The First Amendment · The State-Action Line · The Pandemic Files · Article 10 · The Basic Law · The Funding Lever
Free Speech: What the Record Shows About What the First Amendment Protects, and What Other Democracies Do Instead
United States law protects speech that is hateful, offensive, and Nazi in content, and forbids the government to act against it directly, indirectly, or through pressure on the private companies that carry it; the record runs from the Sedition Act of 1798 to Brandenburg, to the Murthy litigation over pandemic-era pressure on platforms, to a September 2025 letter in which Alphabet told Congress that administration officials pressed it to remove content that broke none of its rules. The other democracies write the limit into the right itself: Britain recorded 12,183 arrests in 2023 under two communications statutes and jailed a woman for 31 months for one post; Germany's Basic Law subordinates expression to the general laws, bans whole categories of symbols, and sent police to search a pensioner's home over a meme; Canada sustains a hate-propaganda law under its Charter's limits clause. In November 2026 the American line is being tested at Rutgers, where a public university has declined to cancel a former Border Patrol commander's speech, and in the federal courts, where the funding lever the government used against Harvard has been held unconstitutional retaliation and the appeal is pending.
Two documents describe the American settlement on speech, and both are short. The first is the First Amendment, ratified December 15, 1791: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press.” The second is the sentence with which Justice Alito closed the Supreme Court's unanimous judgment in Matal v. Tam in 2017: “Speech that demeans on the basis of race, ethnicity, gender, religion, age, disability, or any other similar ground is hateful; but the proudest boast of our free speech jurisprudence is that we protect the freedom to express ‘the thought that we hate.’” Between the text and that sentence stands the whole of the record this report prints: what the amendment forbids, what it permits, the narrow categories it leaves outside, the line it draws at the government's own hand—including the hand applied indirectly, through threats, through pressure on private platforms, and through money—and the contrasting settlements of the other democracies, in which expression is a qualified right, balanced in the text itself against dignity, reputation, and public order, and enforced against posts, memes, and T-shirts by arrest and imprisonment. The comparison is between two legal architectures: one in which the limit sits outside the right, in a short list of harms the courts police case by case, and one in which the limit is written into the right's own text, where legislatures and police apply it in advance.
The spine: from the Sedition Act to Brandenburg
The American law of speech began with a prosecution statute. The Sedition Act of 1798 made it a crime to publish “false, scandalous and malicious writing” against the government, Congress, or the President; some twenty-five people were arrested under it and ten convicted, including a sitting congressman, before the act expired on March 3, 1801, and President Jefferson pardoned those convicted. No court struck it down; its repudiation came through elections and pardons, and the Supreme Court noted in 1964 that the attack upon its validity “has carried the day in the court of history.” The modern doctrine arrived in stages. In Schenck v. United States (1919), Justice Holmes upheld Espionage Act convictions and gave the law its first test—whether words create a “clear and present danger” of evils Congress may prevent—and eight months later, dissenting in Abrams v. United States, he stated the theory the test would eventually serve: “the best test of truth is the power of the thought to get itself accepted in the competition of the market.” In Gitlow v. New York (1925), the Court assumed that freedom of speech is among the liberties the Fourteenth Amendment protects against the states, binding Albany as the First Amendment binds Congress. The endpoint of the line came in Brandenburg v. Ohio (1969), which set the standard that still governs advocacy of force or illegality: the state may not forbid advocacy “except where such advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action.” A Ku Klux Klan leader's filmed threats of “revengeance” against the government did not meet it. The distance between 1798 and 1969 is the distance between punishing scandalous writing about the President and protecting a Klansman's rally speech; the doctrine traveled it in one direction only, each generation's test narrower than the last.
What the law does not protect
The unprotected categories are few, defined, and printed in the cases. Incitement is the Brandenburg category above: intent, imminence, and likelihood, all three required. True threats are outside the amendment; in Counterman v. Colorado (2023), the Court held that the state must prove at least recklessness—that the speaker “consciously disregarded a substantial risk” that the words would be taken as threatening violence—before a statement may be punished as a threat. Obscenity is governed by Miller v. California (1973) and its three-part test, a category confined to sexual material and separate from offensiveness as such. Defamation of a public official requires proof of “actual malice”—knowledge of falsity or reckless disregard for the truth—under New York Times Co. v. Sullivan (1964), the case that constitutionalized the law of libel out of a segregationist commissioner's suit over a civil-rights advertisement. Fraud and speech integral to criminal conduct—solicitation, conspiracy, perjury, extortion—are punishable as the conduct they are part of. Beyond that list, the Court has refused to go: in United States v. Stevens (2010) and Brown v. Entertainment Merchants Association (2011) it declined to create new unprotected categories for depictions of animal cruelty and for violent video games sold to minors, holding that the existing categories are a closed historical set, not a balancing test to be extended by analogy. The categories share a property the desk records because the comparison below turns on it: each is defined by a harm the speech does—imminent violence, a credible threat, a lie that damages—and none is defined by the offensiveness of what the speech says.
