How Have the October 7 Atrocities Been Justified? Reading the Catalogued Frames, the Party’s Own Text and the Treaty Clauses

The justifications offered for the attack of 7 October 2023 sort into three catalogued categories, the typology the 2025 Israel Affairs article on terrorism denial applies to the day: literal denial that the events occurred, interpretive denial that they occurred as described, and implicatory denial that accepts the facts and contests their weight. The most repeated version belongs to the third category and has a dated primary text behind it, the report Hamas released in January 2024 calling the attack a necessary step while denying that civilians were intentionally targeted. The count that text disputes is a peer-reviewed one: more than 1,200 people killed in a single day, the figure the Israel Law Review carried in its 2025 legal analysis. Whether the vocabulary of justification travels with measured antisemitic belief is a separate question, and the later sections take it up with survey data.

Key Findings

• Three categories of denial, literal, interpretive and implicatory, organise the 2025 peer-reviewed literature on how the day was explained, in Israel Affairs and in Studies in Conflict & Terrorism.
• Hamas’s January 2024 document is the dated source of the most-quoted justification, calling the attack a necessary step and denying intentional targeting of civilians, that party’s own account of its own conduct.
• Additional Protocol I of 1977 does both things at once: Article 1(4) extends the law of international armed conflict to conflicts against colonial domination and alien occupation, and Article 51(8) states that violations by one party do not release the other from its obligations toward civilians.
• Across two nationally representative Norwegian surveys (N = 1,575, and a preregistered replication of N = 1,653), Enstad (2026) found endorsement of Holocaust inversion associated with justification of harassment and violence against Jews.
• The Community Security Trust recorded 1,205 antisemitic incidents across the United Kingdom in the 35 days from 7 October 2023, against 191 over the same dates in 2022, on the Trust’s own definitions and by its own provisional count.

Which justification frames have researchers actually catalogued?

Three, and they are named. The 2025 Israel Affairs article “Terrorism denial: 7/10 and the reasons to deny a pogrom” applies Stanley Cohen’s three-part typology of denial to the attack: literal, interpretive and implicatory. Studies in Conflict & Terrorism published a companion 2025 article, “The Denial of Hamas’ Atrocities”, working the same three categories.

Source: Israel Affairs (Taylor & Francis), 2025, DOI 10.1080/13537121.2025.2516864; Studies in Conflict & Terrorism (Taylor & Francis), 2025, DOI 10.1080/1057610X.2025.2494284. Both peer-reviewed.

The typology is not new to this event. Cohen developed it in “States of Denial” (2001) to describe how bystanders, officials and publics respond to atrocity, and the value of three categories rather than one is that they are answered by different kinds of evidence. A literal denial is answered by a record of what happened. An interpretive denial is answered by a record of who was attacked. An implicatory denial concedes both and argues about context, which is why it survives the publication of evidence that closes the first two.

The 2025 Israel Affairs article records the endpoint of the third category in a phrase: the presentation of the perpetrators as “resistance fighters rather than terrorists”. The Journal of Genocide Research published a further 2025 article in the same area, and the Israel Law Review’s 2025 analysis treats denial and reverse accusation as linked moves rather than separate ones.

Claims that the attack was staged, or that the deaths were substantially caused by Israeli fire, are not a fourth category in this scheme. They sit inside the first two, and they carry a different evidentiary burden from the contextual argument, which asserts nothing about the facts of the day.

Is it often said the attack was anti-occupation resistance?

Yes, and it is the most frequently recorded justification of the three. It is often said that the attack was resistance to occupation, blockade and settlement expansion, and that, in the formulation that circulated within days, this war did not start on 7 October. That framing has named holders and dated texts, and the two questions it fuses, the motive claimed and the persons attacked, are answered by different records.

Hamas officials described the attacks as justified by the occupation, a position recorded in the Congressional Research Service In Brief on the Israel and Hamas conflict (R47828), the standing nonpartisan reference prepared for members of Congress. The organisation’s own January 2024 document sets out the same account at length and is the subject of the next section.

