ROGER PARTRIDGE: What Freedom of Speech Is For: The case against silencing

In 1633, the Roman Inquisition condemned Galileo for heresy. His offence was to argue that the Earth moves around the Sun.

The Church was not acting out of malice. It was protecting a politically approved consensus against what was considered to be dangerous nonsense. The theologians and philosophers who condemned Galileo were not fools. They were defending what every educated person knew to be true.

They were also wrong. And being wrong with institutional authority behind you is far more dangerous than being wrong alone.

One of the unfailing truths of human history is that we get things wrong. Not occasionally – routinely. The authorities who silence heretics are not reliably better informed than the heretics they silence – and sometimes they know it. They are simply more powerful.

We do not need to reach back four centuries for examples. When scientists first suggested that Covid-19 might have originated in a laboratory, they were dismissed as conspiracy theorists, denied publication, and labelled purveyors of misinformation. Social media platforms removed their posts. Journalists who took the hypothesis seriously were accused of racism. Yet, by 2023, the FBI and the US Department of Energy had assessed it as the most likely origin.

Philosopher Karl Popper warned that scientific knowledge advances not by confirming what we believe but by subjecting our beliefs to the most rigorous challenge available. The platforms that removed posts about the lab leak hypothesis were doing the opposite: treating a testable scientific proposition as too dangerous to test.

The case for free speech begins here. Not with arguments about the right to offend, but with a darker observation: the people who hold the power to suppress are wrong often enough that society cannot afford to let them use it.

Beyond Natural Rights

Most defences of free speech rest on claims about rights. The tradition runs from Milton through Mill to the First Amendment: speech is a natural liberty, prior to government, which the state may not infringe. On this view, the case is closed before it begins – free expression is simply what we are owed as human beings.

This is stirring but ultimately unsatisfying. Rights-based arguments assert rather than argue. They tell us that free speech matters, not why it matters. And when rights conflict – when one person's expression threatens another's life, liberty, or good name – natural rights provide no mechanism for resolution. The assertion meets a counter-assertion, and we are left with a collision of intuitions.

A deeper problem is that contested claims require contested arguments. In a plural society, appeals to natural law persuade only those who already accept the premises. Those who believe community, tradition, or religious authority take precedence will not be moved by invocations of individual liberty. The argument needs firmer ground.

That ground is consequentialist. The case for free speech does not depend on metaphysical commitments about what humans are owed. It depends on what happens – demonstrably, historically, measurably – when societies protect open inquiry versus when they suppress it. The evidence is clear enough that it can persuade even those who reject the philosophical premises of individual rights.

This essay makes that case. Free speech is not defended here as a sacred right. It is defended as an institution that works – and works better than the alternatives.

What Free Speech Enables

The great puzzle of economic history is not why the Scientific Revolution and the Industrial Revolution happened in Europe. It is why they did not happen elsewhere.

China had gunpowder, the compass, movable type, and the most sophisticated civil service in the world. Rome built roads, aqueducts, and legal systems that endured for centuries. Mughal India produced architectural and administrative achievements unmatched in their era. The Maya developed writing, astronomy, and mathematics independently of the Old World. If technological sophistication or administrative capacity determined scientific and industrial take-off, any of these civilisations might have got there first.

None did. The economic historian Joel Mokyr argues in A Culture of Growth that the difference was institutional. These civilisations shared a common feature: centralised authority that could enforce orthodoxy. One buyer in the market for ideas. Heterodox views could be suppressed empire-wide. When the Emperor, or the Caliph, or the priesthood decided an idea was dangerous, it was dangerous everywhere.

Europe had the opposite problem: it could not suppress ideas even when it tried.

Political fragmentation meant no single authority could enforce orthodoxy continent-wide. When Galileo was silenced in Catholic Italy, his ideas circulated in Protestant Europe. When Spinoza could not publish in one jurisdiction, he found another. Heretical thinkers could always find sanctuary somewhere.

Mokyr calls this the market for ideas. In seventeenth- and eighteenth-century Europe, a transnational community of thinkers known as the Republic of Letters freely circulated discoveries and theories across borders. No king, no church, no academy could monopolise truth. Ideas competed, were tested, and the best survived. This institutional arrangement – not any inherent European genius – explains why the Scientific and Industrial Revolutions occurred in Europe rather than elsewhere.

Free speech does not guarantee that truth will prevail. What it guarantees is that falsehoods can be challenged. The value is not mystical. It is practical. Open societies develop faster feedback loops. They spot problems sooner. They try more solutions. They abandon failures faster. Silence does not prevent harm. It delays its discovery.