What it protects, hate included
There is no hate-speech exception in American law, and the record of its rejection is repeated and recent. In R.A.V. v. City of St. Paul (1992), the Court struck down an ordinance that punished cross-burning and other displays that arouse “anger, alarm or resentment” on the basis of race, color, creed, religion, or gender: even within the punishable class of fighting words, the government may not select the racist instances for special prohibition. In Snyder v. Phelps (2011), the Court held 8–1 that a fringe church's picketing of a Marine's funeral with signs reading “Thank God for Dead Soldiers” was speech on matters of public concern, protected against a tort judgment for emotional distress. In Matal v. Tam (2017), a unanimous Court struck down the federal trademark law's disparagement clause, holding that a statute punishing speech that demeans is viewpoint discrimination, and printing the sentence quoted in this report's first paragraph. The older test of the same principle is the Skokie litigation of 1977–78: when the National Socialist Party of America announced a march in Skokie, Illinois, a village with a large population of Holocaust survivors, the village enacted ordinances against military-style uniforms, racially slurring literature, and demonstrations promoting racial hatred. The Supreme Court's per curiam order in National Socialist Party of America v. Village of Skokie (1977) required immediate appellate review of the injunction against the march, and the Illinois courts struck the ordinances down; the party in the end marched in Chicago in June 1978. Nazi uniform, Nazi literature, and Nazi slogans, displayed by actual Nazis in a town of survivors, are the historical center of the American doctrine: the speech was protected not because any court valued it, but because the categories contain no entry for it.
The state-action line, and this week's test of it
The amendment binds the government. It does not bind a private platform, a private employer, or a counter-demonstrator, and the courts police the boundary in both directions: private suppression is lawful, government evasion of the boundary is not. The evasion doctrine is old. In Bantam Books, Inc. v. Sullivan (1963), the Court held that a state commission's letters to distributors—identifying objectionable books, reminding recipients of the commission's duty to recommend prosecutions, followed by police visits—were state action suppressing speech, although the commission had no power to ban anything itself. The Court restated the principle unanimously in National Rifle Association of America v. Vullo (2024): “Government officials cannot attempt to coerce private parties in order to punish or suppress views that the government disfavors,” and a financial regulator's pressure on insurers to drop a disfavored advocacy group states a First Amendment claim. On the boundary's other side stands the heckler's veto. In Terminiello v. Chicago (1949), the Court reversed the breach-of-peace conviction of a speaker whose address had drawn a violent crowd outside the hall: “a function of free speech under our system of government is to invite dispute.” In Forsyth County v. Nationalist Movement (1992), it struck down a permit fee set by officials according to the expected cost of controlling hostile listeners. The crowd's anger is not a charge the speaker pays, and the crowd's own speech—protest, denunciation, counter-demonstration—is protected on the same terms as the speaker's.
The line runs through the public campus in a specific way, because a public university is the government. In Healy v. James (1972), the Court held that a state college may not refuse recognition to a student organization because it disapproves of the group's views; in Widmar v. Vincent (1981), that a university which opens its facilities to student groups may not close them to a religious group on content grounds. A recognized student organization invites a speaker with its own funds and standing; the university that cancels or burdens the event because of the speaker's content acts as the state suppressing speech, and the students who protest the event exercise the same right in the opposite direction. The distinction is not a courtesy. It is the whole doctrine in one room.
