Social pressure and the Principle of Liberty

Mill sets out the Principle of Liberty in the following passage:

The Principle in this passage applies not only to legal penalties but also to ‘the moral coercion of public opinion’, and says both must be kept out of self-regarding matters. In Mill's view, only the vulgar think tyranny is a problem of political despotism alone. Reflective persons see that society can itself tyrannize and not only through its political structures.Footnote ⁶

Mill has many different terms for what infringes upon liberty. He uses at least these: power, prevention, punishment, coercion, compelling, constraint, hindrance either physical or moral, molestation, moral repression, impediment, and interference. Let the term ‘interference’ stand for all of them. Mill's Principle bans interfering with liberty in the self-regarding sphere. But his Principle does not take all social influences to infringe on liberty. It permits reasoning, remonstrating, persuading, advising, warning of feelings of distaste and contempt, and exercising a qualified right to avoid people of whom one disapproves. Since on the face of it, the persuading, warning, and avoiding look like sources of power, moral repression, and so on, the obvious questions are whether Mill can genuinely distinguish between the influences that interfere with liberty and the influences that do not and whether the influencing the principle permits would threaten the liberty he wants. The first question is about the coherence of Mill's view and the second about how liberal it is.

Let me explain why I think Mill's presentation of his view encourages some mistaken criticisms. In the first place, despite having said how important social influence is, Mill does not make it easy to see how he thinks it relates to liberty. Much of the book focuses on government and law rather than social pressure. For instance, plenty of chapter two is about religious legal disabilities and censorship, chapter three describes the legally permitted physically coercing of supposed eccentrics and lunatics, and large parts of chapter four and nearly all of chapter five are about legal prohibition and such matters as regulation of alcohol, going to war against Mormons, the government provision of public goods, and the validity of contracts. Moreover, when Mill does describe social pressure, he often uses abstractions, such as society's ‘mandates’,Footnote ⁷ ‘tyranny of the prevailing opinion and feeling’,Footnote ⁸ ‘penalties of law or opinion’,Footnote ⁹ and ‘public opinion now rules the world’.Footnote ¹⁰ Obviously, we do not know from these terms how these mandates, tyranny, penalties, or ruling actually work. In sum, the problem of social pressure could do with a more explicit and detailed treatment to avoid misunderstanding and confusion. And then we have more fog because Mill gives a misleading impression when he sets out the Principle of Liberty quoted above. The paragraph follows several announcing that the problem of liberty is not primarily political, that the problem in England is more the yoke of opinion than law, and that we have as yet no principled basis for distinguishing legitimate from illegitimate interferences. When Mill then writes that the object of his Essay is to assert the Principle of Liberty, the reader naturally thinks that Mill intends the Principle to be the complete answer to the problem of ‘the nature and limits of the power which can be legitimately exercised by society over the individual’.Footnote ¹¹ If it were a complete answer, the Principle would be open to the objection that it permits social pressure that threatens liberty. However, the impression misleads; the Principle is not Mill's complete answer. We must distinguish the Principle of Liberty from Mill's overall treatment of liberty and appreciate how limited the Principle is.

Within Mill's overall treatment are various factors, about which the Principle is silent, that causally determine the amount of liberty in a society. For instance, Mill thinks the greatest difficulty in implementing the Principle is that people do not value spontaneityFootnote ¹² and he thinks one condition for liberty in his time is that the educated part of the public comes to value individuality.Footnote ¹³ Here the Principle of Liberty does not aim to create the right attitudes; respect for liberty is the consequence of the right attitudes. Mill also thinks that European nations have avoided the stationariness of China by having diverse individuals, classes, and nations, but this diversity, and so liberty, are under threat.Footnote ¹⁴ Here Mill is writing of diversity as the cause of liberty, not liberty as the cause of diversity. These are factors that determine the amount of liberty but are outside the scope of the Principle of Liberty. I come back to this point at the end when I consider the resources Mill has to shape social influence.

