Free Speech: The Arguments, the History, and Where Every Democracy Draws the Line

Follows: Propaganda, Misinformation & How to Resist It

In this deep dive
1. Before free speech: licensing, and Milton's answer
2. Mill and the case for letting people speak
3. The case for limits: harm, dignity and democratic competence
4. The United States: from "clear and present danger" to Brandenburg
5. Germany: dignity first, and a democracy that defends itself
6. Europe's middle path: Article 10 and the Strasbourg court
7. The new question: platforms and content moderation
8. The strongest case on each side

A note on approach: this piece describes arguments, laws and court decisions; it is not legal advice, and laws change. It doesn't take a side in the debate over where the line should fall. Democracies that disagree sharply about speech are all, by any reasonable measure, still democracies, and each side's position is given here in the form its defenders would recognise.

Three scenes, from three democracies.

Ohio. A Ku Klux Klan leader phones a Cincinnati television reporter and invites him to a rally on a farm. The film shows twelve hooded figures, some carrying guns, gathered around a burning cross, with snatches of racist abuse of Black and Jewish Americans. In his speech the leader warns that if the government keeps suppressing "the white, Caucasian race," there "might have to be some revengeance taken." He is convicted under a state law against advocating violence as a means of political change. In 1969 the US Supreme Court throws out the conviction, and with it the law.

Bavaria, 2005. A man plans an annual open-air march in the town of Wunsiedel "in commemoration of Rudolf Hess," Hitler's deputy. The authorities ban it. In 2009 Germany's Federal Constitutional Court upholds the law behind the ban, even while admitting that the law singles out one ideology, which German constitutional law normally forbids.

Denmark. A journalist makes a documentary about a group of young racists and broadcasts their abusive remarks about immigrants. He is convicted of helping spread racist statements. In 1994 the European Court of Human Rights rules that his conviction violated his freedom of expression, because the programme as a whole aimed to inform the public, not to spread the views.

Different facts, different courts, different outcomes. But each answer comes from a coherent theory of what free speech is for. Propaganda, Misinformation & How to Resist It ended on the dilemma of who gets to decide what's false. This piece goes deeper into the older and larger question underneath it: why protect speech at all, why limit it, and how democracies have answered differently.

Part 1 — Before free speech: licensing, and Milton's answer

For most of history, the question wasn't whether speech should be limited but who should do the limiting. After the printing press arrived in Europe, governments and churches answered with licensing: nothing could be printed without prior approval from an official censor.

England's Parliament, at war with its king in the 1640s, kept that system. Its Licensing Order of 1643 required government approval before any book could be published. The poet John Milton answered on 23 November 1644 with Areopagitica, subtitled "A Speech for the Liberty of Unlicensed Printing to the Parliament of England." It was never delivered as a speech. It was printed as a pamphlet, without a licence.

Milton's arguments set the template for centuries:

• Truth is strong enough. "Let her and Falsehood grapple; who ever knew Truth put to the worse, in a free and open encounter?" Suppression, he says, insults truth by doubting her strength.
• Virtue needs trial. Milton rejects "a fugitive and cloistered virtue, unexercised and unbreathed," that has never met its adversary. Goodness that has never been tested by temptation, or by bad arguments, isn't worth much.
• Books are alive. "As good almost kill a man as kill a good book," he writes. Whoever "destroys a good book, kills reason itself."
• The deepest liberty. "Give me the liberty to know, to utter, and to argue freely according to conscience, above all liberties."

Milton's most famous line about free speech comes with an exception attached, and the exception tells you as much as the line.

Milton was not a modern free-speech absolutist, and he said so. His target was censorship before publication. Once a book was out, he accepted that the state should keep "a vigilant eye how books demean themselves," and could "confine, imprison, and do sharpest justice on them as malefactors." And his tolerance had a hole in the middle: "I mean not tolerated popery, and open superstition," which, he argued, "should be extirpate." Catholics, in other words, were outside the circle.

That's not a footnote. It shows a pattern that runs through the whole history of free speech: nearly everyone who defends it draws a line somewhere, and the line usually falls just short of the views the defender finds most threatening. The question is never really "limits or no limits." It's which limits, justified how, and decided by whom.

