Developments in Human Rights Law and the New Human Right of Free Access to Public Legal Information: The New Human Rights-Advocacy Approach and the Ten Criteria for the Formal Recognition of New Human Rights

By Leesi Ebenezer Mitee

11 min read

Why does it matter? Because the speech restrictions your courts apply are already being measured against a standard most judges were never taught.

The European Court does not wait for appeal. It reads first-instance reasoning. It asks whether the trial court applied the three-part test — prescribed by law, legitimate aim, necessary in a democratic society — or simply deferred to a local standard no one examined.

Most national judges handle an Article 10 case the same way: apply domestic law, note the Convention in passing, assume Strasbourg will sort it out if things go wrong. That assumption is where violations are born. The gap between how courts assume the Convention works and how the Court actually evaluates restrictions is not marginal. It is the center of the doctrine. That explains why the same offensive statement earns full Article 10 protection in one factual context and loses it entirely in another — on factors most lawyers never think to argue.

A 1950 Treaty That Binds Domestic Courts Like Statute — and Never Stops Evolving

Every national judge in forty-six European states must apply the European Convention on Human Rights from the first hearing as binding law. It also means whatever the Strasbourg Court currently says it means — and that second part changes everything.

The change runs deepest if your legal tradition is continental. France, Germany, Poland — their courts were built on codified statutes, not on case law as a source of binding rules. The Convention broke that architecture. Once a state ratifies it, Strasbourg judgments bind domestic courts as hard law. A Romanian magistrate hearing a speech-restriction case today is bound by last year's Strasbourg ruling on the same issue — not as persuasive authority. As law.

The Convention is also designed to keep moving. The Court must track commonly accepted standards across member states, making the 1950 text speak to present-day conditions by design: a living instrument, in the doctrine's own phrase. In 1950, 'the press' meant ink on paper; today the same Article 10 rules govern whether a state can hold an online news portal liable for anonymous reader comments.

For freedom of expression, this architecture carries its sharpest edge. States normally retain a margin of appreciation (discretion to make their own rights judgments within limits). Under Article 10, the Court reduces that discretion, in its own words, almost to vanishing point. That phrase is a signal: freedom of expression is the Convention's most tightly supervised right, and the Strasbourg Court intends to stay close.

The State Must Justify Every Restriction — and 'Prescribed by Law' Is Harder Than It Sounds

A man in Poland tries to register a periodical. The domestic courts refuse: the proposed title, they rule, "would be in conflict with reality." No statute authorized a truth test. The courts invented one. The applicant goes to Strasbourg, and the European Court of Human Rights finds a violation: not because the restriction was disproportionate, but because the law authorizing it didn't exist in any foreseeable form.

That is Gawęda v. Poland. The case shows what practitioners most often get wrong about Convention law: a statute satisfies Article 10(2) only if it meets specific quality requirements. Failing that threshold, the analysis ends.

Under Article 10(2), the Court evaluates restrictions through a three-part cumulative test. Each condition must be satisfied in sequence, and the state carries the full burden. The interference must be prescribed by law. It must serve one of the legitimate aims the Convention lists: national security, public order, protection of reputation, among others. And it must be necessary in a democratic society, meaning proportionate to a pressing social need. Miss any one condition and the Court stops. The other two don't save you.

Most practitioners focus on the third prong — necessity, proportionality, pressing social need. That's where the balancing happens, so it feels like the real fight. Gawęda shows you can lose at the first gate before the balancing begins.

'Prescribed by law' carries a specific meaning: the law must be publicly accessible, and its consequences must be individually foreseeable. A person must be able to predict, with reasonable certainty, what conduct is prohibited and what follows from crossing the line. When Polish courts refused Gawęda's application under a criterion they constructed on the spot, he had no way to know that criterion applied to him. The Court found the registration law insufficiently precise. Refusing to register a title, it noted, is equivalent to refusing to publish. Prior restraints of that severity require clear and foreseeable legal grounding, and the invented test provided none.

The prescribed-by-law test applies across every form of speech restriction, not only publication controls. In Zakharov v. Russia, involving covert surveillance of communications, the Court acknowledged that no individual can foresee when intelligence services will intercept their calls. That structural impossibility makes the requirement more demanding, not less. Because the individual cannot verify what the state is doing, the law itself must specify, in clear and detailed terms, the scope of discretion the authorities hold. The less visible the conduct, the more legible the legal authority must be.

In Gawęda, the answer was no. The state lost before the case had really begun.

