The Oxford Debate That Refused to Fear Offense

Ann Widdecombe’s closing argument against no-platforming at the Oxford Union in 2019 was not a defense of every opinion voiced in public; it was a warning that a society which suppresses offensive ideas may eventually suppress necessary ones. Six years on, as universities struggle over controversial speakers, institutional neutrality and the boundaries of acceptable debate, that warning has lost none of its force. The fight for viewpoint diversity as well as other forms of diversity continues. What is in our heads is more important than the color of your skin or your gender?

The Question Behind the Applause

The motion before the Oxford Union sounded simple: “This House Supports No Platforming.” The argument beneath it was not simple at all. It asked whether a free society should prevent certain people from speaking because their views are offensive, dangerous or repellent, or whether those views should be exposed and defeated in public.

That distinction is often blurred. Opponents of no-platforming are accused of defending the speaker rather than the principle of open debate. Supporters of exclusion are accused of wanting to silence every dissenter. Neither description captures the real difficulty. The issue is not whether a university should approve of a speaker’s views. It is whether the institution believes its students are better protected by hearing and challenging bad arguments or by ensuring that they never reach the room.

Widdecombe’s case was that the second choice carries a cost that is rarely acknowledged. A banned idea does not become false because a committee refuses to host it. A controversial speaker does not vanish because a student society is denied a venue. The argument continues elsewhere, often in places where it is less likely to meet evidence, scrutiny or an opposing voice.

That is the central weakness in the no-platforming argument. It treats public silence as though it were public defeat. But the two are not the same. Defeat requires an argument to be answered. Silence often leaves it untouched.

The Oxford Union’s importance made the decision more than an internal student dispute. Founded in 1823, it has provided a stage for politicians, dissidents, campaigners and intellectuals whose views have often unsettled the wider establishment. Its chamber is valuable not because every speaker is wise, but because the institution has historically understood that intellectual life depends on confrontation.

Widdecombe used that history as part of her case. When a body with the Union’s reputation decides what may be heard, she argued, the decision carries beyond its walls. A refusal to host a controversial speaker is not merely a scheduling choice. It is a statement about whether the institution trusts argument enough to permit its enemies to speak.

Britain’s Older Test of Tolerance

The most powerful part of Widdecombe’s speech did not begin with campus politics. It began with the memory of war.

She recalled growing up in the years after the Second World War, when British families were still living with the loss of sons, husbands and fathers killed in the fight against Nazi Germany. The wounds were not remote. They were present in homes, streets and communities. The symbols and organisers of fascism were not abstract historical curiosities; they represented a political movement that had brought Europe to ruin.

Yet Britain still permitted figures such as Oswald Mosley and Colin Jordan to organise and hold rallies. Their appearances caused anger, hurt and disgust. Many people who had fought fascism or lost relatives to it would have regarded those rallies as an insult. The offence was real. The grief was real. The country nevertheless made a choice to preserve a broad space for political speech.

Widdecombe’s point was not that the British public should have welcomed fascist propaganda. Nor was she suggesting that the state had no right to prosecute crimes or prevent violence. Her point was narrower and more demanding: a society that had fought a war in the name of liberty could not claim to value freedom while allowing only respectable opinions to be expressed.

That is an uncomfortable standard because it requires the majority to tolerate conduct it considers morally offensive. It asks people to separate the right to speak from the worth of what is said. It also asks them to resist an easy emotional argument: that the more offensive an opinion is, the less deserving it must be of protection.

The opposite may be true. A right that protects only harmless or popular views is not much of a right. The real test comes when speech is painful, provocative or repugnant. No one needs a special defence for an opinion that every institution already approves. The need for principle appears when approval disappears.

This does not turn every platform into a sanctuary. Public debate has always included limits relating to threats, incitement and criminal conduct. But those limits are different from a general power to exclude speech because it wounds, embarrasses or disturbs. Once offence becomes the governing test, the boundary moves with the mood of the institution and the pressure of the loudest campaign.

The Cold War and the Burden of Consistency

Widdecombe then reached for another historical test: the Cold War.

For decades, Britain lived under the shadow of Soviet weapons and the possibility of nuclear conflict. Communism was not merely a theory debated in a seminar room. It was the ideology of a hostile superpower whose government controlled much of Eastern Europe and presented itself as an alternative to parliamentary democracy. Support for communist politics could provoke suspicion, anger and fear.