The room, this week, is at Rutgers. NJ.com reported on November 12, 2026 that the Rutgers College Republicans, a recognized student organization at Rutgers University—New Brunswick, the state's public university, have invited Gregory Bovino, the former commander-at-large of the U.S. Border Patrol who directed the administration's urban immigration operations, to speak on Monday, November 16, from 6 to 8 p.m. in Trayes Hall at the Douglass Student Center, with audience questions to follow. A petition organized by the Climate Revolution Action Network, demanding that the university cancel the event, had gathered more than 1,600 signatures when NJ.com reported and nearly 2,000 by November 13; its text states that Bovino's presence “raises serious concerns about the safety and well-being of the Rutgers community” and risks “creating an intimidating and hostile environment for Rutgers students.” Community organizations announced a press conference and demonstration outside the venue an hour before the event. NJ.com's headline carried the protesters' formulation: “We don't stand for Nazis.” As of this filing, the university has taken no action against the event: it has declined to cancel it, and it has said it will provide a designated area for counter-protesters. The event is scheduled, not yet held; its outcome is not in the record. What is in the record is the alignment of the parties with the doctrine printed above—the invitation issued by a student organization, the cancellation demanded of the state institution, the institution declining, and the protest planned as speech against speech. Any change in the university's position would move the case from the protesters' column of the ledger to the government's.
The pandemic record: pressure, in documents
During the pandemic years the boundary question moved from distributors and insurers to the platforms that carry most public speech, and the record of what passed between the government and those platforms is unusually documentary. Beginning December 2, 2022, the new owner of Twitter released internal company records to journalists, published in installments as the Twitter Files: they showed federal agencies and officials flagging specific accounts and posts for the company's attention, the company's internal debate over suppressing the New York Post's October 2020 reporting on Hunter Biden's laptop, and a Stanford-based project that reported true and misleading posts alike to platform partners. The administration's own institutional attempt was brief: the Department of Homeland Security announced a Disinformation Governance Board on April 27, 2022, paused it within three weeks, and dissolved it on August 24, 2022. The litigation record is the spine. Missouri and Louisiana, with individual plaintiffs, sued in 2022, alleging that federal officials had coerced and significantly encouraged platforms to suppress disfavored viewpoints, including on the pandemic and the 2020 election. A district court entered a preliminary injunction on July 4, 2023. The Fifth Circuit, on September 8, 2023, narrowed it but sustained its core: the White House, the Surgeon General's office, the Centers for Disease Control and Prevention, and the Federal Bureau of Investigation, the court found, had likely coerced or significantly encouraged the platforms' moderation decisions; other agencies were removed for want of comparable proof. The Supreme Court, on June 26, 2024, in Murthy v. Missouri, held 6–3 that the plaintiffs lacked standing, because the chain between the officials' communications and the platforms' independent decisions was not traceable on the record as presented. The Court reached no other question: the merits were not decided, and the Fifth Circuit's findings stand vacated with the injunction, without adoption or rejection by the higher court. The documentary record and the legal record therefore say different things, and the desk prints both: the communications happened, in volume, as the letters below attest; whether they crossed the constitutional line the Vullo case draws has never been adjudicated on a full record.
Two letters from the companies themselves close the file. On August 26, 2024, Meta's chief executive, Mark Zuckerberg, wrote to the chairman of the House Judiciary Committee: “In 2021, senior officials from the Biden Administration, including the White House, repeatedly pressured our teams for months to censor certain COVID-19 content, including humor and satire, and expressed a lot of frustration with our teams when we didn't agree.” The letter states the boundary as the company drew it in the same paragraph: “Ultimately, it was our decision whether or not to take content down, and we own our decisions, including COVID-19-related changes we made to our enforcement in the wake of this pressure. I believe the government pressure was wrong, and I regret that we were not more outspoken about it.” On September 23, 2025, counsel for Alphabet, YouTube's parent, wrote to the same committee in materially stronger terms: “Senior Biden Administration officials, including White House officials, conducted repeated and sustained outreach to Alphabet and pressed the Company regarding certain user-generated content related to the COVID-19 pandemic that did not violate its policies.” The letter states that the administration “created a political atmosphere that sought to influence the actions of platforms based on their concerns regarding misinformation,” that government attempts to dictate moderation are “unacceptable and wrong,” and it commits the company to offer reinstatement to every creator terminated under its retired COVID-19 and election-integrity policies. The committee's ranking member answered the same day that the letter had “the ring of a false and coerced confession” extracted by the current majority's own pressure on the company, and noted that Alphabet employees had testified under oath that they were not coerced. Both letters, and the objection to the second, are part of the record; the pressure each describes ran from the government of the day toward the platform, and the pressure alleged against the second letter's production ran from the government of the next day toward the same platform. The doctrine of Bantam Books and Vullo does not distinguish between administrations, and neither does this report.
