qualified interpretation, according to which Mill only defends the freedom to express certain opinions, such as those that are offensive but (otherwise) harmless, thus faces a substantial argumentative burden. Even so, it must be allowed that Mill has a tendency to make broad pronouncements, not to say exaggerations, which he later qualifies. He calls the principle of liberty a very simple principle, for instance, and no one takes that claim seriously. Immediately after stating the principle, he draws from it a ban on paternalistic reasons just as sweeping as his defense of speech rights. An agent’s “own good, either physical or moral, is not a sufficient warrant” for compelling him against his will, Mill declares (Liberty, XVIII: 223). Yet he later seems to qualify that claim by banning agents from selling themselves into slavery, on the grounds that the very point of the principle is undermined by this application of it. The strongest version of the qualified interpretation makes an analogous suggestion that speech can be prohibited when it is harmful specifically to the “deliberative goods” such as autonomy, the promotion of which is supposedly the point of freedom of speech.
The qualified interpretation relies crucially on an example Mill gives of speech that is not immune from censorship: the corn dealer example. The gist of the case is that Mill identifies a context in which an opinion whose expression should ordinarily be tolerated loses its immunity from coercive suppression. Mill writes,
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, XVIII: 260)
Recall that drunkenness, although typically protected by the principle of liberty, ceases to be self‐regarding when it prevents the drunk from meeting a specific obligation, such as to his creditors or family. Mill here makes an analogous point about “opinions” – by which he explicitly means their expression: that an opinion loses its immunity from interference when the context in which it is expressed makes it constitute incitement to murder. Nevertheless, it remains the case that the doctrine that corn dealers are starvers of the poor, or that property is theft, must be tolerated.
What is at issue between the absolutist and the qualified interpretation is whether Mill’s willingness to prohibit the expression of such opinions in specified contexts, such as to an angry mob gathered outside the corn dealer’s house, constitutes a limitation on freedom of speech motivated by its harmfulness. The qualified interpretation holds that opinions that are not just offensive but harmful, or not just harmful but harmful to others’ interests, or that significantly diminish autonomy, can rightfully be censored. Whatever the merits of this position as a matter of policy, it is problematic as an interpretation of Mill because he grants that some opinions are pernicious and yet advocates their toleration nevertheless. The opinion that corn dealers are starvers of the poor threatens to harm the interests of corn dealers simply by being advocated, regardless of context. Indeed, Mill considered Proudhon’s dictum that property is theft – the other opinion mentioned in this passage – to be harmful to the interests of the poor as well as the rich.13 Were he willing to consider silencing harmful opinions as such, he would not endorse its toleration in ordinary discursive contexts where it does not constitute incitement to imminent violence. But he does, and he makes a point of reiterating that in this passage. Hence the corn dealer case does not conflict with Mill’s insistence that an opinion’s falsity, immorality, or pernicious consequences does not put it beyond the pale of toleration.
Rather, the point of the corn dealer example is to explicate the notion of freedom of speech that Mill defends. It is not the freedom to say anything at anytime, anywhere. It does not conflict with noise regulations at libraries or prohibitions on speech in monasteries. Nor does it immunize all of the actions that can be performed with words. As Mill says, no one pretends that actions should be as free as opinions. The corn dealer example illustrates that, in certain peculiar but not unrealistic contexts, an act that would ordinarily be merely an expression of opinion constitutes the more performative action of incitement to violence. In that context, it is not a self‐regarding act protected by the principle of liberty, because it violates an obligation. The fact that certain tokens of the expression of an opinion can be prohibited does not count as an exception to the absolute freedom of speech Mill advocates, because it does not prohibit the expression of any doctrine as a matter of ethical (or political or religious) discussion. Proudhon must be allowed to advance his doctrine, regardless of its merits – which is just what Mill concludes.
In arguing against the pragmatic interpretation, I showed that Mill defends freedom of speech as part of the sphere of liberty protected as a matter of rights. I have now argued, against the qualified interpretation, that Mill’s claim to offer an absolute and unqualified defense of freedom of speech is not an exaggeration. This argument depends partly on having the proper conception of what constitutes freedom of speech, since an overly broad “anything, anytime, anywhere” conception is indefensible. But Mill clearly states what he is defending: the claim that the expression of any opinion or sentiment, propounded as a matter of ethical discussion, ought to be protected from coercive interference regardless of its merits. The “fullest liberty of professing and discussing” any doctrine requires that opinions of any content be tolerated. It does not rule out limitation on what now gets called the time, place, and manner of speech; nor does it immunize everything one can do with words. The importance of the corn dealer example is that it illustrates that the typically self‐regarding act of professing a doctrine can, because of its context, count as incitement to violence and thereby lose its immunity – but not its claim to toleration in other contexts. In this respect, the expression of opinion is analogous to other paradigmatic self‐regarding action, such as drinking, which loses its status as self‐regarding in just those circumstances where it violates an obligation.
