# Hate Speech A contested legal-and-rhetorical category covering speech that attacks or demeans a person or group on the basis of protected attributes (race, ethnicity, national origin, religion, sex, sexual orientation, gender identity, disability, and in some jurisdictions additional attributes including age, occupation, or political affiliation). The category does substantial legal work in most democracies, where hate-speech statutes exist alongside or as exceptions to general free-expression protections. In the United States the category has *no first-amendment status* as a legal class — speech that other democracies prohibit as hate speech is generally protected here — and the *contestation over the term itself* is among the most-active rhetorical battlegrounds of the post-2010 American information environment. ## The international legal lineage The post-WWII international human-rights instruments establish the foundational vocabulary: the Universal Declaration of Human Rights (1948) commits to both freedom of expression and protection against "incitement to discrimination" (Article 7); the International Covenant on Civil and Political Rights (ICCPR, 1966, Article 20) requires states to prohibit "advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence"; the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD, 1965, Article 4) requires criminalization of "all dissemination of ideas based on racial superiority or hatred." Most democracies (UK, Germany, France, Canada, Australia, EU member states broadly) have implementing legislation with criminal or civil penalties for speech that meets statutory hate-speech thresholds. The United States is the principal democratic outlier: the US ratified ICCPR in 1992 *with a reservation* explicitly stating that Article 20 does not authorize legislation that would restrict speech protected by the [[First Amendment]]. ## The American legal frame In the US, the Supreme Court line of cases makes the speech-protective position explicit: - ***Brandenburg v. Ohio*** (1969) — sets the *imminent lawless action* standard; speech advocating violence is protected unless directed at producing imminent unlawful conduct AND likely to do so - ***National Socialist Party v. Skokie*** (1977) — nazi-march speech is protected - ***R.A.V. v. City of St. Paul*** (1992) — cross-burning ordinance struck down as content-based discrimination - ***Virginia v. Black*** (2003) — cross-burning prohibition allowed only when there is specific intent to intimidate - ***Snyder v. Phelps*** (2011) — Westboro Baptist Church protest speech at military funerals is protected - ***Matal v. Tam*** (2017) — *The Slants* band trademark case; "the proudest boast of our free speech jurisprudence is that we protect the freedom to express the thought that we hate" - ***Counterman v. Colorado*** (2023) — true-threat doctrine requires showing speaker's subjective recklessness about threat-content The doctrinal upshot: there is no *hate speech* category in US constitutional law. Speech is either protected (the default) or falls into one of the narrowly-defined unprotected categories (incitement to imminent lawless action, true threats, fighting words, defamation, obscenity). The category *hate speech* does work in academic, policy, journalistic, and platform-content-moderation contexts but has no independent First Amendment status. ## The 1980s–90s academic and campus debates Mari Matsuda's 1989 *Michigan Law Review* article *Public Response to Racist Speech: Considering the Victim's Story* is the foundational legal-academic intervention proposing a narrow hate-speech doctrine that would survive First Amendment scrutiny. Charles Lawrence, Richard Delgado, Catharine MacKinnon, and other critical-race-theory and feminist-legal-theory scholars developed the framework through the late 1980s and 1990s. The campus-speech-code debate of the same period — universities adopting speech codes prohibiting harassment based on protected characteristics — produced the *Doe v. University of Michigan* (1989) and *UWM Post v. Board of Regents of University of Wisconsin* (1991) cases that struck down speech codes as overbroad and vague. The codes returned in narrower forms; the analytical debate over their permissibility remains active. ## The 2010s–2020s platform-content-moderation turn The most consequential operational use of *hate speech* in contemporary American practice is now platform content-moderation rather than legal regulation. Major platforms developed hate-speech policies through the 2010s as user-base scale produced enforcement-pressure on coordinated harassment, organized racism, and post-2016 organized extremism: - **Facebook / Meta** — *Community Standards on Hate Speech* (formalized 2018+); covers attacks on people based on protected characteristics; tiered enforcement - **Twitter / X** — *Hateful Conduct Policy* (formalized 2018); substantially narrowed under Musk ownership 2022+ - **[[YouTube]]** — *Hate Speech Policy* (formalized 2019); meaningful enforcement infrastructure - **TikTok** — *Community