If we consider causal factors that threaten liberty more directly than reduced diversity and undervalued individuality, we can see that the Principle of Liberty is also silent about some of these. When Mill writes that ‘everyone lives as under the eye of a hostile and dreaded censorship’,Footnote ¹⁵ he means that people conform because it does not occur to them not to conform, not because they are frightened. The ‘mind is bowed to the yoke’Footnote ¹⁶ in a process Mill describes as ‘insidious’.Footnote ¹⁷ This insidious process does not work through the interference of penalties so it is outside the scope of the Principle of Liberty. Take another text of Mill's, ‘The Utility of Religion’.Footnote ¹⁸ Mill does not discuss liberty in it; his primary point is that society and individuals can achieve the value of religion without believing in the supernatural. He does, though, helpfully distinguish three leading social influences on the human mind.Footnote ¹⁹ The first is ‘the empire exercised over mankind by simple authority, the mere belief and testimony of their fellow creatures’.Footnote ²⁰ The second is education, understood broadly to include upbringing. The third is ‘the power of public opinion; of the praise and blame, the favour and disfavour, of their fellow creatures’.Footnote ²¹ People want approval and fear disapproval both as ends and as means to such benefits and costs as getting or being denied a job. Authority, education, and public opinion can all affect liberty, understood broadly to include free and open minds but, of these, the Principle of Liberty only regulates public opinion. The Principle is thus but a limited part of what I am calling Mill's overall treatment of liberty which, in On Liberty and others of Mill's works, has plenty to say about authority, education, and other causes of mental conformity.Footnote ²²

Even when it comes to influences that might fall under public opinion and be considered social pressure, the Principle of Liberty does not, as critics point out, forbid all the influences, such as remonstrating and avoiding, that might compromise liberty. But the incompleteness of the Principle is unsurprising given its place in a utilitarian theory and given that, as the next section explains in more detail, social influences have benefits as well as costs. Principles and rules often have a derivative status in utilitarianism and are selected and justified according to whether they would lead to the best outcomes. The Principle of Liberty is said by Mill to have a derivative status of sorts. It derives from combining utility, ‘the ultimate appeal on all ethical questions’, with certain empirical claims about when following the principle would lead to the greatest happiness.Footnote ²³ Because it derives from utility, the Principle of Liberty has certain scope limitations, for instance not applying to children since liberty for them would not promote utility. And also because it derives from utility, the Principle does not apply to all social influences that might threaten liberty. The relation between social influence and utility is too complex to be entirely regulated by a single principle.

On the reading offered here, some influences would increase utility, others decrease it. Society and government might produce more utility overall if, instead of assessing case by case the utility of influencing, they followed a principle ruling out a class of influences. That is what the Principle of Liberty does: subject to its scope limitations, it rules out a class of influences in self-regarding matters. However, the Principle permits other influences that are too various in their effects to be handled by one principle and so leaves it open for other aspects of society to be altered to increase the utility benefits of social influence while decreasing the disutilities. The Principle of Liberty should not then be criticized for failing to solve completely the problem of social pressure because it is embedded in a theory that can provide other solutions. To appreciate this understanding of Mill, it will help to see the values that compete with each other and so make regulating social pressure too complex for a single principle.

Why did Mill allow social influence?

Mill announced the principle of no moral coercion in the self-regarding sphere in chapter one of On Liberty, argued for almost unlimited freedom of expression in chapter two, and praised individuality in chapter three. As chapter four begins, Mill denies that his doctrine so far is ‘selfish indifference’.Footnote ²⁴ People ought to be concerned for and to some extent ought to try to promote the welfare of others. Mill accepts that people can make mistakes in their own lives. He also accepts that people can choose what they themselves consider less valuable because of their infirmity of will. He thinks the actions of others can help. He states that one can learn from the stock of accumulated wisdom even in the self-regarding sphere.Footnote ²⁵ He says that people sometimes need encouragement. For these reasons, Mill thought that people not only may but often ought to advise, encourage, and warn for the sake of the recipients. He also thought that people should be free to express their opinions (so chapter two said) and, for the sake of their own individuality, be free to choose their associates.