Milton's distinction between stopping words before they appear and punishing them afterwards turned out to be one of the most durable ideas in free-speech law. Censorship in advance (lawyers call it prior restraint) is treated as especially dangerous almost everywhere, because it silences speech before anyone can judge it in public, and because a censor's office has every incentive to say no. You can see the idea in all three systems this piece compares. Germany's Basic Law says flatly, "There shall be no censorship." The US Supreme Court's Skokie decision (Part 4) insisted that a state restraining a march in advance must provide immediate appeal. And the European Convention's Article 10 mentions licensing only to permit it for "broadcasting, television or cinema enterprises," the scarce airwaves of the twentieth century. Punishment after the fact, for defamation, incitement or fraud, is where nearly all modern democracies do their line-drawing.

Areopagitica changed nothing at the time; it did not persuade the majority in Parliament. Licensing continued. The system lapsed only in 1695, when Parliament declined to renew it. But Milton's arguments outlived the law he attacked, and were picked up two centuries later by the most influential defender of free speech in English.

Part 2 — Mill and the case for letting people speak

John Stuart Mill's On Liberty (1859) gives its whole second chapter to "the liberty of thought and discussion," and it contains the four arguments still at the core of the case for free speech. Mill summarises them himself at the end of the chapter:

Mill's most uncompromising sentence is this: "If all mankind minus one, were of one opinion, and only one person were of the contrary opinion, mankind would be no more justified in silencing that one person, than he, if he had the power, would be justified in silencing mankind." The wrong of silencing, he says, is "robbing the human race," the people who disagree with the silenced view most of all.

And his best advice to individuals has outlived his political theory: "He who knows only his own side of the case, knows little of that."

Where Mill himself drew the line

Mill wasn't an absolutist either. At the start of chapter 3 he gives the example that still anchors debates about incitement: the opinion "that corn-dealers are starvers of the poor" should be left alone when "simply circulated through the press," but "may justly incur punishment when delivered orally to an excited mob assembled before the house of a corn-dealer." Same words; different context; different answer. What matters is whether the speech, in its setting, becomes "a positive instigation to some mischievous act."

The marketplace of ideas

In the United States, Mill's argument from truth became an economic metaphor. In 1919 Justice Oliver Wendell Holmes, dissenting in Abrams v. United States, wrote that "the ultimate good desired is better reached by free trade in ideas—that the best test of truth is the power of the thought to get itself accepted in the competition of the market." The image became known as the marketplace of ideas.

Eight years later, in Whitney v. California (1927), Justice Louis Brandeis added its most quoted corollary: if there is time "to expose through discussion the falsehood and fallacies," then "the remedy to be applied is more speech, not enforced silence. Only an emergency can justify repression."

Two more arguments for free speech

Truth isn't the only reason people defend speech. Two others are equally influential, as the Stanford Encyclopedia of Philosophy's entry on freedom of speech explains:

• Self-government. The American philosopher Alexander Meiklejohn argued that free speech is not just useful to democracy but part of what democracy is: a people can't govern itself if it can't hear every argument about how to do so. John Rawls later wrote that restricting free political speech "always implies at least a partial suspension of democracy."
• Autonomy. Thomas Scanlon (1972) argued that a legitimate state must treat citizens as capable of making up their own minds. Banning speech because officials fear citizens will be persuaded by it treats them as children.

To these, modern defenders add a fourth, practical argument associated with Frederick Schauer: distrust of government. Even well-justified limits create precedents and tools that later governments, with worse intentions, will inherit. Vague laws chill speech that was never meant to be caught.

The case for free speech isn't one argument. It's four: truth, self-government, autonomy, and the suspicion that whoever holds the censor's pen today won't be the one holding it tomorrow.

Part 3 — The case for limits: harm, dignity and democratic competence

Every democracy limits some speech: fraud, perjury, threats, defamation, incitement to violence. The real disagreement is about speech that is hateful, false or anti-democratic but doesn't directly cause a specific injury. Here the case for limits rests on three main arguments, each at its strongest.

1. The harm argument

This is Mill's own exception, extended. If speech in context is likely to cause violence, it may be restricted. The disagreement is over how direct the link must be. Those who favour wider limits argue that the corn-dealer's mob doesn't have to assemble in one place anymore: a campaign of dehumanising speech, repeated across millions of screens, can make violence against a group more likely without any single post being "incitement." International human-rights law reflects this concern. The International Covenant on Civil and Political Rights (Article 20) requires that "any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law."