How the Same Offensive Act Gets Opposite Verdicts: What the Court Actually Tests

December 2008, Paris. At the end of a show at the Zénith concert hall, comedian Dieudonné calls convicted Holocaust denier Robert Faurisson onto the stage. Faurisson had multiple French criminal convictions for denying the existence of gas chambers. An actor dressed in striped pyjamas, a yellow star sewn on to read "Jew," presents him a prize for "unfrequentability and insolence" — a three-branched candlestick. The man who denied the extermination of Jews receives an award from someone costumed as a Jewish deportee.

When the case reached Strasbourg, the Court never got to Article 10. It invoked Article 17 (the Convention's prohibition on abuse of rights) and stripped protection at the threshold. The show had ceased to be entertainment. It had become a political meeting using comedy as cover. In the Court's language, it was "as dangerous as a head-on and sudden attack," a platform for ideology the Convention was designed to exclude. Dieudonné had used free expression to pursue ends incompatible with the Convention itself.

Now hold that case next to this one. A Turkish politician named Perinçek gave public speeches in Switzerland arguing that the 1915 Ottoman massacres of Armenians did not amount to genocide. Swiss courts convicted him. Strasbourg found a violation of Article 10 and gave him full protection.

Same domain: disputing whether mass atrocity constitutes genocide. Opposite outcome. The Court's explanation is where the doctrine lives.

Six factors separated Perinçek from Dieudonné's category. His statements concerned a matter of genuine public interest. They contained no call to hatred or violence. Switzerland had no particular historical tension around the subject. The dignity impact on Armenians, real as it was, didn't reach the threshold that calls for a criminal law response. No international obligation required Switzerland to criminalise the speech. And — the sharpest factor — the Swiss courts appeared to have punished Perinçek not for dangerous speech but for stating a view that diverged from Switzerland's established consensus.

That last point is where the Court's logic becomes actionable for you. Criminal law deployed to enforce an official historical position, rather than to prevent genuine harm, is itself a Convention violation. The Court is testing proximity to violence, dignity impact against a threshold, and whether a democratic society genuinely requires criminal punishment, or whether the state is suppressing dissent dressed as historical inquiry.

The Zénith performance failed those tests entirely. Faurisson's presence, the deportee costume, the award ceremony: a live political act designed to rehabilitate Holocaust denial before an audience, not entertain one. Perinçek's speeches were contested historiography in a country where no specific tensions made criminal punishment proportionate.

Content didn't decide either case. The cluster of factors did.

The Court Promises Near-Zero Deference on Speech — Then Grants Wide Discretion on Morality

If the Court's deference on freedom of expression shrinks almost to vanishing point, what explains a criminal conviction upheld for publishing 2,000 copies of a novel about theology?

In I.A. v. Turkey, a Turkish publisher released a novel engaging with philosophical and religious questions. Turkish courts convicted him of blasphemy: insulting "God, the Religion, the Prophet and the Holy Book." When the case reached Strasbourg, the majority followed its earlier Wingrove precedent (a 1996 ruling allowing the UK to ban a film for blasphemy) and allowed Turkey a wide margin of appreciation. Believers, the Court reasoned, may legitimately feel themselves the object of offensive attacks; protecting what Muslims hold sacred was a pressing social need.

That reasoning did the work a near-zero deference promise should have prevented. The conviction stood.

Three judges — Costa, Cabral Barreto, and Jungwiert — said the majority had abandoned the Convention's own logic. Their dissent names the fracture precisely. They pointed back to Handyside, where the Court had written that Article 10 protects ideas that "shock, offend or disturb the State or any sector of the population." Those words, the dissenters warned, must not become "an incantatory or ritual phrase," a formula recited without commitment. Applied seriously, those words should have protected the novel. Instead, the majority had extracted isolated passages from a whole book and used them to impose criminal sanctions on the publisher. And, they added flatly: "a democratic society is not a theocratic society."

Their conclusion was structural. The Wingrove line, they argued, placed too much emphasis on conformism and reflected an "overcautious and timid conception of freedom of the press." They called explicitly for the Court to revisit its own precedent.

That call has not been answered. The doctrine draws a two-tier map: narrow margin for political speech and public-interest debate, wide margin for content that offends moral or religious conviction. No clear principle determines when each tier applies. The most reliable signal is whether the expression engages a contested public debate: cases that do tend toward the narrow margin; cases reducible to offence against religious or moral sentiment alone pull toward the wide one.