We can think of Mill as having these three aims: preventing interferences by society in the self-regarding sphere; having people benefit from receiving advice, encouragement, and warnings; and allowing the free expression of one's opinions and the free choice of one's associates. The problem is that the aims seem to conflict. This problem is not just a problem for Mill. We clearly can benefit from advice and encouragement, even advice and encouragement for which we did not ask; and yet advice and encouragement can be oppressive. It seems absurd to tell us who our friends must be; and yet individuals can choose not to associate in a way that amounts to a collective exclusion such as shunning. And even if we would not go so far as Mill in banning all interferences in the self-regarding sphere, Mill seems obviously correct to say that society can interfere too much with liberty even if it confines itself to social pressure rather than political or legal coercion. It is not at all easy to reconcile the values of advice and the freedoms of expression and association with the appropriate freedom from social control.

Mill did not actually seem to have felt himself agonizingly pulled apart by the conflict of values. To be sure, Mill's style is not one of the ‘tragic choices’ kind, but he did on other occasions concede that matters were difficult to decide.Footnote ²⁶ I think it striking that he gave the impression of thinking it easy to reconcile his three aims and wonder whether he underestimated the problem. Be that as it may, we can now understand why his simple principle did not simplistically try to prevent all social influences. That leaves Mill with the task of coherently distinguishing the influences that are permitted from those that are forbidden. What the Principle of Liberty turns out to forbid are influences that interfere, where interfering is characterized as intentionally inflicting a cost for the sake of punishing. The next section explains the distinction and defends its attribution to Mill and its coherence.

Interference as punishment

For Mill, some but not all social influences are interferences. Social influences interfere when they intentionally impose costs for the purpose of punishment. It is interfering influences that the Principle of Liberty rules out in self-regarding matters.Footnote ²⁷ This section presents his view in brief and then develops it in reaction to a rival interpretation, by Joseph Hamburger, who argues instead that Mill wanted an elite to punish people for their mistakes in the self-regarding sphere. The section then explains how interference-as-punishment can avoid the line-drawing problem of establishing how costly an influence has to be to interfere.

To explain Mill's view of interference, the first point to see is that not all the costs one might suffer from one's self-regarding actions are interferences. People who spoil their lives by mismanagement do pay a penalty.Footnote ²⁸ Thus one of the costs of being drunk can be a hangover and one of the costs of living beyond one's means can be running out of money. But one is not, unusual cases aside, being interfered with if left to pay these costs. Of course these are not social influences so the next step is to consider the social influences Mill permits. These are such penalties as being ‘lowered in the opinion of others',Footnote ²⁹ regarded with distaste and contempt,Footnote ³⁰ being avoided, and denied optional good offices. Mill writes, ‘We are not bound, for example, to seek his society; we have a right to avoid it (though not to parade the avoidance), for we have a right to choose the society most acceptable to us', and ‘We may give others a preference over him in optional good offices, except those which tend to his improvement'.Footnote ³¹ Unlike being hungover or impoverished, these are social influences. They impose costs that arise from how people react to behaviour they do not like.Footnote ³² Why, then, are they not interferences? Mill answers as follows:

The crucial claim here is that social influences may impose costs and yet still not interfere so long as the costs are not purposely inflicted for the sake of punishment.

To continue the overview, let us add the site for interference. Mill claims that punishment is appropriate only for breach of duties. The only genuine duties are duties to others. One does not have a duty to oneself.Footnote ³⁴ Therefore punishment is not justified in the self-regarding sphere. To summarize, influences are interferences only if they take the form of punishment; if they do, they are forbidden in the self-regarding sphere.

This reading of Mill is not universally accepted. In his book-length critique of Mill as illiberal, Joseph Hamburger claims that Mill wanted a moral elite to be free to punish the non-elite in self-regarding matters.Footnote ³⁵ As part of his case, Hamburger makes two claims: that Mill himself understands the penalties he permits to be a form of punishment and that Mill envisaged an illiberal scope limit to his Principle of Liberty. Both of these claims are mistaken, as I now try to show.

