Fundamentals

Offensive Speech and Liberty: What Is Protected and Where the Limits Lie

By Daniel Sardá · Published on September 24, 2026

7 min read1,502 words

That speech wounds or provokes outrage does not settle whether it should be prohibited. The key is to distinguish offense from harm, identify who is imposing the consequence, and demand precise rules.

“That offends me” and “that threatens me” do not describe the same situation. A joke can be cruel, criticism can be humiliating, and an idea can provoke outrage. But a credible threat adds something different: a risk that does not depend only on how the listener receives the message.

That distinction is crucial to thinking about offensive speech and liberty. If mere offense were enough to trigger state power, almost any controversial position could be prohibited. At the other extreme, treating every expression as free from responsibility would ignore threats, harassment, defamation, or incitement capable of causing real harm.

The useful question is not whether words make people uncomfortable. It is what conduct is involved, what harm or risk can be shown, who seeks to intervene, and under what rules.

Key idea: Offense describes a response to a message; by itself, it is not sufficient justification for censoring or punishing it.

What is offensive is not a universal legal category

An expression is offensive when someone sees it as insulting, degrading, or contrary to deeply held convictions. That experience can be genuine and intense. Yet it varies across people, communities, and eras. Satire that some regard as legitimate may seem intolerable to others; religious, political, or moral criticism can wound precisely because it challenges a deeply held identity.

For that reason, “offensive” works better as a description of how a message is received than as a self-sufficient legal category. The International Covenant on Civil and Political Rights protects the freedom to seek, receive, and impart information and ideas of all kinds. The Human Rights Committee’s General Comment No. 34 clarifies that this scope includes expression that may be considered deeply offensive, without prejudice to restrictions that meet strict conditions.

This does not mean that “all offensive speech is legal” everywhere or in every circumstance. Laws vary, and an expression may also meet the elements of another regulated form of conduct. It means something more precise: taking offense alone does not end the analysis.

Not all hostile language belongs in the same category

Much confusion comes from placing different kinds of conduct under one label. It helps to separate them:

• An opinion or criticism expresses an evaluation, even if it is severe or unpopular.
• An insult seeks or causes contempt, but does not automatically amount to a threat.
• A threat communicates the possibility of causing harm and requires attention to its credibility, context, and reach.
• Harassment often involves persistent or targeted conduct that cannot be understood well by looking at an isolated remark.
• Defamation concerns, under the rules of each jurisdiction, factual assertions that damage reputation; it should not simply be equated with an unfavorable opinion.
• Incitement requires asking whether a message urges discrimination, hostility, or violence under a threshold far more demanding than mere bad taste.

These distinctions are not semantic tricks. They prevent the vaguest concept from swallowing the rest. Saying “that professor is incompetent” may be a harsh evaluation. Deliberately inventing a damaging fact, repeatedly targeting a person, or announcing an attack raises different problems. Classifying a particular case requires knowing the facts and the applicable law; this guide does not replace that legal analysis.

Nor should “hate speech” be used as a catch-all for hostile words. Article 20(2) of the Covenant refers specifically to advocacy of national, racial, or religious hatred that constitutes incitement to discrimination, hostility, or violence. The formula contains elements that must be established; it is not a license to prohibit every objectionable expression.

Why liberty also protects unpleasant disagreement

Freedom of expression would be far less necessary if it protected only polite, popular, and harmless ideas. Its institutional role appears precisely when an authority or majority wishes to silence a position. Pluralist societies need room to debate beliefs, expose abuses, revise customs, and make mistakes without turning every verbal conflict into a criminal matter.

From a classical liberal perspective, there is also a reason for caution: giving public power broad authority to decide what is offensive creates a tool that a future government may use against critics, minorities, or dissenters. Sensibilities vary; state coercion is concrete. This is part of why political power must have limits.

Protecting expression from public punishment does not require approving of it. Others remain free to respond, refute it, leave an association, withdraw support, or demand a correction. Political tolerance is not moral admiration. It is the willingness not to resort to force merely because an idea is repugnant.

Key idea: Freedom of expression does not promise comfort or immunity from criticism; above all, it limits the use of coercion to impose conformity.

When a state restriction may be justified

Under Article 19(3) of the Covenant, a restriction cannot rest on an authority’s mere preference. It must be provided by law, pursue one of the purposes permitted by the instrument itself, and be necessary to achieve it. General Comment No. 34 adds the requirement of proportionality: the measure must be appropriate, impair the right as little as possible, and bear a proper relation to the interest being protected.

This calls for specific questions:

• Is the rule accessible and sufficiently precise to guide conduct?
• What right, security concern, or legitimate interest is being protected?
• Is there a direct and immediate connection between the expression and the alleged harm or threat?
• Is there a less restrictive response that can address the problem?
• Can the decision be reviewed through independent procedures?

“Harm” is not a magic word either. Invoking it does not remove the need to identify what happened, establish a connection, and justify the response. The burden lies with the person seeking to restrict the right, especially when criminal penalties are involved. That standard belongs to a legal order in which power must be justified and open to review; general laws are one safeguard against arbitrary power.

For cases of incitement, the Rabat Plan of Action proposes examining six factors: context, the speaker’s position or influence, intent, content and form, the extent of dissemination, and the likelihood that harm will occur, including its imminence. It is a guiding framework, not a global criminal code, but it shows why an isolated word rarely resolves the case.

Warning: The vaguer the prohibition, the greater the risk of selective enforcement. Safeguards matter as much as a rule’s stated purpose.

The state, private spaces, and social response

Not every restriction on expression is state censorship. When a government fines, imprisons, or prevents publication, it uses coercive power and must satisfy the most demanding legal safeguards. When an association, media outlet, or online community applies its rules, it operates on a different plane.

That distinction does not make private actors infallible. A platform with great influence can apply ambiguous rules, discriminate, break commitments, or leave users without a route of appeal. This is why rules known in advance, consistent application, explanations of decisions, and reasonable review mechanisms are valuable. Specific legal obligations may also apply depending on the country.

But it is not the same for a community to remove a comment for breaking a published rule as for the state to punish an opinion. Nor do public criticism, a voluntary boycott, or a decision not to invite someone automatically amount to a criminal sanction. Each response may be fair or abusive, proportionate or excessive; judging it requires identifying the actor and the kind of power they exercise.

A simple example: a university may set neutral conditions concerning time, place, and peaceful coexistence for an event, and attendees may protest without preventing it. It is another matter to prohibit a lecture solely because its thesis is uncomfortable, or to tolerate concrete threats in the name of debate. The point is not to eliminate conflict, but to organize coexistence without confusing disagreement with violence.

Responsibility without making offense a crime

The alternative to prohibition is not indifference. In response to degrading expression, counterspeech, documented criticism, support for those targeted, education, voluntary norms of coexistence, and leaving spaces that fail to honor their commitments can all be meaningful. Where false factual claims or demonstrable harms are involved, correction or redress may be appropriate under precise and proportionate rules.

This approach recognizes an important objection: repeated harassment and coordinated campaigns can cause harm that no single sentence reveals on its own. Taking that problem seriously requires attention to pattern, intent, reach, and risk—not reducing the analysis to “someone felt offended.” Precision protects both potential victims and those who might otherwise be punished arbitrarily.

Liberty should remain the general rule. Its limits require clear categories, evidence, and institutional oversight. Between unconditional approval and prohibition lies a broad field of expression, responsibility, reply, and redress. Keeping it open makes it possible to address real harms without giving any authority the power to turn changing sensibilities into the boundaries of what may be said. This is also where the limits of coercion lie: protecting others against demonstrable harms without turning every offense into a prohibition.