The academic literature has given the counterpart framing a name of its own. A 2025 article in Ethnopolitics describes a “Zero Hour” framing in which 7 October is treated as a decontextualised beginning, and argues that the framing works in both directions: it strips the day of what preceded it, and it also supplies the terms in which later violence is presented as justified. The term is that journal’s analytic label, dated 2025, and it describes a discourse pattern rather than adjudicating any grievance the discourse invokes.

What the resistance framing asserts is a characterisation of purpose. It is not, in itself, a claim about who was present in the communities attacked, nor about what the law of armed conflict permits a party fighting an occupation to do. Those are the two questions the record answers directly, and the sections below take them in order: first what investigators and peer-reviewed analysts recorded about the persons attacked, then what the treaty texts say about the conduct. Three distinct claims are folded into one sentence in the popular version, and they come apart the moment each is checked against its own evidence.

What does Hamas’s own published account of 7 October claim?

One document carries most of it: the report Hamas released in January 2024, which calls the attack a necessary step, cites the occupation, the blockade, settlement expansion and settler violence, and denies that civilians were intentionally targeted. Those are that party’s own assertions about its own conduct, published roughly 15 weeks after the events.

The document’s two halves do different work. The first is a grievance account, and it is the source of the justification vocabulary that circulated afterwards. The second is a factual denial: it states that the attack did not intend the killing of civilians and attributes civilian deaths to battlefield conditions and to Israeli fire. The Congressional Research Service In Brief (R47828), an official record, registers the same pairing in the statements of Hamas officials, who described the attacks as a response to occupation while denying intentional targeting of civilians.

In the terms of the typology above, the combination is interpretive denial resting on implicatory denial, and the 2025 Israel Affairs and Studies in Conflict & Terrorism articles classify it that way. That classification is a description of the structure of the claim, not a finding about the events.

The evidentiary status of the document is the point that governs how it can be used. It is an account by a party to the events, of its own conduct in those events, and it is evidence of what that party asserted. It is not evidence of what occurred, and no figure derived from it can be presented as a neutral or independently verified count. The same standard applies in the other direction to counts published by governments party to the conflict, several of which appear in the next section and are labelled there.

What the record does establish about the document is its content and its date. That matters for sequence: the framing under discussion was already circulating in October 2023, before the January 2024 text set it out formally, which means the text codified a vocabulary rather than introducing it.

What did investigators record about who was targeted?

More than 1,200 people were killed on 7 October 2023, the figure the Israel Law Review carried in its 2025 peer-reviewed legal analysis, which also records systematic killing alongside torture, sexual violence and mutilation among the documented conduct. Israel’s own reporting is narrower in scope and close in magnitude: the Begin-Sadat Center for Strategic Studies recorded 1,184 Israelis confirmed killed on the single day.

Hostage-taking is recorded separately and by a party. The Israeli Ministry of Foreign Affairs stated that as of 14 September 2025, 207 hostages had returned or been recovered, 148 alive and 59 deceased, with 48 remaining in captivity, living and deceased. Those are the ministry’s own figures, published by a party to the conflict and not independently verified.

Two organisational determinations sit alongside the peer-reviewed count and are a different kind of statement from it. Human Rights Watch published a report on the 7 October attacks in July 2024 concluding, on that organisation’s own definitions and mission, that the assault was designed to kill civilians and take hostages, and stating in its own words that atrocities do not justify atrocities. The Civil Commission’s 2024 report on sexual violence characterised the acts it examined as war crimes, crimes against humanity and genocidal acts under international law, that commission’s own legal characterisation of the evidence before it.

A legal characterisation by an organisation and a measured count in a peer-reviewed journal answer different questions, and collapsing them is the error the interpretive-denial claim relies on. The claim that only soldiers or armed civilians were attacked is a factual assertion about the persons killed. The figure that meets it is the count, dated to a single day, and the composition of that count is what the 2025 Israel Law Review analysis and the July 2024 organisational report each address in their own registers.

Does the law of armed conflict recognise a resistance exception?

No exception appears in the instruments. Additional Protocol I of 1977 extends the law of international armed conflict to conflicts against colonial domination and alien occupation in Article 1(4), and then binds every party to Article 51(2): “The civilian population as such, as well as individual civilians, shall not be the object of attack.” The recognition and the prohibition sit in the same treaty.

