Why Can Ordinary People Massacre and Still Feel Righteous?

NAZA, the documentary by Yuval Abraham and Rachel Szor, opened in Indonesian cinemas on 16 October 2026 through KlikFilm, after winning the Special Jury Prize at the 83rd Venice International Film Festival. Its evidence is not a death toll. It is a vocabulary, described by 24 Israeli soldiers and intelligence personnel interviewed at night on rooftops in Tel Aviv over three years, with faces shown as silhouettes and voices altered.

The title is the first exhibit. According to the film, NAZA is the Israeli military acronym for nezek agavi, Hebrew for collateral damage, used as a number attached to a potential target: the expected count of civilians killed if the target is struck. An interviewee quoted in the film states the grammar of it plainly: “We say 15 NAZA. We don’t say 15 people killed and certainly not that 15 children and women will be killed.” According to testimony presented in the film, values below 20 required no special authorisation, while higher values could be approved, up to 500 in one case. Interviewees also describe an artificial intelligence system called Lavender, which one of them calls “a factory of targets”, and a practice of checking a child’s phone to establish when the father is at home, referred to in the film as “Where’s Daddy?”.

None of that framing is new to the study of atrocity. In 1973, Herbert Kelman described three conditions under which violence proceeds without moral restraint: authorisation, by which responsibility moves upward; routinisation, by which killing becomes ordinary work; and dehumanisation, by which victims leave the moral community. Albert Bandura catalogued the linguistic half of the mechanism in 1999, listing euphemistic labelling among the ways people disengage from their own conduct. Zygmunt Bauman argued in 1989 that bureaucratic distance, rather than sadism, made the Holocaust administratively possible. Nick Haslam and Steve Loughnan showed in 2014 that dehumanisation rarely needs to be total: denying a group depth of feeling is often enough. These studies were written about other places and other decades. They describe a mechanism, not a verdict on any person who appears in this film.

Israel rejects the film’s account. The IDF Spokesperson’s Unit said it relies on anonymous testimonies “from individuals whose identities cannot be verified, including whether they served in the IDF, in what capacity, or whether they were involved in the matters about which the allegations are being made”. The military said claims about national and military strategy were “clearly beyond the knowledge of junior-level soldiers”, and that decisions on the selection and approval of strikes “are made by human personnel only, in accordance with IDF guidelines, and not by artificial intelligence”.

The consequence of a word like NAZA is documentary before it is moral. Armed forces keep records, and the record fixes what an institution knows about itself. A file that carries an expected civilian figure, rather than a list of expected dead, produces statistics that are internally consistent and externally unreadable. It also relocates disagreement: where the account of a strike exists as a number, contesting it requires the log, the approval chain and the message traffic, not only the graves. That is why the dispute over this film turns on the verification of anonymous testimony, the point the military raises and the point still unresolved.