The lever in the statute book
The platforms' exposure to pressure has a statutory address. Section 230 of the Communications Decency Act of 1996, codified at 47 U.S.C. § 230, provides that “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,” and separately protects good-faith moderation. The provision is the legal foundation of the platforms' business: without it, carrying third-party speech at scale carries a publisher's liability for it. Its repeal has been invoked by officials of both parties in disputes over moderation. President Biden, as a candidate, told the New York Times editorial board in January 2020 that Section 230 “immediately should be revoked.” President Trump, on May 28, 2020, after a platform labeled his posts, signed Executive Order 13925, directing agencies to construe the statute narrowly and to review federal advertising on platforms; the order was revoked by his successor in May 2021. The record shows the statute functioning in these episodes as the consideration in the room: the companies' immunity is a grant of Congress, both parties' leaders have placed its withdrawal on the table while disputing the companies' moderation, and the companies' letters record the resulting atmosphere in their own words. No finding in the record ties a specific removal to a specific threat about Section 230; the desk prints the sequence, the documents, and the absence, together.
Three other democracies, statute by statute
The United Kingdom. Britain has no First Amendment. The right it has is Article 10 of the European Convention on Human Rights, carried into domestic law by the Human Rights Act 1998, and Article 10 is a qualified right on its face: after guaranteeing freedom of expression, its second paragraph provides that the right “may be subject to such formalities, conditions, restrictions or penalties as are prescribed by law and are necessary in a democratic society, in the interests of national security, territorial integrity or public safety, for the prevention of disorder or crime, for the protection of health or morals, for the protection of the reputation or rights of others.” The statutes that do the work are ordinary criminal law. Section 127 of the Communications Act 2003 makes it an offence to send, over a public electronic communications network, “a message or other matter that is grossly offensive or of an indecent, obscene or menacing character.” Section 1 of the Malicious Communications Act 1988 covers matter that is indecent or grossly offensive sent with intent to cause distress or anxiety. The Times, compiling freedom-of-information returns in an investigation published in April 2025, counted 12,183 arrests under the two statutes in 2023 by the forces of England and Wales that answered—about thirty-three a day, up from 7,734 in 2019—and Ministry of Justice figures reported alongside the count show fewer than one in ten of those arrested in 2023, some 1,119 people, were sentenced. The figures are arrests, not convictions; no American counterpart figure exists, because no American statute makes a grossly offensive message as such a crime. Alongside the statutes stand the non-crime hate incidents: since 2014, police guidance has required the recording of incidents perceived by the victim or any other person to be motivated by hostility to a protected characteristic, where no crime is alleged; the Court of Appeal held in the Miller case in December 2021 that the guidance's recording regime unlawfully interfered with Article 10 rights. A joint review by the College of Policing and the National Police Chiefs' Council concluded in March 2026 that the system is not fit for purpose, and the government accepted recommendations to end the recording of non-crime hate incidents; the replacement framework is expected in early 2027, and the existing arrangements remain in force until it arrives. The individual case the record is known by is Lucy Connolly, a childminder of Northampton and the wife of a Conservative councillor. On July 29, 2024, the day three girls were murdered at a dance class in Southport, she posted on X: “Mass deportation now, set fire to all the f****** hotels full of the b******* for all I care… if that makes me racist so be it.” The post was viewed some 310,000 times before she deleted it within hours. She pleaded guilty to publishing material with intent to stir up racial hatred, under the Public Order Act 1986, and on October 17, 2024 was sentenced at Birmingham Crown Court to thirty-one months' imprisonment, forty percent to be served in custody. The Court of Appeal refused her appeal against sentence in May 2025; she was released on licence in August 2025, having served about twelve and a half months including remand. Her sentence sits inside the American categories at no point: the post advocated violence against property and persons in terms the Brandenburg test would measure by imminence and likelihood, and it was punished instead for the hatred it intended to stir.