The clause that answers the reciprocal version of the argument is Article 51(8) of the same Protocol, which provides that any violation of these prohibitions “shall not release the Parties to the conflict from their legal obligations with respect to the civilian population and civilians”. Article 51(7) separately prohibits using the presence of civilians to shield military objectives, and states that the prohibition binds regardless of the other party’s conduct. The text therefore forecloses both directions of the exchange: prior violence by one party neither licenses attacks on the other’s civilians nor transfers responsibility for them.

Hostage-taking is prohibited without qualification as to the character of the conflict. Common Article 3(1)(b) of the four Geneva Conventions of 12 August 1949 forbids “the taking of hostages” in conflicts not of an international character. The Rome Statute of 1998 makes hostage-taking a war crime in Article 8(2)(c)(iii) and attacks intentionally directed against the civilian population a war crime in Article 8(2)(b)(i); Article 7(1) makes murder and rape crimes against humanity when committed as part of a widespread or systematic attack directed against any civilian population.

A 2025 article in the Journal of Conflict and Security Law (Oxford University Press, volume 30, issue 2, page 287) works through how these frameworks apply to mass terrorism and to the 7 October attack specifically, and it is the peer-reviewed treatment closest to the question as commonly asked.

Two things the treaty texts do not do are worth stating plainly. They do not settle any contested characterisation of the occupation that the resistance framing invokes, which is why Article 1(4) exists at all. And they do not, by themselves, establish what occurred on any given day; that is the work of the counts and determinations in the previous section. The instruments answer one narrow question: whether the category of conflict alters the prohibition on attacking civilians. On their own wording, it does not.

How were Israeli civilians recast as legitimate targets online?

The closest measured corpus predates the attack. Becker, Ascone and Troschke (2022, peer-reviewed) coded 1,504 UK comments on leading media Facebook pages reacting to the May 2021 escalation and found 26.9% antisemitic against an IHRA-based scheme, with 27.9% of the antisemitic UK comments attributing sole guilt to Israel and 7.7% denying Jewish self-determination.

The same study found that thread context was needed to infer the antisemitic meaning in 38.7% of the UK antisemitic comments, which is the measurable trace of coded target-framing: the reframing is often carried by the surrounding conversation rather than by the sentence itself. Mendelsohn and colleagues (2023, peer-reviewed conference) document the general mechanism in a glossary of over 300 coded terms and report that content containing them evaded automated toxicity detection.

For the 2023 event, the measured record is one of volume and intensity rather than of frame prevalence. Guerra, Lepre and Karakuş (2024, preprint) scored over 450,000 posts and comments from four subreddits between 7 October 2023 and 3 January 2024, with daily post volume near 10,000 at the onset and extremity-score peaks that coincided with offline conflict events. Chen and colleagues (2024, preprint) assembled 412,258 submissions and 8,089,095 comments across 100 subreddits from August to November 2023, of which the Zionism-and-antisemitism topic subset held 126,107 submissions and 2,516,114 comments; those figures measure how much was said on a topic, not what share of it was antisemitic. The Institute for Strategic Dialogue’s 2025 audit of over 5,500 recommended videos found a profile of a 15-year-old boy shown antisemitic conspiracy content within about an hour on TikTok.

No study in this record measures what share of 7 October discussion recast Israeli civilians as combatants. The vocabulary is documented and dated; its prevalence at that event is not measured, and that absence is where the evidence currently stops.

Was the justifying discourse amplified by coordinated accounts?

11 coordinated groups involving 541 accounts were identified by Elmas and colleagues (2026, preprint) as amplifying the justifying discourse on X.

Elmas and colleagues found this pattern within 4,570,633 tweets from 1,521,433 users on X between 1 September and 15 December 2023.

The groups relied on low-complexity tactics, retweet amplification and copy-paste diffusion, and the authors describe a fragmented landscape without centralised control.

The composition of what those groups pushed is reported directly. Of 191 highly amplified coordinated posts, 12 were assessed as misleading, and those were concentrated in 3 of the 11 groups; the remaining groups were engaged primarily in advocacy, religious solidarity or humanitarian mobilisation. Each group was organised around a small set of amplified accounts that drew little engagement from accounts outside the coordinated set, so the measured amplification was largely internal to it.