Let us return to the various costs from social influences, beginning with being lowered in the opinion of others. An important text is this one:

This passage can be used to explain why the judgements and feelings that people have, and the warnings they give, about other people's self-regarding behaviour are not punitive. To explain, let us agree that some costs of one's behaviour, such as the hangover, are not punishment because a hangover is not voluntarily inflicted whereas punishing someone is voluntary because it intentionally inflicts a cost. The penalty of being the object of distaste and contempt is then like the hangover in being a spontaneous cost and so not intentionally inflicted. As the quotation says, a person ‘could not’ have elevated taste without having certain feelings, someone may ‘compel us to judge him’, the feelings of distaste and contempt occur ‘necessarily’. So the existence of these feelings is not punishment. That leaves the warnings that one will be held in distaste and contempt. Deciding to warn someone is voluntary; but to warn is still not to punish, as it would not punish a man to warn him that he was about to cross an unsafe bridge. Telling someone of the non-voluntarily controlled consequences of their actions does not punish them because it does not intentionally create a cost.

The thought here is that being held in distaste or contempt does not punish because the feelings are compelled or necessary and so at least this type of natural penalty might be explained away as not interference. However, Hamburger points out that Mill not only describes but endorses certain reactions. Mill indeed does say several times, as in the passage above, that someone might be ‘properly a subject of distaste’, and when he writes: ‘I do not mean that the feelings with which a person is regarded by others, ought not to be in any way affected by his self-regarding qualities or deficiencies. This is neither possible nor desirable.’Footnote ³⁷ Moreover, expressing distaste and contempt cannot be explained away entirely as an involuntary reaction. As Hamburger states: ‘But this does not remove the difficulty, unless one assumes that those expressing distaste and contempt cannot control their reactions.’Footnote ³⁸ He remarks that Mill can hardly have thought his elite were unable to control themselves.Footnote ³⁹ On Hamburger's account of Mill, the feelings of distaste and contempt are sometimes desirable and their expression controllable; so when the elite express those feelings, they voluntarily inflict costs and so punish, albeit justifiably.

A fairer interpretation than Hamburger's can explain Mill's view that the feelings of distaste and contempt can be justified, desirable, and controllable without being punishments. Mill believes that people will not only judge that some ways of life are better than others but that some ways of life really are better and these judgements can play a valuable role in getting people to lead better lives. These judgements are not going to be colourless statements. If one thinks someone is leading a particularly good or bad life, then of course one will have strong feelings about them. These feelings are natural and spontaneous. They are also ‘proper’ (Mill's endorsement) because they are inseparably associated with the valuable social practice of judging ways of life. They may in some cases affect behaviour in undesirable ways but on balance the judging is desirable, as well as unavoidable. This interpretation of opinions about good and bad ways of life resembles what Mill says about competition for jobs.Footnote ⁴⁰ Competition for jobs, like the existence of these opinions, produces losers as well as winners, and the losses are genuine losses. But the losses are outweighed by the overall benefits. Hence it can be desirable and justified for people to have feelings of distaste and contempt. But merely having those feelings is not to punish.

Hamburger asked whether people, or at least the people in Mill's elite, should be regarded as unable to control their expressions of their feelings.Footnote ⁴¹ But Mill did not say, contrary to Hamburger, that people should express their feelings of distaste and contempt. He does say that it would be a service to warn people if they would be regarded with distaste and contempt; he does not say that, if those warnings are not given or are ignored, the feelings of contempt should be communicated. Contrary to what Hamburger sometimes says, Mill does not say he wants people to suffer the natural penalties. Mill does, however, permit people to express their feelings. ‘If he displeases us, we may express our distaste, and we may stand aloof from a person as well as from a thing that displeases us.’Footnote ⁴² So Hamburger might say that Mill at least permits the penalties even if he does not welcome them and so he permits interference. But again Mill would not regard these penalties as interference (in the sense of being governed by the Principle of Liberty) because they are not intentionally inflicted for the sake of punishing the person who bears them. The penalties are the by-product of one's own permitted reactions.

Now one might regard Mill's account of interference as inadequate: surely costs might not be intentionally inflicted and yet still cause too much social pressure. That would be a reasonable criticism to make and I try to defend Mill from it in the bulk of the rest of the article. The point here is that, adequate or not, it is Mill's account. Mill does think the natural penalties can be distinguished from being deliberately punished. In saying so, he is not committed to an indefensible view of people lacking self-control and this interpretation is not undercut by his remarks about what is proper.