Mill’s argument for freedom of speech, and for the principle of liberty more generally, require argument – not just about what rights individuals do have but those they do not. A right not to be offended, for instance, would undermine the absolute and unqualified defense of freedom of speech, which extends to offensive opinions. Mill notes this point specifically about drinking, when he denies the claim made by a contemporary temperance group that society has the right to police the moral ecology by demanding sobriety. Mill responds to this claim incisively:
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 me …
(Liberty, XVIII: 288)
This is an especially important passage because it illustrates two points. First, that rights claims are double‐edged. Certain putative rights, such as the right to determine one’s moral ecology – or to be protected from offensive opinions – would not increase freedom but diminish it. Second, that Mill disparages a narrow conception of the liberty of thought, which allows one to hold any opinion but only if one keeps it secret. The liberties of conscience must be understood more broadly than that, Mill contends, so as to include the right to express one’s opinions and sentiments.
Although it is challenging to reconcile Mill’s liberalism with his utilitarianism, he gives a similar account and justification of rights throughout his mature work, including both On Liberty and Utilitarianism. “When we call anything a person’s right,” according to Mill, “we mean that he has a valid claim on society to protect him in the possession of it” (Utilitarianism X: 250). There are no natural (or “abstract”) rights for Mill. But although rights are founded on utility, he insists that this must be understood as “utility in the largest sense, grounded on the permanent interests of man as a progressive being” (Liberty, XVIII: 224), capable of improvement. These individual rights trump considerations of utility in specific cases, because they arise from those moral rules that most concern the essentials of human well‐being; hence they ground more absolute obligations than other rules of conduct.14
In short, Mill defends freedom of speech as an absolute and unqualified right, on the basis of an indirect utilitarian argument about the prerequisites for human well‐being. This conclusion can be challenged, of course, and recently freedom of speech has increasingly come under attack, most notably from the progressive left. In its willingness to sacrifice liberty for equality and other collective goods, progressivism rejects several of the foundations of classical liberalism. Mill’s position thus helps explicate the distinction between liberalism and progressivism, which has been obscured by the current, nearly synonymous use of those terms. Whereas (classical) liberalism is a doctrine of individual rights and personal responsibility, progressivism is overtly skeptical of those individualistic notions. Although a progressivist interpretation suits the political views of many modern readers, it founders as an interpretation of Mill. In the final section, I will show how his classically liberal approach helps buttress the empirical claims necessary to defend the principle of liberty and, in particular, freedom of speech.
We’ve already noted conclusive textual evidence that Mill considers certain harms not to count as reasons to interfere with individual liberty. What these harms have in common is that, in Mill’s view, they would be counterproductive to try to prevent, when considered in terms of the utility of man as a progressive being, capable of improvement. Some of these harms are the products of agency, in that the individual herself is responsible for them. This point comes out most clearly in Mill’s rejection of paternalism, but it also holds for the harms of persuasion and example, which are also ruled out as reasons for compulsion. He thinks that agents will learn from their mistakes only if they are held responsible for doing so rather than being protected from error by others. Another example is offense: Any attempt to protect people from even significant emotional harm would inhibit the atmosphere of intellectual freedom that is a prerequisite both for individual flourishing and the cultivation of genius; and the right to avoid false, immoral, or harmful opinions would, in several respects to be discussed, undermine justification and knowledge. Finally, to protect people from the harm of losing would impede the great engine of improvement, namely competition. These would be counterproductive secondary principles, according to Mill, whereas the principle of liberty – including its recognition of individual rights of freedom of conscience in the broadest sense, which expressly includes freedom of expression – can be justified as contributing essentially to human well‐being.
These arguments amount to a relatively straightforward, though hardly uncontroversial, indirect utilitarian justification of certain individual rights (but not others) and a similarly founded critique of paternalism. They rely on substantive empirical claims about what does and does not generally promote human flourishing in the long run. What interpreters of Mill have largely failed to appreciate, in my view, is how profoundly agent‐centered his approach to morality is.15 Mill holds that the development of humanity must primarily be a matter of self‐development, because we learn best when we are allowed to enjoy the fruits of our successes and to suffer the pains of our mistakes. In effect, Mill places a principle of personal responsibility alongside his principle of individual liberty. While he does not name this responsibility principle, as he does the liberty and utility principles, his commitment to it is quite clear; and it helps reconcile his liberalism with his utilitarianism. Mill thinks it would be a mistake to compromise these principles, which he holds to be so crucial to human development and flourishing, by attempting to distinguish exceptional circumstances with more fine‐grained moral rules. These dual principles of individual rights and personal responsibility form the core of the classical liberal tradition, which too often gets obscured not only by the ambiguity in the term liberal but also by the current tendency to construct a false dichotomy between progressivism and laissez‐faire.
