The Case Against Censorship: Alan Haworth, John Stuart Mill and Why Fallible Governments Cannot Be Trusted to Decide What We May Think

Freedom of speech is frequently praised as one of the foundations of Western civilisation, yet governments throughout the democratic world appear increasingly willing to restrict it. The justifications vary. Sometimes censorship is presented as protection against misinformation, sometimes as a necessary response to hate speech, and increasingly as a means of protecting citizens from psychological or emotional harm. Whatever the justification, the underlying assumption remains remarkably consistent: some authority must determine which ideas are sufficiently dangerous, offensive or misleading to justify their suppression. This raises an old philosophical question. What entitles governments, regulators or other powerful institutions to decide what citizens may hear, read and discuss? More importantly, how can those institutions be confident that their judgments are correct? The British political philosopher Alan Haworth has provided an important contemporary defence of freedom of expression by reconsidering the arguments of John Stuart Mill, particularly those developed in Chapter II of On Liberty (1859). Haworth's major study, Free Speech (1998), and his article, "On Mill, Infallibility, and Freedom of Expression" (2007), deserve renewed attention in an era when censorship is increasingly presented as an administrative necessity rather than an assault upon liberty.

Mill's argument begins with a deceptively simple observation. Human beings are fallible. Individuals make mistakes, governments make mistakes, scientific communities make mistakes, and even the most widely accepted beliefs can eventually be overturned. If an opinion is suppressed because those in authority believe it to be false, society may lose access to a truth. Even if the suppressed opinion is mistaken, it may contain an element of truth overlooked by the prevailing orthodoxy. And even where the accepted opinion is substantially correct, confronting serious objections can deepen understanding and prevent true beliefs from degenerating into unexamined dogma. Mill famously argued that silencing an opinion amounts to an assumption of infallibility. Those who prohibit discussion behave as though their judgment is sufficiently authoritative to justify denying others the opportunity to examine the disputed position. Haworth recognises that this argument faces a philosophical difficulty. A censor need not literally believe himself infallible. A government might acknowledge that its judgment could be mistaken while nevertheless concluding that the expected harms of allowing certain speech outweigh the benefits of permitting it. For example, officials might argue that a particular claim is probably false and potentially dangerous, and that suppressing it is therefore a reasonable precaution. Such reasoning does not necessarily involve a claim to perfect knowledge. Does this objection defeat Mill's argument? Haworth's achievement is to show why the deeper Millian case survives. The crucial issue is not whether every censor consciously believes himself incapable of error. It is whether fallible institutions should possess the authority to settle intellectual controversies by preventing citizens from examining competing arguments. The problem is institutional as much as psychological. A government may acknowledge its fallibility in theory while establishing a system that treats its preferred judgments as effectively beyond challenge. When an opinion is prohibited, the public loses opportunities to test the official position against alternative evidence and reasoning. The institution responsible for suppression also becomes less exposed to the corrective effects of criticism. This produces a dangerous asymmetry. The government retains the authority to decide what may be discussed, while citizens lose part of their capacity to evaluate whether that authority has exercised sound judgment. The central insight is that censorship can undermine the very processes through which errors are discovered and corrected.

Mill's argument therefore concerns more than the rights of individual speakers. It concerns the public interest in maintaining an intellectual environment where ideas can be challenged, revised and sometimes rejected. This is particularly important in science. Scientific progress has repeatedly involved the correction of established theories. The history of science is not simply a succession of discoveries added to an ever-growing body of unquestionable truths. It also involves conceptual revolutions, methodological disputes and the recognition that apparently secure explanations were incomplete or mistaken. The point is not that every unconventional theory deserves equal credibility. Most speculative claims are wrong, and some are demonstrably absurd. But the proper method of distinguishing serious criticism from error is generally investigation and argument, not administrative prohibition. A society that protects prevailing scientific opinion from challenge risks confusing institutional consensus with epistemic certainty. The same problem arises in politics. Governments routinely make judgments about economic policy, public health, national security, immigration, education and social organisation. These judgments may be supported by substantial evidence, but they remain vulnerable to error, unforeseen consequences and changes in circumstances. When governments acquire the power to suppress criticism of their policies, they weaken the mechanisms through which mistakes can be identified. This is especially troubling because political authorities have interests of their own. A government may sincerely believe that a particular opinion is harmful, but it may also benefit politically from preventing that opinion from gaining influence. The danger is not limited to deliberate dishonesty. Institutional self-interest can influence what officials perceive as dangerous, irresponsible or unacceptable.