What, then, of Hamburger's attempt to render Mill's thought coherent, albeit objectionable, by limiting the scope of the Principle of Liberty? Mill states certain limits to the scope of the Principle; for instance, it applies ‘only to human beings in the maturity of their faculties’.Footnote ⁴³ Hamburger thinks the Principle has a further scope limit. In his interpretation, the Principle applies to the dealings of ‘society’ with individuals, where ‘society’ refers to mid-Victorian Britain and societies like mid-Victorian Britain. But the Principle does not apply to a superior moral elite. Mill not only allowed but wanted this elite to interfere with the non-elite even in self-regarding matters. That is how he can coherently allow the social pressures of obtrusive comments, displays of distaste and contempt, and ostracism. The elite are allowed to use these forms of pressure in self-regarding matters because the Principle does not apply to them.Footnote ⁴⁴

On what grounds does Hamburger take this unlikely interpretation of Mill? The only ground is this: Mill lets people try to influence in self-regarding matters in forms that Hamburger thinks so obviously interfere with liberty that the only way to both permit them and keep the Principle is to interpret the Principle as not applying to an elite. Now Hamburger may be right that the influences Mill permits really would interfere with liberty and he may be right that his scope limit on the Principle renders it consistent with those liberty-limiting influences; but it is obviously another question whether Mill actually advocated that scope limit. If Mill would permit the elite to interfere, as Hamburger claims, why did he not just say so instead of trying to claim that these social influences were not interferences at all? Hamburger's interpretation makes no sense unless Mill realized his own view of interference was wrong, and if he did, why did he say it? In any case, when it comes to allowing the elite to interfere in self-regarding matters, Mill's own language is unambiguous. He would not allow them to interfere.Footnote ⁴⁵

Two examples from chapter four will make the point. (I chose this chapter because it is sometimes thought to mark Mill's pulling back from his liberal position in the book so far.) Mill says that when it comes to people's behaviour that affects only their own interests or the interests of others who have consented to be affected, ‘there should be perfect freedom, legal and social, to do the action and stand the consequences'.Footnote ⁴⁶ Second, to restate a point made above, Mill said that punishment is justified only for breaches of duty and one has no duties to oneself. Neither point leaves any scope for an elite to punish people for their self-regarding faults, contrary to what Hamburger claims.

I close this section by deploying the conception of interference as punishment to avoid a line-drawing problem. Alexander Bain, in an often-cited observation, wrote that one can divide up the forms of social pressure into many different degrees ‘from the most intense individual preference at the one end to the severest reprobation at the other’.Footnote ⁴⁷ Bain did not think it too difficult to decide which forms punish, but Geoffrey Scarre says that ‘the absence of a clear-cut divide between suffering inconvenience and being punished is problematic for Mill’.Footnote ⁴⁸ The problem Mill supposedly faces is that one cannot draw a workable distinction between the costs that do and do not constitute punishment. However, it is a mistake to think, as Bain and Scarre seem to, that the distinction must occur somewhere on a scale of cost. Differences in cost are not the way to differentiate between being fined $5 for refusing to be vaccinated, which is punishment, and being denied entry to public places, which is only a precaution against contagion but much more costly than the fine to those who are denied.Footnote ⁴⁹ The distinction lies in whether the cost was deliberately inflicted with the intention of deterrence or retribution. This distinction may not always be easy to apply. Differentiating punishing from not punishing relies on access to intentions and it can be hard to know one's own intentions let alone those of other people. Nonetheless, the distinction is not subject to the problem of drawing a line between the size of costs.

The interim conclusions are that Mill's Principle really does ban social influences that take the form of punishment in self-regarding matters; and that the distinction he draws between interfering and non-interfering influences is coherent so far. That leaves the question of whether the distinction is an important one, that is, whether it matters when it comes to preventing excessive social pressure.

Why focus on punishment and is the focus too narrow?