Yet communists were permitted to stand for Parliament. The Communist Party operated openly. The Morning Star could be sold on British streets, including to readers who regarded its politics as hostile to the country’s interests. The state did not treat political advocacy as equivalent to treason merely because the ideas behind it were dangerous or discredited.

Again, the principle was not that the public had to admire the message. It was that citizens should be allowed to encounter it and reject it. The government could act against espionage or violence. It did not claim a general right to prevent the circulation of an unpopular political doctrine.

This history exposes a weakness in modern arguments for no-platforming. The standard often appears to shift according to the identity of the speaker and the political fashion of the moment. Views once treated as dangerous are later defended as part of legitimate debate. Views currently treated as beyond the pale may, in time, be reconsidered. Institutions that create broad exclusionary powers rarely know how those powers will be used under a different leadership or against a different cause.

The question is therefore not simply whether today’s target deserves a platform. It is whether the rule used to deny that platform would be acceptable if the political balance changed. A principle that appears attractive only when applied to one’s opponents is not a principle. It is a weapon.

Widdecombe’s historical examples were designed to impose that burden of consistency. If Britain could endure the public expression of communism during a period of intense geopolitical danger, why should universities assume that students cannot hear a controversial speaker today? If the answer is that modern students are more vulnerable, the claim requires evidence rather than assertion. If the answer is that some ideas are too harmful to be heard, the institution must explain who decides and how the decision can be challenged.

Without those safeguards, no-platforming becomes a form of intellectual paternalism. It assumes that students must be protected from arguments before they have had the opportunity to learn how to answer them.

Offence Is Not a Policy Test

Widdecombe’s humour in the Oxford debate served a serious purpose. She told the chamber that she had been offended, upset and insulted many times in a single day. She joked about being wounded by fellow debater Celeste McGinley and about the scorn poured on her Paso Doble.

The joke worked because it exposed a confusion at the heart of contemporary speech disputes. To be offended is to experience a reaction. It is not, by itself, a finding that another person should be denied a hearing. If offence becomes sufficient evidence for suppression, every disagreement can be converted into a demand for institutional intervention.

People are offended by arguments about religion, immigration, sex, race, national identity, war, policing and public spending. They are offended by jokes, flags, historical monuments and the opinions of people they believe should know better. Some of those reactions are justified. Others are shallow. Many are sincere but still insufficient to determine public policy.

A university that treats offence as a veto will not create a calmer intellectual environment. It will create a permanent contest over who can claim the greatest injury. The person with the strongest emotional response will have an advantage over the person with the strongest evidence. Student groups will learn that protest, outrage and reputational pressure are more effective than debate. Administrators will be encouraged to avoid risk rather than defend principle.

That is not tolerance. It is a system of informal censorship administered through fear of complaint.

Widdecombe divided the enemies of free expression into two broad categories. On one side were what she called “snowflakes”: people who seek insulation from discomfort and regard exposure to unwelcome views as a form of harm. On the other were totalitarians, who seek not merely to avoid offence but to eliminate dissent and establish an orthodoxy that cannot be challenged.

The two tendencies are different in scale and intention, but they can reinforce each other. A culture that treats discomfort as a reason to suppress speech creates habits that a more authoritarian movement can exploit. Once institutions accept that some ideas must disappear for the sake of social peace, the argument over which ideas qualify becomes a struggle for control.

The danger is not that every university will become a dictatorship. The danger is more ordinary. It is that administrators, lecturers and students will gradually learn that certain subjects are professionally or socially unsafe, and that the safest course is to avoid them. A university can remain formally open while becoming substantively cautious.

When Social Orthodoxy Replaces State Censorship

Traditional censorship is visible. A government bans a newspaper, arrests a dissident or prohibits a meeting. Responsibility is clear, and resistance can be organised against a recognisable authority.

The modern pressure Widdecombe identified is often less direct. A speaker may not be banned by the state. Instead, a campaign forms online. A petition gathers signatures. A student society is threatened with reputational damage. A venue withdraws. A sponsor expresses concern. An administrator announces that the event cannot proceed because of security costs or the possibility of disruption.

The result may look less like censorship because no official order has been issued. Yet the practical effect can be similar: the speaker does not appear, the audience does not hear the argument and the institution avoids the controversy. The power is dispersed, but the silence is real.

Widdecombe described this as social orthodoxy. The phrase matters because it draws attention to the way authority can operate without a censor’s office. Social media campaigns and institutional shaming can establish a boundary around permissible speech. People learn where the boundary lies not through a clear rule but through examples made of those who cross it.