The pattern persists today. School closures during Covid were presented as following the science. Sweden was denounced as reckless for keeping schools open. The evidence now shows catastrophic learning loss, severe mental health damage, and minimal public health benefit – particularly for younger children who were never at serious risk. The people who questioned closures in 2020 were closer to the truth than the people who dismissed them. Wherever institutions have the power to silence critics, errors persist longer than they should.

What Silencing Costs

Suppression harms societies. It also harms the people it silences.

Think of the scientists who endured the suppression of the lab leak hypothesis – and said nothing. The standard account treats this as a failure of inquiry – a missed correction, a delayed discovery. That is true. But something else is also true: the scientists bore a personal cost that no improvement in social outcomes can fully compensate. They were denied the basic human capacity to speak their mind and have their views count.

As Nobel laureate Amartya Sen argued in Development as Freedom, political freedoms – the freedom to speak, to be heard, to participate in the debates that shape one’s community – are not merely instrumental to human development. They are constitutive of it. A life in which these capacities are foreclosed is a diminished life, regardless of whether the person is materially comfortable. Suppression is not merely an epistemic failure. It is a harm visited on persons.

The distinction matters. Speech restrictions do not only risk producing bad social outcomes. They also silence people: real individuals whose capacity to speak their convictions, contest authority, and participate in public life is curtailed by the law’s reach. The person who learns not to speak has lost something. Something no settlement, no apology, no policy reversal restores.

Sustained suppression does something worse still. It doesn't just silence people – it conditions them. The person who learns not to speak eventually loses the habit of forming thoughts worth saying.

Practical Limits

For all this, free speech has never been absolute. The question is where to draw the line – and how to tell whether the line is drawn well.

The common law developed three categories of speech restriction, each protecting a specific interest against a specific action.

Defamation protects reputation against damaging statements of fact or opinion about an identifiable person. Truth is a complete defence for statements of fact; for opinion, the defence requires only that the opinion be genuinely held. Incitement protects physical safety against speech that directly causes imminent violence. The US Supreme Court drew this line in Brandenburg v. Ohio: abstract advocacy is protected; direct incitement to imminent lawless action is not. Fraud protects economic interests against deliberate deception. The listener relies on a false statement and suffers material loss.

What unites these categories is that they address setbacks to interests – resources, opportunities, capacities – not offence or distress. As legal philosopher Joel Feinberg, one of the leading liberal theorists of the harm principle, observed: offence is a transitory unpleasant state that does not impede interests and is not harm in the relevant sense. The victim is identifiable. The damage is traceable to a false statement or harmful act. Courts adjudicate against objective standards, not subjective feelings. That is why they work – and why liberal democracies spent much of the twentieth century abolishing blasphemy and sedition laws that didn’t.

Modern hate speech laws risk making the same mistake. They treat offence as harm, viewpoints as conduct, and discomfort as danger.

Three problems follow. The first is who decides? Every speech restriction creates a new form of power – the power to define the boundary. That power attracts capture. Once restrictions exist, they are routinely interpreted expansively – from incitement to insult to merely normalising attitudes. They are enforced asymmetrically – the powerful use them against the powerless more often than the reverse.

Former American Civil Liberties Union president Nadine Strossen has documented case after case of dissidents and minority advocates prosecuted under laws designed to protect them. The European Commission against Racism and Intolerance – the body responsible for monitoring hate speech laws across Europe – has warned that these laws risk being used to silence minorities and suppress political opposition, and has expressed concern that vulnerable groups may have been disproportionately prosecuted. The power to define “hate” does not stay where its creators intended.

The second problem is that invisible costs outweigh visible ones. The harm from offensive speech is usually local and visible: a person insulted, a group demeaned. The harm from suppressed speech is diffuse, cumulative, and invisible: the policy error unchallenged, the grievance that festered, the scientist who saw problems with the modelling but knew that saying so publicly would end grant funding and collegial standing.

Surveys consistently find large majorities self-censoring on contested topics. Specific cases reveal the cost: the Rotherham grooming scandal went uninvestigated for years while investigators feared being called racist. Suppression does not produce harmony. It produces brittleness – dissent driven underground, grievances radicalising in echo chambers, trust collapsing when the public sphere feels managed rather than authentic.