Mill's early critic, James Fitzjames Stephen, thought it impossible to confine the moral coercion of public opinion: ‘People form and express their opinions on each other, which, collectively, form public opinion, for a thousand reasons’ and the ‘opinion produces its effect in precisely the same way whatever was its origin’.Footnote ⁵⁰ Stephen's thought implies that the effect on liberty of public opinion does not depend on the origin of the opinion, only on such matters as what the opinion is and who holds it. In particular, we might think from Stephen that it makes no difference whether the opinion is a form of punishment. One question is therefore whether Mill's account of interference really tracks a genuine difference from the point of view of liberty. I shall argue that it does. Another question is whether Mill's ban on interferences in self-regarding matters is too narrow in leaving aside influences that could cause excessive conformity or social tyranny.Footnote ⁵¹ This question is not answered by the reasons one can derive from Mill to focus on punishment.

Mill states that: ‘The distinction between the loss of consideration which a person may rightly incur by defect of prudence or of personal dignity, and the reprobation which is due to him for an offence against the rights of others, is not a merely nominal distinction.’Footnote ⁵² The distinction is a real one, he goes on to say, because we, the social influencers, will feel and behave differently according to whether we are punishing or not:

This passage is written from the perspective of those doing the influencing. However, we are asking why it matters to social pressure whether an opinion punishes or not. Since the concern is about the loss of individuality, the question must be answered from the perspective of those who are the influenced as opposed to the influencers. To restate Stephen, why should punitive influences have any more effect on the influenced than non-punitive influences?

If we consider the passage from the perspective of the influenced, one aspect is straightforward. The influenced would suffer costs from punishment on top of such natural penalties as being hungover or impoverished and would be more prone to conform to avoid those costs. A hangover plus a fine is more of a deterrent than a hangover alone. It is less straightforward to see how the ‘vast difference in our feelings’ would affect the behaviour of their targets. Mill may mean that the element of moral criticism in punishment is itself especially motivating. For example, when recommending making a crime of infecting one's spouse with syphilis, Mill said that criminalizing would be effective not just because it would impose costs but because of what it would communicate: ‘the mere effect of placing its mark on the conduct in this way would have very great influence, and would make this crime be considered, as in truth it is, one of the gravest a man could possibly commit’.Footnote ⁵⁴

Thus punishment plausibly would pressurize more than natural penalties alone because punishing would inflict extra costs and communicate moral disapproval.Footnote ⁵⁵ Mill is not then arbitrarily banning punishment while permitting advice, warnings of distaste and contempt, and avoidance. On behalf of Mill, one can not only distinguish influences that punish from those that do not but also see how the distinction makes a difference. Nonetheless, as the next section shows, the Principle permits social influences that could still impose costs and so could pressure people into conforming. Thus we still have the question posed earlier: is the focus on punishment too narrow?

Residual sources of social pressure

The main point of this section is that Mill's Principle does not rule out all the social pressure that threatens liberty. As we shall see, the Principle permits social criticism that Mill himself would agree constrains individuality; and the right of avoidance Mill accepts turns out to permit organized ‘shunning’. That said, the section begins by defending Mill against the criticism that he permits too much social pressure in wanting to override good manners.

In a passage already quoted, Mill wrote that people can benefit from being warned of their faults. He continues: ‘It would be well, indeed, if this good office [of warning] were much more freely rendered than the common notions of politeness at present permit, and if one person could honestly point out to another that he thinks him in fault, without being considered unmannerly or presuming.’Footnote ⁵⁶ This passage has invited the objection that ‘Mill has simply discarded a whole department in the institution of privacy and thereby exposed people to social pressures to account for themselves and to conform to others’ expectations’.Footnote ⁵⁷ I think this objection depends on reading Mill uncharitably.