2. The dignity argument

The legal philosopher Jeremy Waldron, in The Harm in Hate Speech (2012), argued that hate speech is best understood as group defamation. Its harm isn't mainly that it offends people's feelings. It's that it attacks their public standing, "undermining their standing as respected equals" in their community. A decent society, on this view, is one in which every member can go about public life confident of being treated as an equal. Posters, leaflets and broadcasts declaring that some group doesn't belong erode that confidence, and the burden falls on the people targeted, who have to live in a public space that tells them they're less than others.

Related arguments from feminist and critical legal scholars add that some speech silences. Catharine MacKinnon argued that pornography alters how women's speech is heard, including the force of a woman's "no." More recent philosophers make a parallel point about intimidating hate speech: if it drives people out of public debate, then permitting it can reduce the total amount of free expression rather than increase it.

3. The democratic-competence argument

The third argument turns the democracy case for free speech around. Democracy needs more than a right for everyone to speak. It also needs citizens who can get reliable knowledge and a political order that survives. The legal scholar Robert Post has distinguished speech that gives democracy its legitimacy (everyone's right to take part in public debate) from the conditions of democratic competence: a public that can rely on expert knowledge requires institutions such as science and professional standards, which constantly judge some claims better than others.

A harder version of the argument is about survival. In 1937, watching democracies fall to movements that used democratic freedoms to destroy them, the German émigré scholar Karl Loewenstein argued in the American Political Science Review for militant democracy: democracies may legitimately restrict the freedoms of movements that aim to abolish democracy itself. This idea runs straight into the German Basic Law (Part 5).

The critique of the marketplace

Defenders of limits also question the marketplace metaphor. The Stanford Encyclopedia notes that few contemporary theorists hold the strong version, the claim that truth reliably wins out in open debate, and that critics draw on behavioural economics and social psychology to argue that human reasoning is badly suited to it. Propaganda, Misinformation & How to Resist It described one piece of that evidence: a 2018 study of Twitter found false stories spread faster and further than true ones. A market, the argument goes, can fail.

The case for limits isn't "some speech is offensive." It's that some speech injures people's standing as equals, damages the conditions for knowledge, or threatens the democracy that protects everyone's speech.

Defenders of broad free speech have answers to each, which Part 8 sets out. But it's worth noticing that all three arguments for limits are made by people who value free expression and believe they are protecting it.

Part 4 — The United States: from "clear and present danger" to Brandenburg

The First Amendment says "Congress shall make no law … abridging the freedom of speech, or of the press." The modern American approach to those words was built mostly in the twentieth century, through cases about dissenters, radicals and people almost everyone found repellent.

1919, Schenck v. United States. During the First World War, Charles Schenck, general secretary of the Socialist Party, sent leaflets urging resistance to the military draft. The Supreme Court upheld his conviction. Justice Holmes wrote the famous line that "the most stringent protection of free speech would not protect a man in falsely shouting fire in a theatre and causing a panic," and set out the test: whether words "create a clear and present danger" of evils Congress may prevent. The theatre line, endlessly quoted to justify limits, came in a case that sent a man to prison for an anti-war leaflet, and the test it supported was later abandoned.

1919, Abrams v. United States. Months later, Holmes changed course, dissenting from the conviction of leaflet-writers and writing the "free trade in ideas" passage above.

1964, New York Times v. Sullivan. An elected city commissioner in Montgomery, Alabama, sued the newspaper for libel over a full-page advertisement, "Heed Their Rising Voices," that described the nonviolent civil-rights demonstrations of Southern Black students and the official response to them. The Court ruled that a public official can win a libel suit only by proving "actual malice," meaning "knowledge that it was false or with reckless disregard of whether it was false or not." It grounded this in "a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open."

1969, Brandenburg v. Ohio. The Klan case from the opening. The Court held that the government may not forbid "advocacy of the use of force or of law violation except where such advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action." It noted that its 1927 Whitney decision, which had upheld a similar law, "has been thoroughly discredited," and overruled it.

The Brandenburg test is the heart of the American approach. To be punished, speech advocating violence must be (1) intended to produce lawless action, (2) aimed at imminent action, and (3) likely to produce it. Abstract advocacy, even of revolution, is protected.

1977, National Socialist Party of America v. Village of Skokie. When a neo-Nazi group planned to march in the village of Skokie, Illinois, a local court banned the march. The Supreme Court ruled that a state imposing that kind of restraint must provide "strict procedural safeguards," including immediate appellate review, or else allow the march to go ahead while the appeal is heard. The case became a symbol of the principle that the government may not decide which viewpoints are too hateful to express.