The third is that the evidence shows these laws fail. Anti-Defamation League surveys have repeatedly found higher levels of anti-Semitism in most European countries with hate speech laws than in the United States without them. The case most often invoked to justify these laws – Weimar Germany – makes the point even more sharply: it actually demonstrates the opposite. The Nazis were prosecuted repeatedly under group libel laws. They used the courtroom as a propaganda platform. Hitler shouted defiance from the dock, generating headlines that spread his message further than any rally could. The decisive problem was not speech, but the failure to confront violence and lawlessness. Courts punished words while tolerating paramilitary terror.

The pattern is not confined to Weimar. Researchers studying the hate speech conviction of Geert Wilders found that prosecution damaged support for the legal system and democracy among those who broadly shared his views – with effects accumulating across a nine-year panel study. Prosecution of hate speech, the authors concluded, risks damaging the democratic system it is designed to protect.

What does work? Experimental evidence shows that empathy-based counterspeech on social media can reduce hate speech: in a large Twitter field experiment, empathy-based messages made users more likely to delete hate speech and less likely to repeat it – though effect sizes are modest and the most prolific offenders prove harder to reach. The European Commission against Racism and Intolerance has likewise concluded that counterspeech and related measures are much more likely to prove effective than relying on criminal sanctions alone.

Engagement works. Suppression backfires.

Revealed Preference

There is a simpler test than academic studies. Where do people go when they can choose?

Freedom House data shows that the vast majority of asylum seekers come from countries rated Not Free or Partly Free. People flee toward societies with robust speech protections, not away from them.

Migration correlates with poverty, of course. But the clearest cases run the other way. After Hong Kong’s National Security Law criminalised political expression in 2020, hundreds of thousands left one of the world’s wealthiest cities for the United Kingdom, Canada, and Australia – accepting lower incomes in exchange for the freedom to speak. The economic logic ran in reverse.

The migration patterns of the twenty-first century are an emphatic verdict on which institutional arrangements people actually prefer. The theorists who advocate speech restrictions will never bear the costs of getting it wrong. The migrants who choose free societies are betting their lives and their children’s futures. Revealed preference on this scale is not easily dismissed.

The Question Remains Live Here

This is not abstract. The question remains live here.

After the Christchurch attack, successive Justice Ministers ultimately resisted the pressure to legislate. Kiri Allan made the decisive call in 2022, concluding that the reforms would be “more corrosive than protective.” The current coalition government went further, ruling out hate speech legislation altogether – it is explicit in the coalition agreement.

Then came Bondi. Following the massacre of fifteen Jewish Australians at a Hanukkah celebration in December 2025, Australia passed new hate speech legislation within five weeks. Those laws are already being cited as a model for New Zealand – and a general election later this year could return a Labour government with unfinished business on this issue.

The problem is not that harassment and violence motivated by hatred are trivial – they are not, and the criminal law already addresses them regardless of motivation. The problem is that hate speech laws go further: they target expression, not conduct, and consistently fail that test.

New Zealand faces a particular risk. We are a small society that prizes consensus and deference to expertise – virtues in many contexts, but vices when they suppress the diversity of viewpoint that complex problems require. The pressure to agree can become the power to silence. And enforced consensus does not produce social cohesion. It produces the brittleness that shatters when tested.

The Christchurch attacker was radicalised online, in spaces beyond New Zealand law. Domestic restrictions would not have made New Zealanders any safer.

The Function of Freedom

What, then, is free speech for?

Not for the pleasure of giving offence. Not because all opinions are equally valid. Free speech is the error-correction mechanism of an open society – and something more besides. It is the condition under which people can speak their minds, challenge authority, and participate in the life of their community. Both matter.

Every expansion of human freedom began as heresy. Religious toleration, the abolition of slavery, the rights of women, civil rights for racial minorities – all were championed by people whose speech was considered dangerous by the authorities of their time. We cannot know in advance which of today’s heresies will prove to be tomorrow’s common sense. That is precisely why we cannot afford to let the powerful decide which ideas may be heard.

The case for free speech is not sentimental. It is institutional. Societies that protect open inquiry outperform those that do not – scientifically, economically, and in the dignity they afford their members. The market for ideas is not a metaphor. It is the mechanism that made the modern world.

The confident censor is always certain they are protecting the public good. Galileo’s judges were sure. The Inquisition was sure. The platforms that suppressed the lab leak hypothesis were sure.

The pattern never changes. The powerful are certain, and the heretics are silenced – until, sometimes, the heretics turn out to be right.

An open society is not one that has found the truth. It is one that remains capable of correcting its mistakes. That is what free speech is for.

This essay is part of an ongoing series on liberalism, democracy, and the international order. Related writing in Persuasion, Quadrant and Quillette is collected here. Roger Partridge writes at Plain Thinking