This form of pressure is difficult to regulate because it hides behind individual choice. A university can say that no one prohibited the event; the venue simply became unavailable. A society can say that no one silenced the speaker; members merely withdrew support. A campaign can say that it only criticised the invitation; the organisers made their own decision.

Each statement may be technically true. Together they can produce a system in which controversial views are heard only when nobody powerful objects. That is a fragile definition of freedom.

The problem is not protest itself. Protest is part of free expression, and people who object to a speaker should be allowed to say so. The problem arises when protest is used to demand that others lose the ability to hear, question or challenge that speaker. The right to object is then turned into a right to control the room.

Universities should be particularly wary of this shift. Their purpose is not to guarantee agreement. It is to make disagreement more informed. If a speaker is wrong, students should be able to identify the error. If the speaker is dishonest, they should learn how to expose the dishonesty. If the argument is hateful, they should understand why it is hateful and how it can be answered.

A social system that removes the speaker before the audience can test the claim may protect reputations, but it does little to educate.

Bad Ideas Do Not Disappear When Banned

Widdecombe’s most important practical argument was that bad ideas should be defeated in public rather than driven underground.

She offered the example of a Holocaust denier. The instinctive response to such a person is understandable: why give a platform to someone who denies a historical crime of overwhelming evidence and immense human consequence? But Widdecombe argued that she would prefer to confront the denial directly, dismantle it point by point and show the audience why it fails.

That approach is not risk-free. A skilled propagandist can exploit a debate to create confusion. A speaker may use the appearance of impartial discussion to give a false claim an undeserved aura of legitimacy. Publicity can amplify a message that would otherwise have reached fewer people.

Those are serious objections. They do not, however, prove that exclusion is always the better answer. They show that public institutions need strong formats, informed opponents and clear standards. A debate can include a knowledgeable critic, a chair who enforces the rules and an audience capable of judging competing claims. The answer to manipulation is not always disappearance; it may be better preparation.

The underground argument is stronger still when applied to political movements that already have access to alternative channels. A banned speaker can move to online platforms, private networks or sympathetic media. Supporters can portray the exclusion as proof that the establishment fears the truth. The ban then becomes part of the movement’s story about persecution.

That does not mean every banned person is vindicated. It means suppression can produce unintended effects. It may harden the speaker’s followers, attract people who enjoy forbidden ideas and prevent moderates from seeing the claims challenged in an open setting.

Public argument has the opposite virtue. It allows an institution to demonstrate confidence in evidence. It gives the audience a record of the disagreement rather than a rumour about what was supposedly said. It creates a chance for an opponent to make the case against the speaker before listeners who may not otherwise encounter it.

Widdecombe’s phrase was blunt: bad causes destroyed through public argument remain destroyed. Silenced causes may continue gathering grievance. A defeated argument loses credibility because it has been answered. A censored argument can claim that it was never fairly heard.

This is why no-platforming is often a poor substitute for intellectual work. Suppression can be arranged by committee. Refutation requires preparation. It demands facts, historical knowledge, confidence and the willingness to listen closely enough to understand what must be answered. A university that prefers cancellation may be choosing administrative convenience over education.

The Heretics Who Changed the Terms

The strongest argument against permanent exclusion is historical uncertainty. Institutions rarely know which challenges will later be regarded as necessary.

Widdecombe invoked William Wilberforce, Emmeline Pankhurst and Martin Luther King Jr. Their causes are now associated with major advances in public morality and political rights. But none began as a comfortable contribution to the prevailing conversation.

Wilberforce confronted a slave trade that was embedded in Britain’s economy and defended by powerful interests. His campaign was not simply a request for a minor adjustment to public policy. It challenged a system that many respectable people regarded as normal, useful or too entrenched to change.

Pankhurst and the suffragettes rejected the political arrangements of Edwardian Britain. Their tactics offended contemporary standards of propriety and order. Many people who opposed them regarded their actions as reckless, disruptive and threatening to the social fabric.

King’s civil-rights campaign likewise challenged laws and practices that governments and majorities defended as legitimate. His civil disobedience was not designed to make the established order feel comfortable. It was designed to expose its injustice.

The lesson is not that every radical is a future Wilberforce or King. History contains many foolish and destructive causes. The lesson is that a society cannot reliably identify the future value of a dissenting idea at the moment it appears. The people judging a challenge are often the people whose assumptions it threatens.