Germany. Article 5 of the Basic Law guarantees freedom of expression, of the press, and of broadcasting, and qualifies the guarantee in its second paragraph: “These rights shall 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 general laws include the criminal code's speech offences. Section 86a of the Criminal Code punishes the display of symbols of unconstitutional organizations—the swastika above all—with up to three years' imprisonment, subject to exceptions for art, scholarship, teaching, and reporting; a swastika on a T-shirt worn in public is a criminal matter in Germany, where the same shirt in the United States is the Skokie record. Section 130 punishes incitement to hatred against segments of the population. Section 188, expanded in 2021, gives persons in political life enhanced protection against insult and defamation; police statistics reported in 2026 counted 4,792 investigations under the provision in 2025, a record. The NetzDG—the Network Enforcement Act of 2017—obliged platforms to remove manifestly unlawful content within twenty-four hours on pain of fines up to €50 million, until the statute was largely superseded by the European Union's Digital Services Act in 2024. The enforcement case that made the system legible abroad began on November 12, 2024, when Bavarian police searched the home of Stefan Niehoff, a 64-year-old retired soldier, in the early morning and confiscated his tablet, after he reposted a meme describing the vice chancellor and economics minister, Robert Habeck, as a “Schwachkopf”—a professional moron—in a parody of a shampoo advertisement. The search followed a criminal complaint in a system in which a state reporting office had registered the post and forwarded it to the police. The insult proceedings were later dropped, prosecutors concluding the insult lacked the weight the statute requires. The search itself was lawful, warrant-backed, and unremarkable under the governing texts: that is the datum.
Canada. Section 2(b) of the Charter of Rights and Freedoms guarantees “freedom of thought, belief, opinion and expression, including freedom of the press and other media of communication,” and section 1 subjects every Charter right to “such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.” In R. v. Keegstra (1990), the Supreme Court of Canada upheld, 4–3, the Criminal Code's prohibition on the wilful promotion of hatred, under which a teacher had been convicted for teaching his students antisemitic doctrine: the limit was demonstrably justified, the majority held, by the harm hate propaganda does to its targets and to the society the Charter constitutes. The Canadian settlement is the British and German settlements' sibling: the right is real, and its limit is internal, applied by balancing at the point of legislation and sustained by the courts as a feature of the constitutional text rather than a violation of it.
The factual burden of the comparison can now be stated without emphasis. In the United States, expression is limited by a closed list of harm categories, construed by courts, and hateful content as such appears in none of them; in the United Kingdom, Germany, and Canada, the governing texts qualify the right in advance, and the official category of unlawful speech includes speech whose offence is the hatred, offence, or insult it expresses. American officials describing the European enforcement and European officials defending it use the same words for the two systems and mean different things by them, because the instruments differ. When Vice President Vance addressed the Munich Security Conference on February 14, 2025, he named the difference from the American side: “What I worry about is the threat from within. The retreat of Europe from some of its most fundamental values: values shared with the United States of America.” He cited Brussels, where, he said, “EU commissars warn citizens that they intend to shut down social media during times of civil unrest the moment they spot what they've judged to be, quote, ‘hateful content,’” and he stated the American premise as a rule of democracy: “dismissing people, dismissing their concerns or worse yet, shutting down media, shutting down elections or shutting people out of the political process protects nothing.” European leaders, including Finland's president and Germany's chancellor, rejected the characterization and defended their jurisdictions' authority to regulate hate speech and disinformation. The disagreement is genuine and the instruments are as printed. Each side's description of the other's system is, on the documents, substantially accurate; what differs is the settlement each text embodies about where the limit on expression is allowed to sit.
The money lever, 2025–2026
The current administration's instrument toward universities has been neither prosecution nor injunction but money. Its stated rationale, repeated across the 2025 funding actions, is the enforcement of federal anti-discrimination law: that universities receiving federal funds permitted antisemitic harassment after October 7, 2023, in violation of Title VI of the Civil Rights Act, and that federal research money may be withdrawn from institutions out of compliance. In March 2025 the government cancelled some $400 million in Columbia University's grants and contracts on that ground. Columbia settled. Under the agreement of July 23, 2025, the university pays $200 million to the federal government over three years and $21 million to settle Equal Employment Opportunity Commission investigations; the government reinstated the vast majority of the terminated grants and restored the university's eligibility for future funding; a jointly selected independent monitor oversees implementation; and Columbia states, and the agreement provides, that the university retains authority over faculty hiring, admissions, and academic decision-making, and admits no liability. The Secretary of Education said the government hoped the agreement would serve as a template for other universities. Harvard University declined the administration's April 2025 demands; some $2.2 billion in its research funding was frozen and grants terminated. Harvard sued, and on September 3, 2025, Judge Allison D. Burroughs of the District of Massachusetts held for the university: the freeze and terminations, she ruled, violated the First Amendment as retaliation for protected speech and as unconstitutional conditions and coercion, and violated Title VI's procedures and the Administrative Procedure Act besides; the freeze orders were vacated and the government permanently enjoined from reimposing them on the reasoning used. The opinion's conclusion contains the sentence by which it is known: on the record before her, she wrote, it was “difficult to conclude anything other than that Defendants used antisemitism as a smokescreen for a targeted, ideologically-motivated assault” on the nation's premier universities. The government appealed on December 31, 2025; the First Circuit heard argument on October 5, 2026, and, as of this filing, has not decided. Research payments to Harvard have resumed under the district court's order while the appeal pends. The two outcomes stand side by side in the record and the desk draws no line between them: one university purchased the restoration of its funding at the settlement table, accepting a monitor and policy commitments; the other obtained restoration from a court, on the ground that the government's use of the funding lever against a university's speech and governance is the indirect action the First Amendment forbids. Whether the appellate court sustains that ground is the largest open question in the American speech record as this report files.