Two limits travel with the finding. The authors document amplification and make no finding of centralised or external financing, so the study cannot be read as evidence about who paid for anything. And 541 accounts out of 1,521,433 users is the proportion the study establishes; it does not establish how many people encountered coordinated material, which is a different measurement the paper does not make.

Ng and Carley (2023, peer-reviewed) supply the general baseline that keeps the coordination question in proportion. Applying a combined synchronization index to six political and social activism events on Twitter, covering about 1.8 million unique users and 4.3 million original tweets, they found that in most of the six events human users scored higher on synchronization than bots, with bot-and-human pairs the most synchronised pair type. Their conclusion is that synchronised posting around contentious events is predominantly organic. The six events do not include the Israel-Hamas conflict, so the result is mechanism evidence rather than a measurement of this case.

On the available measurement, then, the justification vocabulary spread mostly through ordinary accounts, with a small, detectable, internally amplified coordinated layer inside it.

Do justification attitudes track measured antisemitic belief?

Enstad (2026, peer-reviewed) measured it directly: across two nationally representative Norwegian surveys, N = 1,575 and a preregistered replication of N = 1,653, endorsement of Holocaust inversion, the belief that Israel treats Palestinians as badly as Jews were treated during the Second World War, was associated with justification of harassment and violence against Jews, and with refusal to take a stance against such aggression.

The same study found endorsement of Holocaust inversion associated with refusal to answer questions measuring blatant antisemitic prejudice, which is a measurement caveat as much as a finding: non-response is itself patterned, so the association is estimated on respondents who answered. The design is cross-sectional and the language is associational; the replication establishes that the pattern reproduces, not that one attitude produces the other.

Moon, Barlev and Neuberg (2026, peer-reviewed) tested how a justification changes reactions to a person expressing antisemitism, across three preregistered US experiments totalling 979 participants. Absent any justification, people expressing antisemitism were generally disliked, more so by liberals than conservatives. When the same antisemitism was justified by disapproval of Israel and the war in Gaza, or by Israeli violations of Palestinian human rights, the speaker was liked more by liberals and not by conservatives, an effect the authors describe as licensing. The measured outcome is how much participants liked a person expressing antisemitism, not a claim that disapproval of Israel is itself antisemitism.

Harber and colleagues (2026, peer-reviewed) report the mediation structure across three preregistered studies (N = 373, N = 243, and a three-wave longitudinal N = 493), and it is specific about which beliefs carry the association: anti-Zionist conspiracy belief mediated the link between antisemitism and anti-Israel attitudes, while conspiracy beliefs about Jews unrelated to Israel did not, and a general conspiratorial mindset did not explain the results.

Their model accounted for over 55% of the variance in anti-Israel attitudes, an association measured in preregistered designs and reported by the authors as correlational.

How did Jewish students describe encountering these justifications?

253 Jewish college and university students, surveyed weekly in early 2024 across 812 responses, showed within-person increases in antisemitism-related stress and in stress attributable to the Israel-Palestine conflict each independently associated with heightened depressive symptoms (Morstead and DeLongis, 2025, peer-reviewed).

The same longitudinal data identifies which responses tracked with worse outcomes. Increased avoidance coping was associated with heightened depressive symptoms, while approach coping, seeking support or information, was protective. Students in the United States reported higher depressive symptoms than those in Canada. The associations are within-person and correlational, which is the strength of the weekly design: each student serves as their own comparison across weeks rather than being compared to other students.

The vocabulary question was studied separately. A 2025 article in Socius (SAGE, peer-reviewed) applies neutralization theory, the sociological framework describing the terms by which harmful acts are rendered acceptable to those who accept them, to how Jewish college students in the United States described the justifications they encountered around the violence in Gaza. The framework is the same intellectual family as Cohen’s denial typology in the first section, applied at the level of received discourse rather than published claims.