That is the problem with no-platforming as a general principle. It asks the present to certify the limits of the future. It gives committees the power to decide not only which ideas are wrong but which ideas are too illegitimate to be heard. Such power may be exercised with good intentions and still produce intellectual stagnation.

There is also a deeper irony. Universities celebrate historical figures who challenged the orthodoxies of their time while imposing narrow boundaries on present-day dissent. Students are encouraged to admire old rebels after their arguments have become safe. The harder task is to tolerate an argument before its merits have been settled.

A serious educational institution should preserve room for that uncertainty. It should not confuse moral confidence with historical wisdom. The fact that a claim is unpopular may be evidence of nothing more than its unpopularity. The fact that an institution feels threatened may indicate that the speaker has identified a weakness, or it may indicate that the speaker is wrong. Only engagement can begin to distinguish the two.

The Personal Cost of Meeting an Enemy

Widdecombe’s defence of open debate was not delivered from a position of emotional distance. She described an encounter with a woman who had supported the Republican cause in Northern Ireland and whom Widdecombe believed had backed a movement associated with violence against her.

The context was personal. Widdecombe had narrowly missed the 1984 bombing at the Conservative Party conference in Brighton, an attack that killed five people and injured many others. She recalled looking at the woman and asking, in effect, whether she had wanted to blow her up. The answer involved support for the Republican cause, though not necessarily for the specific bombing.

Many people would have considered refusal to share a platform understandable. They might have concluded that the political disagreement was too closely tied to a violent campaign and that engagement would confer respect on an opponent who had not earned it.

Widdecombe made a different choice. She said she did not object to sharing a platform because argument offered a way to expose what was wrong in the woman’s position. The choice was not sentimental. It was not a declaration that the past did not matter or that victims should forgive those who supported violence. It was an insistence that moral revulsion should not automatically end the possibility of political argument.

This distinction is crucial. Free speech does not require emotional neutrality. It does not ask victims to pretend that their opponents have caused no harm. It asks whether the response to harm should be an attempt to silence or an attempt to persuade, rebut and hold the offender accountable.

There are occasions when a person may reasonably refuse to appear with another speaker. Consent, safety and the purpose of the event matter. No one is obliged to accept every invitation. The institutional question is different. A university deciding whether an audience may hear a controversial figure is exercising power on behalf of others, including students who may wish to challenge the speaker.

Widdecombe’s example shows what the principle costs when it is taken seriously. It is easy to defend debate with an opponent whose views are merely irritating. It is harder to defend debate with someone linked to a cause that has threatened one’s own life. The point of the example is precisely that free speech cannot depend on comfort.

If openness survives only in easy cases, it has not been tested.

The Voltaire Principle and Its Limits

Widdecombe ended by invoking the famous principle commonly attributed to Voltaire: that one should defend another person’s right to say something even while profoundly disagreeing with the content.

The quotation is often treated as a slogan, but its real force lies in the separation it demands. The right to speak is not the same as the right to be believed. The right to publish is not a guarantee of a readership. The right to hold a view does not protect the view from ridicule, investigation or condemnation.

What it protects is the possibility of making the case.

That possibility matters because public knowledge is not produced by agreement alone. It emerges through challenge, correction and the testing of claims. An argument that cannot survive scrutiny should be exposed as weak. An argument that survives may deserve further consideration. A claim that is never heard cannot be evaluated by those who have not encountered it.

The principle also places duties on those who defend open debate. They must not pretend that every speaker is equally credible. They must not describe all criticism as censorship. They must not use free speech as a shield against consequences that are part of ordinary public argument. A speaker can be invited and still be answered. A university can host a controversial event and still make clear that it rejects the speaker’s views.

That is why the argument for free expression is not an argument for institutional passivity. It is an argument for better institutions. A university should provide context, ensure that opposing views can be heard and protect participants from threats or violence. It should distinguish between a controversial opinion and conduct that crosses a legal or moral boundary.

The difficulty is that these distinctions require judgement. No-platforming often appears attractive because it avoids the need to make them. If the speaker is not present, there is no need to chair the discussion, prepare a rebuttal or explain to students why the claims fail. But avoiding judgement is not the same as exercising it well.

The Voltaire principle therefore does not say that all speech is harmless. It says that the existence of harm must be assessed without abandoning the presumption that arguments are best answered by arguments. The alternative is to place too much confidence in those who decide that a view is beyond the reach of debate.

What the Oxford Vote Actually Decided

When the Oxford Union rejected the motion in 2019, the result was reported as a victory for the free-speech side. That description is accurate, but incomplete.