The quotation from Mill says that it would be better if common notions of politeness were changed but it does not explicitly tell us whether people should warn or point out faults much more freely while those notions remain unchanged. So we cannot attribute to Mill in this passage the view that good manners should be overridden, only that what is regarded as good manners should ideally be changed. The objection to Mill could then be that if, as he hoped, the common notions of politeness did change so as to allow blunt comments, people would become too vulnerable in the self-regarding sphere. In Mill's defence, a change in the notions of politeness could modify the meaning of warnings and advice. Bluntness in a society with a code that rules it out might be more upsetting than bluntness in a society where it is not considered bad manners. People might be much more direct than, say, the English and also pride themselves on being direct. In such a society, people might feel less pressured than the English would when they are the target of direct comments. Suppose, though, that they do not.Footnote ⁵⁸ I think it is open to Mill to say that if people remain as sensitive as the English to personal criticism, then the existing code of manners should remain. Thus Mill might say that manners should change to permit more warnings and advice only if the existing code of manners were no longer so necessary to protect people from social pressure.

Having defended Mill against the charge of bad manners, let us consider some real problems that arise from his narrow focus on punishment. Consider drunkenness: for Mill, so long as it is self-regarding, others may not punish a drunk but may advise, warn, and avoid. What if instead we consider holding any unpopular religious or political view? Unless advising, warning, and avoiding somehow become interferences when they are in a direction he dislikes, Mill must allow people to advise, warn, or avoid those who might hold unpopular views. But these permitted ‘natural penalties’ impose or threaten to impose costs and thereby make people conform religiously or politically. Moreover, Mill himself seems to agree. In advocating individuality, he describes how people who do not conform ‘become a mark for society . . . to point out with solemn warning as “wild,” “erratic,” and the like’,Footnote ⁵⁹ and he says ‘the man, and still more the woman, who can be accused either of doing “what nobody does,” or of not doing “what everybody does,” is the subject of such as much depreciatory remark as if he or she had committed some grave moral delinquency’.Footnote ⁶⁰ The context makes it clear that this social criticism is effective in cramping individuality. Earlier, Mill said that ‘unmeasured vituperation employed on the side of the prevailing opinion really does deter people from professing contrary opinions, and from listening to those who profess them’.Footnote ⁶¹ Since vituperation can be for finding non-moral faults and since people can regard someone as wild and to be avoided without thinking them immoral, the social criticism could be permitted by the Principle of Liberty despite threatening individuality.

The right to avoid someone for self-regarding faults also allows social pressure. Mill writes:

Hamburger, pressing his thesis that Mill is illiberal, says ‘this is ostracism’.Footnote ⁶³ Ryan says, in effect, ‘this is not ostracism’.Footnote ⁶⁴ As we shall now see, Ryan is strictly correct but the right to avoid does allow more social pressure than he and some other commentators have acknowledged.

Ryan assumes that ‘no right to parade’ should be read as ‘ought not to parade’. Actually, for Mill ‘no right to X’ does not entail ‘a duty not to X’ (one does not have a right to win in fair competition for a job but if one does win, one may hold the job without breaching a duty). Still, Mill probably did mean that parading one's avoidance would be wrong. What then follows? Ryan interprets the duty not to parade as part of a ban on coercion in the self-regarding sphere. He states that ‘coercion is involved where harm is organized to deter someone from an action, or in retribution for action’.Footnote ⁶⁵ Jeremy Waldron interprets ‘no right to parade’ as ruling out any ‘concerted effort to cast [someone] out of society altogether’.Footnote ⁶⁶ Both these quotations run together two separate ideas: first, the methods of organizing or concerting, and second, the purpose of the organizing or concerting (deterrence, retribution). If, as Mill says, one has a right to warn others, why would one not have a right to do so in an organized way? If one member of a religious community may warn other members that one of their flock is setting a bad example by failing to fulfil his religious obligations, why may that member not warn in the community's newsletter? Of course, Mill ruled out ‘organized efforts to humiliate in public’,Footnote ⁶⁷ but then he would rule out unorganized efforts to humiliate too. What in fact Mill opposed is not organizing but punishing in self-regarding matters. The distinction between the right to avoid and the right to parade is a version of a now-familiar distinction between costs that are and are not intentionally inflicted for the sake of punishment. In this case, the distinction is between avoiding for one's own sake and avoiding with the aim of inflicting a cost. If this characterization is correct, and if ostracism is to be taken to be punishment, Mill's right to avoid does not permit ostracism. Nonetheless, organized avoidance – call it ‘shunning’ – is not ruled out even though, obviously, it can be a source of social pressure.Footnote ⁶⁸

Thus the Principle does not deal with all the social pressure that On Liberty warns against. The Principle only regulates interference and Mill defines interfering so narrowly that the Principle does not regulate genuine cases of social pressure, such as being criticized or shunned. How damaging this observation is to Mill's treatment of liberty depends in part on what other resources he might have to direct social influence. They are the subject of the next section.