2011, Snyder v. Phelps. Members of a church picketed near a soldier's funeral with signs saying that God was punishing the United States. The soldier's father sued, and a jury awarded millions in damages. An appeals court overturned the verdict, and the Supreme Court agreed, eight justices to one. Speech, Chief Justice Roberts wrote, can "inflict great pain," but "we cannot react to that pain by punishing the speaker," because the nation has chosen "to protect even hurtful speech on public issues to ensure that we do not stifle public debate."

America's position is unusual in the world, and it knows it. When the United States ratified the International Covenant on Civil and Political Rights in 1992, the Senate attached a reservation saying that Article 20, the clause requiring bans on advocacy of hatred, "does not authorize or require legislation or other action by the United States that would restrict the right of free speech." The Stanford Encyclopedia describes the American position on hate speech as one "unshared by virtually all other liberal democracies."

Part 5 — Germany: dignity first, and a democracy that defends itself

Germany's Basic Law of 1949 was written in the shadow of the Weimar Republic's collapse (see How Democracies Erode, and How They Hold). Its first sentence is not about freedom. It is about dignity: "Human dignity shall be inviolable. To respect and protect it shall be the duty of all state authority" (Article 1).

Free expression comes in Article 5, in strong terms: "Every person shall have the right freely to express and disseminate his opinions," and "There shall be no censorship." But paragraph 2 sets limits: these rights "find their limits in the provisions of general laws, in provisions for the protection of young persons and in the right to personal honour."

The Basic Law also builds in Loewenstein's militant democracy. Article 18 says anyone who abuses the freedom of expression and other listed rights "in order to combat the free democratic basic order shall forfeit these basic rights," a forfeiture only the Constitutional Court can declare. Article 21 allows the court to declare unconstitutional any party that seeks "to undermine or abolish the free democratic basic order."

The criminal law follows. Section 130 of the Criminal Code, Volksverhetzung (usually translated as "incitement of the masses" or "incitement to hatred"), punishes anyone who, "in a manner suited to causing a disturbance of the public peace," incites hatred against national, racial, religious or ethnic groups, calls for violent or arbitrary measures against them, or attacks their human dignity by insulting or defaming them. The penalty is three months to five years in prison. Further subsections punish, among other things:

• publicly approving, denying or downplaying genocide committed under National Socialism, in a manner suited to disturbing the public peace (up to five years);
• publicly approving, glorifying or justifying National Socialist rule in a way that violates the dignity of its victims (up to three years);
• approving, denying or grossly downplaying other genocides and war crimes in a way suited to inciting hatred or violence.

The Wunsiedel decision

The 2009 Wunsiedel case, from the opening, shows how seriously German law takes the tension. Normally a law limiting opinion must be a "general law," not aimed at a particular viewpoint. The ban on glorifying Nazi rule plainly is aimed at one. The Constitutional Court admitted this: the provision "is not a general law." It upheld it anyway, "by way of exception," reasoning that the Basic Law "can virtually be regarded as a counter-concept to the totalitarianism of the National Socialist regime," and that a limited exception for propaganda approving that historical regime is built into the constitution itself.

Notice what else the court said. Even the spread of National Socialist ideas "does not from the outset fall outside the area of protection of the freedom of opinion," and the free order of the Basic Law "primarily assigns the task of counteracting the associated dangers to civic commitment in free political discourse." In other words, even Germany's highest court treats its exception as an exception, justified by one specific history.

The American rule asks: will this speech cause imminent violence? The German rule also asks: does it attack the dignity of the people it targets, or the democratic order itself?

The strongest case for the German approach, as its defenders put it: a democracy that has been destroyed once by a movement exploiting free expression has earned the right to refuse that movement a second chance. Protecting dignity protects the equal standing on which everyone's free speech depends.

The strongest criticism: deciding which ideas threaten the democratic order hands officials and courts exactly the power free speech exists to deny them, and the categories can expand over time beyond their original target.

Part 6 — Europe's middle path: Article 10 and the Strasbourg court

The European Convention on Human Rights, signed in Rome on 4 November 1950, binds the member states of the Council of Europe, which include nearly every European country, and is enforced by the European Court of Human Rights in Strasbourg. Its Article 10 protects "freedom to hold opinions and to receive and impart information and ideas without interference by public authority and regardless of frontiers."