Haworth's reconstructed Millian position offers a powerful response. Freedom of discussion should be defended as a general public-interest principle because society benefits from preserving the conditions of critical inquiry. This defence is not dependent upon the extravagant claim that every opinion is equally valuable or that discussion invariably produces truth. Public debate can be confused, emotional, manipulative and sometimes profoundly irrational. Falsehoods may circulate widely, and persuasive speakers can attract followers despite weak arguments. Nevertheless, the existence of these difficulties does not establish that political authorities possess the knowledge, impartiality or competence required to determine which ideas should be removed from public consideration. Indeed, the fallibility of public discussion must be compared with the fallibility of censorship itself. There is no reason to assume that government officials are immune from the errors, prejudices, intellectual fashions and institutional pressures affecting ordinary citizens.

Haworth also examines the consequentialist character of Mill's argument. Freedom of expression is valuable partly because of the beneficial consequences of open inquiry. Critics have long argued that consequentialism appears to permit censorship whenever authorities believe suppression will produce better outcomes. This objection is serious. If liberty is justified solely by its consequences, why should it not be restricted when the consequences appear unfavourable? The Millian response requires attention to long-term institutional effects rather than merely the immediate consequences of a particular controversy. Suppose a government suppresses a disputed opinion because officials believe it will cause social harm. Even if that particular decision appears beneficial, the precedent may encourage further restrictions. Citizens may become reluctant to express unpopular views. Researchers may avoid controversial questions. Journalists may decline to investigate sensitive subjects. Public institutions may gradually become less tolerant of criticism. The costs of censorship therefore extend beyond the individual statement being prohibited. A regulatory system can create what is commonly called a chilling effect, where people refrain from otherwise lawful expression because they fear punishment, reputational damage or administrative intervention. Such effects are particularly difficult to measure. We can observe an article being removed or a speaker being punished. We cannot easily identify all the arguments never written, investigations never undertaken or discoveries never made because individuals concluded that the risks of speaking were too great.

This leads to an important development of Haworth's argument. Human beings are not merely fallible in their judgments about whether particular opinions are true. They are also limited in their ability to predict the consequences of suppressing those opinions. A censor cannot know with certainty what intellectual developments might have followed from permitting a disputed argument to circulate. Nor can officials reliably anticipate every indirect consequence of establishing a regulatory power that future governments may interpret differently.

The epistemological case against censorship is therefore stronger than a simple appeal to the possibility that an unpopular opinion might be correct. Censorship involves an attempt to control an intellectual process whose future development cannot be comprehensively predicted. The government must make judgments not only about the truth or falsity of disputed claims, but also about their social consequences, the benefits of suppressing them, the effects upon future inquiry and the institutional behaviour encouraged by the restrictions. Each judgment introduces additional uncertainty. From the standpoint of epistemic limitationism, this is particularly significant. Human cognitive capacities are limited, and the complexity of social and intellectual systems places substantial constraints upon our ability to predict their development. The problem becomes especially acute when governments attempt to regulate broad and ambiguous categories such as harmful misinformation, offensive expression or material considered psychologically damaging. What constitutes harm? Who determines the relevant standards? How are disagreements resolved? What happens when the official definition of harm reflects the political or cultural assumptions of those exercising regulatory authority? A government may regard certain opinions as dangerous precisely because they challenge prevailing institutional arrangements.

This does not mean that freedom of expression must be absolute. Mill himself distinguished between expressing an opinion in public discussion and expressing it in circumstances where it may contribute directly to immediate harm. His famous example concerns the claim that corn dealers starve the poor. Publishing such an opinion is different from delivering it to an angry crowd assembled outside a corn dealer's house under circumstances where violence may follow. Context matters. Credible threats, targeted harassment, criminal exploitation and direct incitement to violence raise different questions from peaceful political criticism or intellectual disagreement. The philosophical challenge is to establish restrictions narrowly enough to address genuine harms without transforming the mere possibility of offence or distress into a general justification for censorship.

Haworth's work is especially relevant to contemporary debates about online regulation. Governments increasingly seek to impose duties upon digital platforms to prevent vaguely defined forms of harm. Rather than directly prohibiting every controversial opinion, regulators can create financial and legal incentives for technology companies to remove material that might attract official scrutiny. The result may be censorship exercised through private corporations rather than through explicit government orders. This arrangement introduces further epistemological problems. Automated moderation systems can misunderstand context, satire, quotation and legitimate disagreement. Commercial platforms may lack the incentive to investigate disputed claims carefully. Removing controversial material can be cheaper than defending its publication. Meanwhile, the government can maintain that it has not directly censored anyone, since the final decision was made by a private company. Yet the institutional consequences may be similar. Citizens lose access to arguments, speakers become reluctant to challenge prevailing positions, and public debate is shaped by authorities whose decisions are difficult to scrutinise.