Self-reported exposure figures place the campus cohort in a wider age pattern. In the American Jewish Committee’s 2025 survey, an advocacy organisation reporting its own paired samples of American Jews and the general public, 47% of Jewish adults aged 18 to 29 said they had been personally targeted by antisemitism in the past year, against 28% of those aged 30 and older, and 87% against 70% reported online antisemitism. Those are self-reported experiences within that organisation’s survey samples, not incident counts and not a measured population rate.

What the student evidence establishes is an aggregate association between encountering this discourse and measured psychological symptoms, in defined samples and dated windows. It does not identify which specific justification frame, of the three catalogued above, students were responding to; the studies measure exposure and stress, not frame attribution.

Did recorded incidents rise while the attack was being justified?

1,205 antisemitic incidents were reported to the Community Security Trust across the United Kingdom in the 35 days from 7 October 2023, according to the Trust’s own provisional tally on its own definitions.

This compares with 191 incidents reported to the Trust over the same dates in 2022. The Trust described the total at publication as the highest 35-day total ever reported to it.

Bundesverband RIAS verified 202 antisemitic incidents in Germany connected to the attacks between 7 and 15 October 2023, against 59 documented nationwide in the same window of 2022, and stated the increase as at least 240% because roughly 300 reports had arrived and only the verified subset was counted by its cut-off. Of the 202, 183 were classed as abusive behaviour and 6 as attacks; political background could not be assigned in 64% of cases, so the count carries no perpetrator profile.

The Vienna reporting centre of the Israelitische Kultusgemeinde recorded 76 incidents in the first 13 days, a rapid partial count, and its settled figures for the year put October 2023 at 200 incidents and calendar 2023 at 1,147 against 719 in 2022, with a daily average of 1.55 before 7 October and 8.31 after. In the United States, the Anti-Defamation League counted 9,354 incidents in 2024, the highest in the 46 years of its own audit, alongside its own disclosure that a definitional change adopted after October 2023 added 1,350 incidents, 15%, to the 2023 total and carries into 2024.

These are four organisations counting on four sets of definitions over four windows, and none of them is an official crime statistic. They cannot be summed, ranked against each other, or set beside police figures. What they establish is that recorded incidents rose sharply in the same days that the justification vocabulary was circulating, an association in time and nothing more; no study here measures whether the discourse moved the counts.

Methodology and limitations

This report measures the discourse and the documentary record around the attack of 7 October 2023: what justifications were stated, by whom, in what dated texts, what the treaty instruments and investigative records say in reply, and what has been measured about the discourse and its correlates. It does not adjudicate any contested claim about the conflict that the justification vocabulary invokes.

Source types, by tier. Primary and official records: the Geneva Conventions of 12 August 1949 (Common Article 3), Additional Protocol I of 8 June 1977 (Articles 1(4), 51(2), 51(7) and 51(8)), the Rome Statute of 1998 (Articles 7 and 8), and the Congressional Research Service In Brief R47828. Peer-reviewed research: the 2025 articles in Israel Affairs, Studies in Conflict & Terrorism, the Israel Law Review (Cambridge University Press), the Journal of Conflict and Security Law (Oxford University Press), Ethnopolitics and Socius; and Becker and colleagues (2022), Mendelsohn and colleagues (2023), Ng and Carley (2023), Morstead and DeLongis (2025), Enstad (2026), Moon and colleagues (2026), Harber and colleagues (2026). Preprints, labelled at first use in the body: Guerra and colleagues (2024), Chen and colleagues (2024), Elmas and colleagues (2026).

Party-published material is labelled as such throughout: Hamas’s January 2024 document, the Israeli Ministry of Foreign Affairs hostage figures of September 2025, and the Begin-Sadat Center for Strategic Studies casualty totals. None of these is independently verified, and each is cited as that body’s own account or count.

Monitoring tallies (Community Security Trust, Bundesverband RIAS, the Vienna reporting centre of the Israelitische Kultusgemeinde, the Anti-Defamation League, the American Jewish Committee) rest on voluntary reporting against each organisation’s own definition, are not official crime statistics, and are not additive across organisations. The CST and RIAS figures for October 2023 were published as provisional. The ADL total carries that organisation’s own disclosed definitional change.