Then comes the crucial second paragraph. Because the exercise of these freedoms "carries with it duties and responsibilities," it may be subject to restrictions that are "prescribed by law and are necessary in a democratic society," for a list of purposes including national security, public safety, preventing disorder or crime, protecting health or morals, and protecting "the reputation or rights of others."

That wording produces a three-step test the court applies to every restriction:

1. Prescribed by law: is the restriction set out in a law clear enough for people to know what's forbidden?
2. Legitimate aim: does it pursue one of the purposes listed in Article 10(2)?
3. Necessary in a democratic society: does it answer a "pressing social need," and is it proportionate?

Within that framework, the court has said some things that sound almost American. In Handyside v. United Kingdom (1976), it held that Article 10 applies not only to ideas "favourably received or regarded as inoffensive" but also to those "that offend, shock or disturb; such are the demands of that pluralism, tolerance and broadmindedness without which there is no 'democratic society.'" It has held that there is "little scope" for restricting political speech or debate on questions of public interest.

But it has also built a firm floor under its protection. Article 17 of the Convention says nothing in it may be read as giving anyone a right to engage in activity "aimed at the destruction of any of the rights and freedoms set forth herein." The court uses this in the most extreme cases. In Garaudy v. France (2003), a Holocaust denier's complaint was thrown out entirely: denying clearly established crimes against humanity, the court said, was not historical research but one of "the most serious forms of racial defamation of Jews."

Three cases show how finely the court draws its lines:

• Jersild v. Denmark (1994): the journalist from the opening. His conviction violated Article 10, because the documentary aimed to inform the public about racism, not to promote it.
• Perinçek v. Switzerland (2015, Grand Chamber): a Turkish politician was convicted in Switzerland for saying that the mass deportations and massacres of Armenians in 1915 did not amount to genocide. The court found his conviction violated Article 10, balancing his speech against the dignity of the victims and of Armenians today, which it held is protected by Article 8.
• Sanchez v. France (2023, Grand Chamber): a local politician standing for parliament was convicted for failing to delete hateful comments posted by others on his Facebook wall. The court found no violation: as a politician, he bore responsibility for unlawful content he left up.

Part 7 — The new question: platforms and content moderation

All of this law was built for a world in which the main threat to speech was the state, and the main venues were newspapers, broadcasters and public squares. Today, much public speech happens on platforms owned by a handful of private companies, which remove, demote, label and recommend billions of posts under their own rules. That raises questions the old frameworks weren't designed for.

The American answer: immunity and editorial freedom

In 1996 the US Congress passed what is now Section 230 of the Communications Decency Act. Its key sentence reads: "No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider." A second clause protects platforms from liability for good-faith removal of content they consider "obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected."

The effect is double: platforms generally can't be sued for what users post, and can't be sued for taking it down. Since the First Amendment restrains governments, not private companies, platforms can moderate largely as they choose. In Moody v. NetChoice (decided 1 July 2024), concerning 2021 laws in two US states that restricted large platforms' moderation, the Supreme Court sent the cases back to the lower courts but explained that when platforms curate their feeds, they are engaged in their own expressive activity, comparable to a newspaper's editorial judgment, which the government can't simply override.

The European answer: duties and procedures

Europe took the opposite path: impose obligations. Germany's Network Enforcement Act (NetzDG) of 2017 required large social networks to deal quickly with complaints about unlawful content, backed by fines. Its main duties for platforms were repealed in May 2024, after EU-wide rules began to apply. The European Union then adopted the Digital Services Act (DSA), which entered into force in November 2022 and applied to all regulated services from 17 February 2024. Among other things, it:

• requires platforms to let users flag illegal content and to respond, and to explain their removal decisions;
• gives users the right to appeal, inside the platform or to an out-of-court dispute body;
• requires "very large" platforms (over 45 million monthly users in the EU) to assess and reduce systemic risks;
• does not impose a general obligation to monitor everything users post (Article 8);
• allows fines of up to 6 percent of a provider's global annual turnover.

What counts as "illegal content" isn't defined by the DSA itself: it depends on EU and national law, so something illegal only in one country generally needs to be removed only there.

The European Court of Human Rights has also addressed the issue. In Delfi AS v. Estonia (2015), its first case on liability for user comments, it found no violation of Article 10 when a commercial news portal was held liable for clearly unlawful comments under one of its articles, which it removed only about six weeks later.

Private governance

Some platforms have built quasi-judicial systems of their own. Meta's Oversight Board, for example, reviews content decisions on Facebook, Instagram and Threads; its case decisions are binding on the company "unless implementing them could violate the law."