The vote did not establish that every speaker should be invited to every event. It did not settle the legal boundaries of speech. It did not prove that controversy is always educational or that audiences can never be manipulated. What it did was reject a broad institutional commitment to no-platforming.

That distinction matters. A university may decide that an event is unsafe because of a credible threat of violence. A society may determine that a proposed speaker has no meaningful connection to the subject under discussion. An organiser may decline an invitation for practical reasons. Those decisions are not automatically no-platforming in the political sense.

The Oxford vote concerned the principle that certain views or speakers should be excluded because they are offensive or unacceptable. Rejecting that principle affirmed the Union’s role as a place where intellectual confrontation is part of the institution’s purpose.

The symbolism was especially strong because the Union has long understood that its reputation rests on the range of views it can accommodate. It is not important because it is always polite. It is important because it provides a setting in which political leaders, activists and critics can be made to answer questions in public.

Widdecombe’s speech connected that institutional identity to the practical work of debate. She did not present free speech as an abstract entitlement floating above history. She treated it as a discipline. Students must learn to hear arguments that unsettle them, identify errors and withstand the pressure to join a crowd against a dissenter.

The vote therefore had consequences beyond the motion. It set a standard against which future decisions could be judged. When a later event is cancelled, students can ask whether the decision reflects a genuine safety concern or a reluctance to face controversy. When a speaker is denied a venue, they can ask whether the institution has considered an alternative format that would permit scrutiny.

A decision to preserve a platform is not a declaration that the platform is morally neutral. It is a declaration that the audience is capable of judgement and that the institution is prepared to assist rather than decide on the audience’s behalf.

The University as a School of Civic Courage

The most practical lesson from Widdecombe’s argument concerns what universities are for.

A university is not a padded room in which students are protected from every claim that might offend them. Nor is it merely a marketplace where anyone can say anything without standards or responsibility. It is supposed to be a place where people acquire the knowledge and habits needed to judge competing claims.

Those habits include more than academic technique. Citizens must be able to hear an argument without assuming that exposure equals endorsement. They must separate evidence from rhetoric, identify emotional manipulation and ask whether a claim is supported. They must know how to disagree without turning every disagreement into a demand for expulsion.

None of those abilities develops when institutions remove every difficult speaker. Students may become skilled at recognising approved opinions while remaining unprepared for arguments they encounter outside campus. The wider world will not be organised around university speech codes. People will meet political claims online, at work, in elections and in their communities. They will need judgement, not permanent supervision.

This is why the language of protection can be misleading. A student who never hears a hateful claim at university may still encounter it elsewhere, without a lecturer, chair or fellow student present to provide a rebuttal. The university may have removed the immediate discomfort while leaving the deeper vulnerability intact.

Civic courage is not the absence of distress. It is the ability to remain rational and humane in the presence of distress. That ability can be practised through structured disagreement. A well-run event can ask students to listen, question and respond without endorsing the speaker. It can demonstrate that an institution is strong enough to reject a view without pretending that the view does not exist.

The same principle applies to lecturers and researchers. If academics believe that certain topics are professionally dangerous, they will avoid them even when those topics matter. The result is not a more truthful university but a narrower one. Students sense the boundaries, adapt their questions and learn that intellectual safety may depend on saying less.

A university that teaches courage cannot model fear. It must show that difficult ideas can be confronted without becoming contagious, and that a rebuttal is not a sign of weakness. The institution’s confidence should be visible in the quality of its responses.

The Questions Britain’s Universities Still Face

The issues surrounding campus speech have changed in detail since 2019, but the underlying conflict remains. Student societies continue to confront pressure over controversial invitations. Staff worry that a poorly phrased comment may trigger a campaign. Administrators weigh the cost of security and disruption against the value of allowing an event to proceed.

The existence of those pressures does not make no-platforming inevitable. It makes clear rules more necessary.

Universities should begin with a presumption that lawful speech can be heard. If an institution intends to refuse an event, it should identify a specific reason rather than rely on general language about reputational risk or possible offence. It should distinguish a credible threat from the possibility that protesters will object. It should explain why security, moderation or an alternative venue cannot address the problem.

The burden should not rest entirely on the proposed speaker. Institutions that cancel events often describe the decision as a neutral act of risk management. It is not neutral when one group’s disruption determines whether another group can meet. If the threat of protest becomes a routine basis for cancellation, the most aggressive campaigners acquire an effective veto.

That is a perverse incentive. It rewards those who threaten disruption and penalises those who seek to organise debate. A university committed to free expression must protect peaceful protest while refusing to let protest become coercion.