Three specific limits govern how far the findings reach. Becker and colleagues (2022) coded comments on the May 2021 escalation, not the 2023 attack, so it measures the framing vocabulary at an earlier event. Ng and Carley’s six activism events do not include this conflict. And no study located here measures what share of 7 October discussion recast Israeli civilians as combatants, nor whether the justification discourse moved recorded incident counts; every temporal statement in this report is correlational.

Conclusion

The question people ask about 7 October 2023 is not usually a legal one. It is closer to bewilderment: how did an argument for the day arrive so quickly, and what was in it? The answer on the record is that it arrived in three catalogued shapes, and each is answered by a different kind of evidence.

Set the whole record side by side and it holds together tightly. The 2025 peer-reviewed literature names the three categories and quotes the endpoint of the third, the perpetrators as resistance fighters rather than terrorists. Hamas’s January 2024 document supplies the dated grievance account and the factual denial in one text, that party’s own assertions about its own conduct. The Israel Law Review’s 2025 analysis carries more than 1,200 killed in a day, with torture, sexual violence and mutilation among the documented conduct, and Israel’s own ministry counts 207 hostages returned or recovered and 48 still held as of September 2025. The instruments that the resistance framing appeals to are the same instruments that answer it: Additional Protocol I recognises conflicts against alien occupation in Article 1(4), forbids making civilians the object of attack in Article 51(2), and in Article 51(8) says in as many words that one party’s violations release no other party from its obligations to civilians. On the measurement side, the coordinated layer detected in 4.57 million tweets came to 541 accounts in 11 groups, so the vocabulary travelled mostly through ordinary accounts; and where survey researchers have looked, the justification attitudes track measured antisemitic belief, in two Norwegian national samples, in three US experiments on licensing, and in a mediation model accounting for over 55% of variance.

One thing in that list is documented and one is not. The vocabulary is on the record with dates and speakers, the associations are measured with confidence intervals and preregistrations, and the 1,205 incidents the Community Security Trust logged in 35 days sit in the same weeks as the argument, with nothing in this record establishing a path between them.

Which leaves the part the numbers reach up to and stop. An explanation that makes a mass killing intelligible does not stay in the argument where it was made; it becomes something a student hears in a seminar room, something a comment thread supplies as context, something the surrounding conversation carries without saying. The trope that Jews are the authors of what happens to them is older than any of these texts. Is the vocabulary catalogued here one of the channels through which that older thing now moves, and if it is, what does a society do with an argument that was ready before the funerals?

Frequently Asked Questions

Where does the three-part denial typology come from?

Stanley Cohen developed the literal, interpretive and implicatory categories in “States of Denial” (2001) to describe how bystanders, officials and publics respond to atrocity generally. The 2025 Israel Affairs and Studies in Conflict & Terrorism articles apply that existing framework to the 7 October attack rather than devising a new one.

Do official crime statistics show the same pattern as community monitoring tallies?

The FBI’s Uniform Crime Reporting programme recorded 1,938 anti-Jewish single-bias incidents in the United States in 2024, the highest since collection began in 1991 and up 5.8% from 1,832 in 2023, reported by 16,419 participating agencies covering 95.1% of the population. It is an official record and a known undercount: only 3,127 of those agencies (19.0%) reported any hate crime at all.

How much hate crime never reaches a police report?

About 42% of violent hate-crime victimizations went unreported to police between 2015 and 2019, per the Bureau of Justice Statistics National Crime Victimization Survey (2021). That survey estimated 305,390 hate-crime victimizations in 2019, a figure that is not directly comparable with FBI incident counts because the two use different methods, units and definitions.

Why can incident counts from different countries not be added together?

Each monitor counts on its own definition over its own window: the Community Security Trust states its definition is narrower than the criminal justice system’s and mixes crimes with non-crime incidents, while Bundesverband RIAS classifies against categories broader than criminal law. The Anti-Defamation League and the FBI likewise count from different universes, ADL’s own definitions against criminal reports, so their totals are never summed or compared directly.

Is coded language built to evade automated moderation?

Mendelsohn and colleagues (2023, peer-reviewed conference) assembled a glossary of over 300 coded terms and found that harmful content containing them evaded automated toxicity detection. Taylor, Peignon and Chen (2017, preprint) document the same design in the 2016 substitution scheme in which ordinary words such as “skypes” and “googles” carried an alternate hate-speech meaning.