As important as he takes the development of character and civilization to be, Mill does not approach them as goods to be maximized through exogenous means, whether those means involve compulsion or benevolence. Instead, the recurring theme in Mill’s work is his focus on self‐development. This was one of his main criticisms of Bentham:
Morality consists of two parts. One of these is self‐education; the training, by the human being himself, of his affections and will. That department is a blank in Bentham’s system.
(Bentham, X: 98)
Mill criticized Bentham’s conclusions about even the half of morality he recognized – that is, those external actions that can be considered without reference to their moral qualities – on the grounds that he systematically ignored the effects of action on the agent herself. In an early letter to Thomas Carlyle, where he calls himself a utilitarian only in an idiosyncratic sense, Mill proposes that the good of humanity can be furthered only “by each taking for his exclusive aim the development of what is best in himself” (Letter to Thomas Carlyle, 12 Jan 1834, XII: 207; emphasis in original). This was not just an early deviation from utilitarian orthodoxy but a lifelong commitment. In his late work, The Subjection of Women, Mill warns against “unenlightened and shortsighted benevolence” that, by taking control of their lives out of people’s own hands and relieving them of responsibility for their own action, “saps the very foundations of self‐respect, self‐help, and self‐control which are the essential conditions both of individual prosperity and social virtue” (Subjection, XXI: 330).
In order for people to flourish, they must be at liberty to shape their own lives and held responsible for their choices. This does not mean that other people, and society at large, should not attempt to guide and educate them; on the contrary, Mill thought we should do more by way of persuasion. But the misguided attempt to coerce others toward their own good, or to provide it for them, proves counterproductive because it undermines virtue. Similarly, the attempt to shield them from offense and error, whether factual or moral, proves counterproductive because it undermines knowledge and practical reason. Instead the “[c]omplete liberty of contradicting and disproving our opinion, is the very condition which justifies us in assuming its truth for purposes of action” (Liberty, XVIII: 231). Thus freedom of speech is to knowledge as anti‐paternalism is to virtue. Mill has classically liberal worries about the unintended consequences of benevolently motivated action, including censorship for the sake of the common good, which no doubt many modern progressives will consider exaggerated. But the merits of these arguments are not here at issue. The question is how to interpret Mill, and on that score there can be little doubt.
Perhaps the most telling example of Mill’s commitments to the principles of individual rights and personal responsibility arises from his view of democracy. He held that the best form of government at any given time will be the one that most contributes to the advancement of society; and representative government is the ideal form for those societies that have the prerequisites to sustain it, because it best promotes the self‐development of its members. Nevertheless, Mill saw several dangers inherent to democracy. The most famous is the tyranny of the majority, against which the best protection is the institution of individual rights. But Mill also thought that democracy tends to promote conformity of thought even without compelling it. In order to palliate this tendency, he advocated a form of democracy that “breaks the headlong impulses of popular opinion … [by] adverse discussion” (Tocqueville on Democracy in America [I], XVIII: 180). By adverse discussion Mill meant confrontation with opposing points of view, which he considered necessary for knowledge and virtue. This means that democracy in particular requires, “above all, the utmost possible publicity and liberty of discussion” (Considerations, XIX: 436).
The second chapter of On Liberty offers various Millian arguments for the necessity of freedom of speech and the cultivation of diversity of opinion. Rather than go into the details of those specific arguments, which are profound in their essence but can be quibbled with in detail, it is more fruitful to see Mill as anticipating the phenomena now known as epistemic closure and confirmation bias. Epistemic closure refers to the tendency to surround oneself with those of like opinions and thereby avoid adverse discussion, which people typically find onerous. And confirmation bias refers to the common habit of finding evidence that confirms our antecedent opinions especially salient, while ignoring or discounting disconfirming evidence. These are among the psychological phenomena explaining Mill’s observation that “the weakest part of what everybody says in defence of his opinion, is what he intends as a reply to antagonists” (Liberty, XVIII: 251). Mill’s insight is that it is not just intellectual dishonesty that causes people to argue against straw men, but unwitting ignorance of the best opposing arguments. Yet in order to be justified in one’s beliefs, one must be able to answer the strongest arguments against them. Mill concludes this argument just as the absolutist interpretation would have it:
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, XVIII: 257–8)
Mill’s classically liberal views about self‐development and the importance of cultivating diversity of opinion render far more plausible the empirical claims necessary to ground his defense of freedom of speech (and the principle of liberty more generally). 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 and unqualified as he expressly and repeatedly claims. Contrary to the pragmatic interpretation, Mill considered the expression of opinion to be quintessentially self‐regarding action, which lies at the center of the sphere of liberty protected by rights. And contrary to the qualified interpretation, even the expression of harmful opinions, such as those targeted by modern progressives, must be tolerated in standard contexts where they do not constitute more performative speech acts that are not similarly immune. These commitments reflect not the weakness of speech but its paramount importance to human flourishing.