The questions underneath

The legal scholars the Stanford Encyclopedia cites frame the new debate in three ways. Jack Balkin (2018) argues that free speech is now a triangle, with governments, platforms and speakers, and that governments increasingly regulate speech by pressuring platforms rather than speakers. Evelyn Douek argues that courts' case-by-case model of speech decisions doesn't suit moderation at the scale platforms operate, which has to be governed more like a system. And some philosophers argue that the largest platforms "have become public fora" and should be held to free-speech duties like a government.

The old question was what the state may forbid. The new question is what private companies should do, and how much the state should tell them.

The strongest case for leaving platforms free: they are private publishers with their own expressive rights; forcing them to carry or remove speech means the government choosing what's said; and immunity is what allowed open platforms to exist at all.

The strongest case for regulating them: a few companies now control a share of public debate no newspaper ever did, often by opaque algorithms tuned for engagement; procedural duties, transparency and appeal rights are the minimum that such power owes the public; and leaving it unregulated just means rules are written privately instead of democratically.

Part 8 — The strongest case on each side

Here are the two broad positions, each in its own best terms, followed by the questions still genuinely open.

The case for broad protection (closest to the American approach), as its defenders would make it:

• Fallibility. Every censor in history was sure of their judgment. Many views once suppressed as dangerous, from religious dissent to arguments for equal rights, are now accepted. We can't know which of today's heresies are tomorrow's truths.
• The power problem. Whoever writes the speech rules writes them for everyone, including future governments with worse intentions. Hate-speech laws drafted against extremists have, in many places, been used against minorities, journalists and dissidents.
• Counter-speech works better. Driving ugly views underground doesn't end them; it makes martyrs and hides them from rebuttal. Brandeis's "more speech" remedy keeps arguments where they can be answered.
• Equality cuts both ways. Strong protection has historically sheltered the speech of movements for the powerless, such as the civil-rights advertisement in Sullivan and anti-war leaflets, as much as the speech of the powerful.

The case for dignity-based limits (closest to the German and wider European approach), as its defenders would make it:

• History. Democracies have been destroyed by movements that used free expression to dehumanise minorities and then abolish freedom itself. Learning from that is not paranoia.
• Real injury. Hate speech harms the people it targets in their standing as equal citizens, not just their feelings, and can drive them out of public life. Free speech that silences others isn't neutral.
• Markets fail. The idea that truth wins in open competition is an act of faith that the evidence on misinformation doesn't support. Some floor of protection for dignity and truth makes the rest of the debate possible.
• It works without tyranny. Countries with hate-speech laws remain vigorous democracies with free presses, independent courts and open debate. The slippery slope is a prediction, and in these countries it largely hasn't come true.

What both sides share

It's easy to miss how much common ground there is. Every system described here protects harsh criticism of the government, which is the core of political freedom and the first thing authoritarian governments attack. Every one punishes true threats, fraud and direct incitement to violence. Every one treats prior censorship as especially suspect. And every one relies on independent courts to decide hard cases, rather than on the government of the day. The real disagreement lives in a band between those shared points: hateful speech, denial of historical atrocities, anti-democratic advocacy and, increasingly, the rules platforms apply. That band matters a great deal to the people targeted by such speech and to the people whose speech is restricted. But it is a disagreement among democracies about how to protect democracy, not between democracy and its opposite.

What's still open

• Where is the line between hate and opinion? Every system, including the American, has to draw one somewhere (true threats, harassment, incitement). The disagreement is about where, and who decides.
• Does the American model fit a world of platforms? Brandenburg's "imminent" requirement was written for a speaker and a crowd. Whether it makes sense for speech that reaches millions instantly and builds up over months is a real question.
• Does the European model travel? Laws designed with independent courts and a free press in mind can work very differently in countries without them.
• Who should moderate the moderators? Governments, companies, independent boards, users, or some mix. Nobody has a settled answer.

The takeaway

The case for free speech rests on four arguments: that we might be wrong (Milton and Mill), that self-government needs open debate (Meiklejohn), that citizens deserve to be treated as able to judge for themselves (Scanlon), and that no one should trust the censor's pen to future governments (Schauer). The case for limits rests on harm (incitement, as in Mill's corn-dealer), dignity (Waldron's group defamation), and democratic competence and survival (Post, Loewenstein's militant democracy). The United States protects even hateful advocacy unless it's intended, likely and directed to imminent lawless action (Brandenburg, 1969). Germany puts dignity first and punishes incitement to hatred and Holocaust denial (§130). The European Court protects speech that "offends, shocks or disturbs" but lets states impose necessary, proportionate limits and excludes the most extreme cases altogether. Platforms have turned an old question about state power into a new one about private power, and democracies are answering it in opposite ways.