Guidelines on Hate Speech* (formalized 2020); aggressive enforcement - **Reddit** — *Content Policy* prohibiting promoting hate based on identity (revised 2020 to allow subreddit-level bans for hate-speech violations) Each platform's specific definition of *hate speech* differs; enforcement is famously inconsistent; the political contestation over the definitions and enforcement is itself substantial. ## The contemporary right-wing rebrand attempt A coordinated rhetorical project from approximately 2020 has worked to make *hate speech* itself an *illegitimate category* — framing the term as a vehicle for partisan censorship, conflating any enforcement with viewpoint-discrimination, and producing the rhetorical formulation that has become canonical in right-wing media: "hate speech is just speech [redacted partisan] disagrees with." The strategic-rhetorical work: 1. **Redirect attention from speech-content to enforcer-identity** — frame any moderation decision as evidence of bias on the moderator's part 2. **Cite asymmetric enforcement** — point to specific moderation decisions that hit conservative accounts; treat absence of analogous left-wing enforcement as evidence of systematic bias rather than as evidence of different content patterns 3. **Conflate platform moderation with government censorship** — treat private-company terms-of-service as if they were First Amendment violations; the *Murthy v. Missouri* litigation explicitly tested this conflation 4. **Reclaim the term *hate speech* for the right** — apply it to anti-conservative content (e.g., to academic CRT, DEI, anti-MAGA criticism, transgender-affirming content); the term is partially reversed in the process 5. **Frame moderation as *censorship industrial complex*** — Michael Shellenberger / [[Matt Taibbi]] / *[[Twitter Files]]* framing; Renee DiResta's *Invisible Rulers* (2024) documents the project from the receiving-end perspective The campaign has had measurable success: major-platform hate-speech enforcement has substantially weakened (X under Musk; Meta civic-integrity contractions; YouTube content-policy revisions); academic disinformation research programs have been defunded or wound down ([[Stanford Internet Observatory]] 2024); congressional pressure ([[House Judiciary Committee]] subpoenas) has narrowed government-platform cooperation on hate-speech-and-disinformation issues. ## Why the term is in the disinformation lexicon *Hate speech* sits inside the disinformation lexicon less because it names a specific disinformation technique and more because: 1. **The contestation over the term is itself doing analytical work** — both sides of the contemporary American argument deploy the term; the analytical interest is in *how* it gets deployed, with what asymmetric application, with what policy consequences 2. **The term sits at the intersection of speech, harassment, and downstream-violence concerns** — connects to [[Stochastic Terrorism]], [[Doxxing]], [[Sockpuppet Accounts]], [[Troll Farms]] in the operational vocabulary 3. **The right-wing rebrand-attempt parallels the [[Fake News]] rebrand pattern** — the *concept capture* mechanism documented in *Fake News* is now being attempted with *hate speech*, *misinformation*, *disinformation*, and *CIB* (coordinated inauthentic behavior) 4. **The platform-moderation question is one of the central post-2020 democratic-discourse policy questions** — the term is in active legislative, regulatory, and judicial use, with stakes for the long-run shape of the American information environment ## Some careful framings worth keeping Three honest framings worth recording: 1. **Reasonable people genuinely disagree** about where the line should be drawn; the contestation is not solely manufactured. The American constitutional default for very broad speech protection has substantive arguments on its side (chilling-effect risks of vague hate-speech statutes; minority-protection concerns about whoever holds the enforcement power; difficulty of consistent application) 2. **The European model has substantive arguments on its side** too (post-WWII evidence that unregulated speech enabled atrocity; majority-group-speech-against-minorities pattern of harms; dignity-and-equality concerns) 3. **The contemporary American debate is not primarily a debate about the underlying philosophical question** — it is a debate about specific platform-moderation decisions, specific campus events, and specific legislative proposals, with the philosophical positions deployed instrumentally on each side ## See Also - [[Stochastic Terrorism]] — the downstream-violence mechanism hate-speech often implicates - [[Censorship]] — the term contested speech-restrictions are often labeled with - [[Cancel Culture]] — adjacent contested vocabulary - [[Doxxing]] — the harassment workflow hate-speech-enforcement often targets - [[The Big Lie]] / [[Misinformation]] — adjacent contested-vocabulary terms in the same rhetorical project - [[Disinformation Hub]] — Subject hub - [[Disinformation Lexicon]] — alphabetical MOC