Influencing influence outside the Principle of Liberty

I can see nothing like a theory in Mill's writing about how to deal with residual social pressure.Footnote ⁶⁹ What one can find are some examples and ideas that suggest ways to have the good social influence without too much of the bad. The theme of these examples is the relation between independence and non-conformity.

I begin by pointing out some options for Mill. All sorts of different factors besides the nature of an influence determine the extent of social pressure, so perhaps changing the factors would alter the pressure. While Mill does not give us a list of these factors, here are some, all of which can be found in On Liberty:

1. 1. What the influence is. Moral criticism, warnings of distaste, and avoidance are examples. This is the factor discussed so far in this article.

2. 2. Who it is coming from, such as a dominant social class, one's own social groups, or one's spouse. (See below.)

3. 3. How sensitive someone is to social pressure given their personalities (e.g. whether they are strong or timid).Footnote ⁷⁰

4. 4. How much people stand to lose by withstanding social pressure (also see below on independence).

The first factor is regulated by the Principle of Liberty: punishments in self-regarding matters are not allowed, other forms of influence are. As for the other factors on the list, they could perhaps be altered. However, even if altering them were feasible, it may not be desirable. Remember that for Mill social influences have benefits as well as costs. Suppose people could somehow be made not to care about what others think in self-regarding matters. For Mill, they may then pay too high a price in well-being by not learning from others or not having their will strengthened by others.Footnote ⁷¹

So much for background. Now for the first example:

This passage draws attention to the instrumental costs of not conforming and how independence thus removes a motive to conform. A later passage says that having the support of those with rank would also reduce the motive; what matters is being independent of those doing the pressuring, not independence as such.Footnote ⁷³ Mill also writes above about the degree of resistance to pressure one can reasonably expect from people. Independent people ought not to fear being ill-thought of or ill-spoken of.

Consider next an example from The Subjection of Women. Writing in one part about men, Mill considers why some do not stick with their minority political or religious beliefs. The explanation is social pressure, which does not, in Mill's story, directly affect the men. It comes instead through their wives. ‘Whoever has a wife and children has given hostages to Mrs Grundy. The approbation of that potentate may be a matter of indifference to him, but it is of great importance to his wife.’Footnote ⁷⁴ The wives care so much because they have little opportunity for social reward other than status and because, if they have daughters, the absence of opportunities for them makes a good marriage so important. In this example, removing the legal disabilities for women would greatly reduce the pressure on husbands to conform. Of course, Mill thinks the legal disabilities should be removed anyway, primarily for the sake of women's opportunities but also because removing the disabilities would reduce the pressure exerted on women by a society that is less tolerant of their individuality than it is of men's.Footnote ⁷⁵ But to return to the example, we can draw the lesson from it and from the previous one that people can be socially influenced for the wrong reasons. We do want people to be open to social influence, but when it comes to finding a life that suits, and leading a life that is one's own, we might not want people deviating from what they would like to do in order to gain the social benefits.

These examples could thus be taken to offer a generalization about the good and bad effects of social influences. We had one generalization already with influences that interfere: in general, interfering would have bad enough effects in the self-regarding sphere that a principle, the Principle of Liberty, should rule them out. Now we can generalize further about social influences: influences which work though dependence would tend to have bad effects. These bad social influences are not, though, to be dealt with by a principle regulating what sorts of influences are allowed in what circumstances; they are to be dealt with by reducing or removing dependence, which could be done in many different ways, from removing legal disabilities for women to reducing class inequality to introducing more workers’ cooperatives.Footnote ⁷⁶