There must also be an honest recognition that speech can cause harm without justifying suppression. Holocaust denial, support for political violence and extremist propaganda deserve vigorous opposition. Students who encounter such claims should be given evidence and context. Speakers should face hard questions. Institutions should state clearly where they stand.

But a university that says “we reject this argument” is doing something different from a university that says “you may not hear it”. The first teaches discrimination between ideas. The second asks students to trust the censor’s judgement.

The choice is not between an open campus and a safe campus. It is between different understandings of safety. One model defines safety as the absence of disturbing speech. The other defines it as the presence of rules, evidence, protection from threats and confidence that disagreement can be managed. The second model is harder to administer, but it prepares students for citizenship rather than dependence.

The Cost of Choosing Silence

No-platforming is often defended as a narrow response to extreme cases. The difficulty is that institutions do not operate in a vacuum. A rule created for one type of speaker can be applied to another. A standard introduced to exclude threats can expand to cover offensiveness, then disagreement, then reputational inconvenience.

The expansion may happen without anyone deciding to abolish free speech. It happens through small precedents. One invitation is withdrawn because the speaker’s views are regarded as harmful. Another is cancelled because the audience may feel unsafe. A third is rejected because the event could attract negative publicity. Each decision is presented as exceptional. Over time, the exceptions define the rule.

The result is a university culture in which the absence of controversy is mistaken for intellectual success. Students may encounter fewer offensive opinions, but they also encounter fewer opportunities to understand how those opinions work. Lecturers may avoid disputes, but they also avoid subjects whose importance lies in their capacity to provoke.

Silence has a price beyond the event itself. It deprives the public of a record. People who did not attend may hear only that a speaker was deemed unacceptable, not what the speaker argued or how the opposition answered. The controversy migrates to social media, where claims are shortened, motives are assigned and outrage becomes detached from evidence.

That is a poor environment for truth. A formal debate creates a shared object: a speech, a question, a rebuttal. A cancellation creates a story about a speech that never happened. The second is easier to distort.

There is also a moral cost. If students learn that injury to feelings is sufficient to end a discussion, they may become less willing to distinguish between suffering and disagreement. Genuine threats, discrimination and harassment can then be placed in the same category as an unpopular opinion. When everything is described as harm, the word loses its ability to identify serious danger.

Widdecombe’s argument is ultimately an appeal to proportion. Not every insult is persecution. Not every controversial speaker is a threat. Not every demand for exclusion is an act of protection. Institutions that fail to make those distinctions do not become more humane. They become easier to manipulate.

A Principle That Will Be Tested Again

The Oxford Union’s 2019 debate did not end the dispute over no-platforming because no vote can settle the tension between liberty and protection. Every generation faces new controversies, new technologies and new claims that a particular idea is too dangerous to hear.

What the debate supplied was a standard for meeting those controversies. The standard is not indifference. It is not a demand that victims suppress anger or that universities abandon responsibility. It is the conviction that open societies should be reluctant to silence and eager to rebut.

That conviction rests on historical memory. Britain tolerated the public organisation of fascists after a war against fascism. It allowed communists to campaign during the Cold War. Reformers who now occupy honoured places in public history once disturbed the assumptions of their societies. None of these examples proves that every offensive speaker deserves an audience. They prove that present institutions should be cautious about declaring themselves qualified to close the future.

Widdecombe’s personal example adds a harder truth: free speech is most meaningful when defending it is uncomfortable. It is easy to support debate when the opponent is merely irritating. The principle matters when the opponent is linked to a cause that has caused real fear, grief or danger.

Universities will continue to face pressure to make difficult speakers disappear. The pressure will come from students, campaigners, staff, donors, politicians and online audiences. Some of the complaints will be justified. Some will be opportunistic. Many will combine legitimate concern with a desire to control who is heard.

The institution’s task is to judge carefully rather than surrender the decision to the loudest demand. It must protect people from violence and unlawful intimidation. It must permit protest. It must keep events orderly. But it should not mistake peace for truth or silence for victory.

The question posed at Oxford remains the one that matters: when an idea is offensive, do we trust our ability to answer it, or do we trust our power to make it disappear?

Widdecombe’s answer was direct. A free society should defeat its opponents in argument, not wish them away. That answer is demanding because it offers no permanent shelter from offence. It offers something more durable: the chance to test every orthodoxy, including the orthodoxy that some ideas are too dangerous to confront in public.