Haworth's defence also has implications for universities. Universities traditionally exist to pursue knowledge through teaching, research and critical discussion. Their intellectual legitimacy depends upon the willingness to question assumptions and examine competing explanations. A university that suppresses controversial arguments merely because they offend prevailing institutional sensibilities undermines its own purpose. Academic freedom does not require universities to treat every claim as equally credible. Scholarly standards, evidence and rigorous criticism remain essential. But these standards should be applied through intellectual evaluation rather than ideological conformity. The distinction between rejecting an argument because it is poorly supported and prohibiting it because it is politically inconvenient is fundamental. There is also a wider democratic concern. Citizens cannot meaningfully evaluate governments if important political questions are removed from public debate. Elections alone do not guarantee democratic freedom. A political system may preserve regular voting while progressively restricting the range of opinions citizens can encounter. Democracy requires more than the periodic selection of rulers. It requires continuing opportunities to criticise policies, challenge institutions and expose official mistakes. Haworth's Millian defence helps explain why freedom of expression should not be treated as a minor privilege to be balanced casually against every new administrative objective. It is one of the conditions that makes responsible political judgment possible.

The strongest objection remains that unrestricted discussion can itself produce harm. False claims may influence behaviour, inflammatory rhetoric may intensify social conflict, and manipulative propaganda may exploit human vulnerabilities. These concerns cannot simply be dismissed. But the appropriate response is to ask whether particular restrictions are necessary, proportionate and supported by convincing evidence, and whether less restrictive alternatives are available. Counterargument, public education, transparency, independent journalism and the exposure of false claims may often provide better responses than suppression. The burden of justification should rest heavily upon those proposing censorship, because restrictions on discussion threaten the processes through which society evaluates the justification for those very restrictions.

This is the deeper importance of Haworth's contribution. He preserves the anti-authoritarian force of Mill's argument without depending upon the logically questionable claim that every censor must believe himself infallible. Governments may acknowledge uncertainty while still seeking to control public discussion. What matters is whether their authority is exercised within institutions that permit continuing criticism and correction. A genuinely fallibilist political order would recognise that no government, expert committee, university administration or technological corporation possesses a final monopoly upon truth. Such an order would protect the freedom to challenge established positions precisely because even highly confident judgments may be mistaken.

There is a paradox at the centre of modern censorship. Governments often justify restrictions by claiming to protect citizens from error. Yet the restrictions themselves may prevent society from discovering errors committed by governments. The more extensive the censorship, the greater the danger that official mistakes will become insulated from correction. The principle can be stated simply: because human beings are fallible, no institution should be casually entrusted with the power to determine the boundaries of permissible thought. And because human beings are also limited in their ability to predict the future consequences of suppressing ideas, the case for intellectual liberty becomes stronger still. This does not establish that every form of expression must be permitted under every circumstance. It does establish that censorship requires a much more demanding justification than the assertion that particular opinions are offensive, misleading or potentially harmful. Mill recognised that the pursuit of truth depends upon the possibility of error being exposed through criticism. Haworth has helped reconstruct that insight for contemporary political philosophy. In an age of expanding digital regulation, bureaucratic management of public discourse and growing pressure upon independent commentators, their arguments deserve renewed attention.

Freedom of speech is not merely the right to express opinions that governments already approve. Its real importance emerges when individuals challenge what authorities regard as settled. A society confident in its institutions should be prepared to defend them through reasoned argument rather than administrative suppression. The greatest danger of censorship is not simply that governments may silence truths they dislike. It is that they may progressively destroy the intellectual conditions required to discover when they themselves are wrong. That is why a genuinely free society must tolerate disagreement, controversy and the persistent possibility that its most cherished assumptions will eventually require revision. The alternative is an intellectual order in which fallible authorities increasingly exercise powers that only infallible beings could safely possess.

Haworth, A. (2007). On Mill, Infallibility, and Freedom of Expression. Res Publica, 13, 77-100.

Haworth, A. (1998). Free Speech. Routledge.

Mill, J. S. (1859). On Liberty. John W. Parker and Son.