How do younger and older respondents differ on overt antisemitic statements?

In the Yale Youth Poll’s Fall 2025 national survey of US registered voters, 10% of voters aged 18 to 34 agreed with all three antisemitic statements tested, against 2% of voters aged 65 and older, while two-thirds of voters rejected all three. Hersh and Royden (2022, peer-reviewed), surveying about 3,500 US adults, similarly found the highest concentration of overt antisemitic attitudes among young adults on the far right.

What do surveys record about Holocaust knowledge in the cohorts now encountering this discourse?

The Conference on Jewish Material Claims Against Germany’s 2020 survey of 11,000 US adults aged 18 to 39 found 63% did not know that six million Jews were murdered in the Holocaust and 36% thought two million or fewer were killed, while 90% said they believed the Holocaust happened. In the same survey, 11% believed Jews caused the Holocaust.

Sources

• American Jewish Committee, 2025. The State of Antisemitism in America 2025. ajc.org. Monitoring organization (advocacy mission), survey.
• Anti-Defamation League Center on Extremism, 2025. Audit of Antisemitic Incidents 2024. adl.org. Monitoring organization (advocacy mission).
• Becker, M., Ascone, L., Troschke, H., 2022. Antisemitic Comments on Facebook Pages of Leading British, French, and German Media Outlets. Humanities and Social Sciences Communications. DOI 10.1057/s41599-022-01337-8. Peer-reviewed.
• Begin-Sadat Center for Strategic Studies, 2025. The Gaza Terror Offensive, situation report to 5 September 2024. besacenter.org. Research institute.
• Bundesverband RIAS, 2023. Monitoring: Antisemitische Reaktionen in Deutschland auf die Hamas-Massaker in Israel, 7 to 15 October 2023, published 18 October 2023. report-antisemitism.de. Monitoring organization.
• Bureau of Justice Statistics, 2021. Hate Crime Victimization, 2005 to 2019. National Crime Victimization Survey, US Department of Justice. Official record.
• Chen, K., He, Z., Burghardt, K., Zhang, J., Lerman, K., 2024. IsamasRed: A Public Dataset Tracking Reddit Discussions on the Israel-Hamas Conflict. arXiv:2401.08202. Preprint.
• Cohen, S., 2001. States of Denial: Knowing About Atrocities and Suffering. Polity Press. Scholarly book.
• Community Security Trust, 2023. Antisemitic incidents, 10 November update. cst.org.uk. Monitoring organization.
• Community Security Trust, 2024. Antisemitic Incidents Report 2023. cst.org.uk. Monitoring organization.
• Conference on Jewish Material Claims Against Germany, 2020. First-Ever 50-State Survey on Holocaust Knowledge of American Millennials and Gen Z. claimscon.org. Organization survey.
• Congressional Research Service. Israel and Hamas Conflict In Brief: Overview, U.S. Policy. Report R47828, congress.gov. Official record.
• Elmas, T., Silva, S., Pote, M., Dey, K., Chang, H., Ye, Y., Luceri, L., Buntain, C., Ferrara, E., Flammini, A., Menczer, F., 2026. Israel-Hamas War on X: A Case Study of Coordinated Campaigns and Information Integrity. arXiv:2604.10566. Preprint.
• Enstad, J., 2026. Holocaust Inversion and Justification of Anti-Jewish Aggression: Evidence from Two National Surveys in Norway. American Psychologist. DOI 10.1037/amp0001704. Peer-reviewed.
• Ethnopolitics (Taylor & Francis), 2025. Analysis of “Zero Hour” framing of 7 October 2023. Peer-reviewed journal.
• Federal Bureau of Investigation, 2025. Reported Crimes in the Nation, 2024, Hate Crime Statistics. Uniform Crime Reporting Program, Crime Data Explorer. Official record.
• Geneva Conventions of 12 August 1949, Common Article 3. Treaty text. Primary record.