For national courts, this creates a specific obligation. The vanishing-point promise has a genuine exception, but European supervision holds just as strictly inside it. You must still test necessity; deferring to local moral consensus without that test violates your Convention obligations. Open Door and Dublin Well Woman shows where the floor sits: Irish courts issued a permanent injunction against two nonprofits providing pregnant women information on abortion services in the United Kingdom, covering all women "regardless of age or state of health." The Court struck it down. Magazines and telephone directories already carried the same information. Absolute, perpetual bans fail even in morality's heartland.

The Punishment Itself Can Be the Violation: Why Amount Doesn't Settle the Question

A British jury finishes deliberating. The defendant, Count Nikolai Tolstoy, has written a defamatory pamphlet. The damages: £1,500,000, payable to Lord Aldington. The case reaches Strasbourg — and the European Court does not ask whether the defamation finding was correct. It asks whether the award itself breaches Article 10. It does.

That is the move you need to understand. The sanction is not a downstream severity question. It is an independent variable in the Article 10 analysis. Even a justified conviction can produce a disproportionate penalty that constitutes a separate violation.

The Court's reasoning in Tolstoy Miloslavsky turned on two requirements. Civil damages must bear reasonable proportionality to the actual harm to reputation — not to a jury's instinct about punishment. And appellate courts must have real power to override an inflated award. At the time in England, they had almost none. That structural absence was fatal. Lord Aldington may have suffered genuine reputational harm. The Convention does not authorise an uncapped response to it.

The same principle reaches small penalties. In Lingens v. Austria, a journalist received a fine modest enough that it left him free to keep publishing. Still a violation. A fine that doesn't silence one journalist still broadcasts a warning to every journalist watching: make this kind of criticism and pay this kind of price. That collective deterrent is exactly what Article 10 guards against. Nominal amount is irrelevant when the chilling effect operates across the whole press corps.

Challenge the sanction as a separate Article 10 question, not just as part of the conviction's proportionality. On civil damages, show the mismatch between award and actual reputational injury, and identify any gap in judicial oversight mechanisms. On criminal fines, document the deterrent effect on future journalism: the person who didn't write the story because of what happened here. The violation lives in that silence.

Two New Frontiers, One Shared Logic: Employees Who Speak Out and Platforms That Host Comments

A geriatric nurse in a German care home spent months telling management that staff shortages were causing patients to go without basic care. Management ignored her. She filed a criminal complaint against her employer. The employer dismissed her without notice.

When Heinisch v. Germany reached Strasbourg, the Court protected her. But the reasoning is what matters for your practice. The Court noted she had "allowed herself a degree of exaggeration and generalisation" and found it irrelevant. She had raised her concerns internally, repeatedly and to no avail. The patients she cared for were too vulnerable to speak for themselves, making her uniquely positioned to act in the public interest. Her immediate dismissal broadcast a warning to every employee watching: file a complaint and lose your job. The Convention reads the punishment's reach beyond the individual — a sanction that silences the next person to know something is itself the harm.

The framework this produces is a checklist, not a sliding scale. The Court asks: Did the discloser exhaust internal channels first? Was the information genuinely in the public interest? Did the person act in good faith, verifying what they could even if imperfectly? And was the sanction proportionate? A dismissal severe enough to silence the next employee who knows something fails that test regardless of whether the original disclosure was fully justified.

That framework carries into platform liability, though the operative variables are different. Whistle-blower cases turn on exhaustion of internal channels and good faith; platform cases turn on content severity and host responsiveness. In Delfi v. Estonia, a large commercial news portal left anti-Semitic reader comments unaddressed for six weeks. Strasbourg upheld liability. In Magyar Tartalomszolgáltatók v. Hungary, a non-commercial portal hosted vulgar but non-hateful criticism of two real estate companies. Strasbourg found a violation because the Hungarian courts had accepted at face value that the comments were unlawful without conducting any balancing exercise. The analytical questions were the same: was the content extreme, did the host respond promptly, did the court actually weigh the competing interests? The inputs were different; the outcomes were opposite.

Magyar and Delfi are the same test with different facts. You bring the right variables for whichever side of that line you're arguing.

Every Restriction Is Already on Trial

The judge who asks "is this proportionate?" before asking "is this prescribed by law?" has already lost the analysis. That sequence isn't procedural formality — it's how the Convention forecloses arguments before they begin. Start there. On sanctions: quantify the gap between the award and actual reputational harm, then document the silence of whoever didn't write the next story. On whistle-blowers: run the checklist — internal channels exhausted, genuine public interest, good faith even where imperfect, sanction proportionate enough not to warn everyone watching. On platforms: establish whether the content meets the hate speech threshold, whether the host is commercial, and whether the court conducted any real balancing. That's what the Court checks, case by case. Practitioners who know each criterion precisely work in a different courtroom than those who don't.