The timeline
1791 — The First Amendment is ratified as part of the Bill of Rights. 1798 — The Sedition Act criminalizes scandalous writing against the government; it expires in 1801 unrepealed by any court. 1919 — Schenck states the clear-and-present-danger test; Holmes's Abrams dissent states the market theory. 1925 — Gitlow binds the states. 1949 — Terminiello reverses a speaker's conviction won by a hostile crowd. 1963 — Bantam Books holds informal official pressure to be state action. 1964 — Sullivan constitutionalizes libel. 1969 — Brandenburg sets the incitement standard that still governs. 1972–81 — Healy and Widmar carry the doctrine onto the public campus. 1977–78 — The Skokie litigation protects a Nazi march. 1990 — Canada's Keegstra sustains a hate-propaganda law under the Charter's limits clause. 1992 — R.A.V. and Forsyth County reject hate-speech selection and the heckler's fee. 1996 — Section 230 is enacted. 2011 — Snyder v. Phelps protects funeral picketing. 2017 — Matal v. Tam prints the doctrine's boast; Germany enacts the NetzDG. 2020 — Candidate Biden says Section 230 should be revoked; President Trump signs Executive Order 13925. 2022 — The Disinformation Governance Board is announced in April and dissolved in August; the Twitter Files begin in December; Missouri's suit proceeds. 2023 — Counterman sets the recklessness standard for true threats; the Fifth Circuit sustains the core of the coercion findings against four federal actors; British police make 12,183 arrests under the communications statutes. 2024 — Vullo restates the coercion doctrine unanimously; Murthy disposes of the platform case on standing; Zuckerberg's letter in August; Connolly sentenced in October; the Niehoff search in November. 2025 — Vance at Munich in February; Columbia settles in July; Burroughs rules for Harvard in September; Alphabet's letter in September; the government appeals the Harvard judgment in December. 2026 — Britain's police bodies declare the non-crime hate incident system unfit in March, replacement due in 2027; the First Circuit hears the Harvard appeal in October; Rutgers declines to cancel the Bovino event in November, and the event stands scheduled for November 16.
The record as it stands on November 14, 2026, is therefore this. The American text of 1791 and the case law built on it protect expression that the other democracies punish, including expression that is hateful, Nazi in content, and aimed at the bereaved, and they forbid the government four roads to the same end: direct prohibition outside the harm categories, selection among viewpoints within them, coercion of the private parties who carry speech, and—on the one appellate-pending ruling in the file—retaliation through federal money. The pandemic file shows the third road traveled in volume and adjudicated nowhere: the communications are documented, the companies' letters are on congressional paper, and the only Supreme Court disposition is on standing. The European and Canadian files show the alternative architecture operating as designed, at volumes the documents state plainly—12,183 British arrests in a year under two communications statutes, a thirty-one-month sentence for one post, 4,792 German investigations in a year into insults against politicians, a pensioner's home searched at dawn over a meme, and a Canadian hate-propaganda law sustained by the Charter's own limits clause. On Monday a former Border Patrol commander is scheduled to speak at a public university in New Jersey because the state institution that owns the room has declined to stop him, while protesters prepare to answer him in the street outside—the two protected acts facing each other across the state-action line, which is where the American record, from the Sedition Act forward, has always located them.