• Guerra, A., Lepre, M., Karakuş, B., 2024. Quantifying Extreme Opinions on Reddit Amidst the 2023 Israeli-Palestinian Conflict. arXiv:2412.10913. Preprint.
• Hamas, 2024. Published document on the operation of 7 October 2023, released January 2024, as recorded in Congressional Research Service R47828 and in Studies in Conflict & Terrorism (2025). A party’s own published account.
• Harber, K., Bulska, D., Malloy, T., Vila, J., 2026. Antisemitism, Conspiracy Beliefs, and Anti-Israel Attitudes. American Psychologist. DOI 10.1037/amp0001635. Peer-reviewed.
• Hersh, E., Royden, L., 2022. Antisemitic Attitudes Across the Ideological Spectrum. Political Research Quarterly. DOI 10.1177/10659129221111081. Peer-reviewed.
• Institute for Strategic Dialogue, 2025. Amplifying Antisemitism: How Recommender Algorithms Serve Harmful Content to Children. isdglobal.org. Established research organization.
• Israel Affairs (Taylor & Francis), 2025. Terrorism denial: 7/10 and the reasons to deny a pogrom. DOI 10.1080/13537121.2025.2516864. Peer-reviewed.
• Israel Law Review (Cambridge University Press), 2025. Hamas’ October 7th Genocide: Legal Analysis and the Weaponisation of Reverse Accusations. Peer-reviewed.
• Israel Ministry of Foreign Affairs, Permanent Mission to the Multilateral Organizations in France, 2025. Israel’s Objectives and Current Operation, figures stated as of 14 September 2025. embassies.gov.il. Official record, a party’s own count.
• Israelitische Kultusgemeinde Wien, Antisemitismus-Meldestelle, 2023 and 2024. Press release of 23 October 2023 and Antisemitic cases 2023 in Austria. antisemitismus-meldestelle.at. Monitoring organization.
• Journal of Conflict and Security Law (Oxford University Press), 2025. Legal frameworks for mass terrorism: the October 7th attack. Volume 30, issue 2, page 287. Peer-reviewed.
• Journal of Genocide Research (Taylor & Francis), 2025. DOI 10.1080/14623528.2025.2483546. Peer-reviewed.
• Mendelsohn, J., Le Bras, R., Choi, Y., Sap, M., 2023. From Dogwhistles to Bullhorns: Unveiling Coded Rhetoric with Language Models. ACL 2023, arXiv:2305.17174. Peer-reviewed conference.
• Moon, J., Barlev, M., Neuberg, S., 2026. Justifying Antisemitism: Political Liberalism and Perceptions of Prejudices. American Psychologist. DOI 10.1037/amp0001702. Peer-reviewed.
• Morstead, T., DeLongis, A., 2025. Antisemitism on Campus in the Wake of October 7: Examining Stress, Coping, and Depressive Symptoms Among Jewish Students. Stress and Health. DOI 10.1002/smi.3529. Peer-reviewed.
• Ng, L., Carley, K., 2023. A Combined Synchronization Index for Evaluating Collective Action on Social Media. Applied Network Science. DOI 10.1007/s41109-022-00526-3. Peer-reviewed.
• Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of International Armed Conflicts (Protocol I), 8 June 1977, Articles 1(4), 51(2), 51(7) and 51(8). Treaty text. Primary record.
• Rome Statute of the InternationalCriminal Court, 17 July 1998, Articles 7(1), 8(2)(b)(i) and 8(2)(c)(iii). Treaty text. Primary record.
• Socius (SAGE), 2025. Neutralization and Violence in Gaza: How Jewish College Students in the United States Describe the Justifications They Encountered. DOI 10.1177/23780231251340798. Peer-reviewed.
• Studies in Conflict & Terrorism (Taylor & Francis), 2025. The Denial of Hamas’ Atrocities. DOI 10.1080/1057610X.2025.2494284. Peer-reviewed.
• Taylor, J., Peignon, M., Chen, Y., 2017. Surfacing Contextual Hate Speech Words within Social Media. arXiv:1711.10093. Preprint.
• Yale Youth Poll, 2025. Fall 2025 Poll on Antisemitic Attitudes and Israel-Palestine. Yale University, youthpoll.yale.edu. Academic poll.