Try this: take one piece of speech you strongly object to and run it through all three systems in the comparison table. Would it be protected in the United States? In Germany? In Strasbourg? Then do the same with a piece of speech you strongly agree with that others find offensive. Notice which test you'd want applied in both cases, and whether it's the same one.

Sources & Further Reading

Sources
- John Milton, Areopagitica (23 November 1644), Project Gutenberg edition; Licensing Order of 1643 and its lapse in 1695 as summarised in standard references.
- John Stuart Mill, On Liberty (1859), chs. 2–3, Project Gutenberg edition.
- First Amendment text (Cornell LII). US Supreme Court opinions (Legal Information Institute, Cornell): Schenck v. United States, 249 U.S. 47 (1919); Abrams v. United States, 250 U.S. 616 (1919), Holmes dissenting; Whitney v. California, 274 U.S. 357 (1927), Brandeis concurring; New York Times v. Sullivan, 376 U.S. 254 (1964); Brandenburg v. Ohio, 395 U.S. 444 (1969); National Socialist Party of America v. Village of Skokie, 432 U.S. 43 (1977); Snyder v. Phelps, 562 U.S. 443 (2011); Moody v. NetChoice, No. 22–277 (2024).
- 47 U.S.C. § 230 (Cornell LII); US Senate reservations to the ICCPR, 138 Cong. Rec. S4781 (2 April 1992).
- Basic Law for the Federal Republic of Germany, Arts. 1, 5, 18, 21, and German Criminal Code § 130, official English translations (gesetze-im-internet.de); Federal Constitutional Court, Order of 4 November 2009, 1 BvR 2150/08 (Wunsiedel), press release no. 129/2009. Netzwerkdurchsetzungsgesetz of 1 September 2017, as amended 6 May 2024 (§§2–3f repealed).
- European Convention on Human Rights, Arts. 10 and 17; European Court of Human Rights, Guide on Article 10 (updated 28 February 2026) and Factsheet – Hate speech (November 2023): Handyside (1976), Lingens (1986), Jersild (1994), Garaudy (2003), Delfi AS (2015), Perinçek (2015), Sanchez (2023).
- International Covenant on Civil and Political Rights, Arts. 19–20.
- European Commission, Digital Services Act Q&A (entry into force November 2022; application from 17 February 2024; fines up to 6% of turnover; no general monitoring obligation; 45 million user threshold).
- Stanford Encyclopedia of Philosophy, "Freedom of Speech" (Meiklejohn, Rawls, Scanlon, Schauer, Waldron, MacKinnon, Balkin, Douek, the critique of the marketplace of ideas, the US position on hate speech).
- Jeremy Waldron, The Harm in Hate Speech (Harvard, 2012); Robert Post, Democracy, Expertise, and Academic Freedom (Yale), ch. 2, "Democratic Competence and the First Amendment"; Karl Loewenstein, "Militant Democracy and Fundamental Rights," American Political Science Review 31 (1937). Oversight Board, oversightboard.com.

Go deeper
- 🌐 Project Gutenberg — Areopagitica by John Milton (short and rousing; read it alongside his exceptions)
- 🌐 Stanford Encyclopedia of Philosophy — Freedom of Speech (the best free map of every argument in this piece, with their critics)
- 🌐 Legal Information Institute — Brandenburg v. Ohio (1969) (the short opinion that defines the American approach)
- 🌐 European Court of Human Rights — Guide on Article 10 (the court's own summary of its case law; long, but searchable)
- 🌐 European Court of Human Rights — Factsheet on hate speech (dozens of real cases in a few pages each)
- 📘 The Harm in Hate Speech — Jeremy Waldron (the strongest dignity-based case for limits)
- 📘 HATE: Why We Should Resist It with Free Speech, Not Censorship — Nadine Strossen (the strongest counter-case, by a former president of the American Civil Liberties Union)
- 🌐 Companion pieces in this library: Propaganda, Misinformation & How to Resist It, What Is Freedom? and Rights, Constitutions & the Rule of Law