The Apollo: SAVAK's Torture Record

The Room

The place had a proper name: the Anti-Sabotage Joint Committee. Iranians knew it as Komiteh-ye Moshtarak — the Joint Committee. It was housed in a building in central Tehran that later became the Ebrat Museum, a grim showcase the Islamic Republic maintained to document the Shah’s crimes while committing far worse ones of its own.

Inside Komiteh-ye Moshtarak, SAVAK’s Third Directorate — the internal security branch, directed by Parviz Sabeti — conducted interrogations of suspected guerrillas, leftists, and political dissidents. It was here that the methods were refined, tested, and systematized.

The methods were real. They are documented not by opposition propaganda but by the International Committee of the Red Cross, by academic historians working from declassified files, and by the Islamic Republic’s own post-revolutionary investigations (which, while politically motivated, corroborated the physical evidence).¹

What follows is the forensic record. It is not pleasant. It is necessary — because the credibility of every statistical argument in this series depends on the willingness to document real abuses with the same rigor applied to debunking inflated numbers.

The Methods

The Bastinado

The oldest method — whipping the soles of the feet with cables or rods. Effective for producing extreme pain without visible external injury or lethal damage. A method shared across authoritarian regimes in the region, from the Ottoman Empire through the twentieth century.

The Apollo

The device that gave this article its name. A metal helmet placed over the prisoner’s head, designed to amplify the wearer’s own screams while electric shocks were administered to sensitive areas of the body. The sensory feedback loop — hearing your own amplified screaming while experiencing pain — was calculated to accelerate psychological breakdown.

The Apollo was not crude. It was engineered. It represented the application of behavioral science to interrogation — a sophistication that distinguished SAVAK’s later methods from the improvised violence of its early years.

The Hot Grill

An electrically heated metal bed frame. The prisoner was strapped to the frame while the metal was heated, burning the skin of the back and legs. Survivors described the smell of their own burning flesh as the most disorienting element — the brain struggled to reconcile familiar sensory input (the smell of cooking) with the source (oneself).

Psychological Methods

Beyond the physical devices, SAVAK employed a repertoire of psychological techniques: extended sleep deprivation, mock executions staged with convincing detail, and threats against family members — including credible threats to arrest, interrogate, or harm wives, children, and parents.

Allegations of sexual violence — including rape and sexual humiliation of both male and female prisoners — appear consistently in survivor testimonies, opposition literature, and academic studies of the period. The use of snakes and sexual humiliation against female prisoners has been reported by survivors of both the Fedayeen and MEK, and verified in scholarly research.

The Turn

SAVAK’s methods evolved over time. The turning point was 1971.

On February 8, 1971, a group of Fedayeen-e Khalq guerrillas attacked a gendarmerie post in the village of Siahkal, in the forests of northern Iran. The attack was militarily insignificant — the guerrillas were quickly routed — but it marked the beginning of armed insurgency against the Shah’s government.

Before Siahkal, SAVAK dealt with dissidents through a mix of surveillance, intimidation, and crude physical violence. The political opposition was largely intellectual — student activists, writers, lawyers. SAVAK’s response was calibrated accordingly: arrest, interrogation, occasional beatings, release or imprisonment.

After Siahkal, the adversary changed. The Fedayeen and MEK were trained fighters. They conducted bank robberies, assassinations, and bombings. They were willing to die. Standard interrogation techniques failed to extract operational intelligence from committed revolutionaries.

SAVAK’s response, as documented by historian Ervand Abrahamian in Tortured Confessions, was to professionalize its brutality. Interrogators were “sent abroad for ‘scientific training’ to prevent ‘unwanted deaths from brute force.’”² The shift was explicitly pragmatic: dead prisoners yielded no intelligence. The goal was to break the will without killing the subject — to extract information about networks, safe houses, weapons caches, and planned operations.

This “scientific” approach did not make the torture less cruel. It made it more effective. And it concentrated the worst methods at the Joint Committee, where the highest-value detainees were interrogated.

The Verification

The evidence for SAVAK’s torture comes from multiple independent sources with different biases — making the convergence significant.

The ICRC (1977): When President Jimmy Carter’s human rights policy pressured the Shah to permit international inspection, the International Committee of the Red Cross was granted access to Iranian prisons. Their confidential reports to the Shah noted “evidence of physical mistreatment” consistent with torture.³ The ICRC’s findings were diplomatic in language but unambiguous in substance. The Shah responded by ordering a halt to torture in late 1977 — a concession that came too late to change the political trajectory.

Ervand Abrahamian: The Columbia University historian cross-referenced SAVAK’s own files (captured after the revolution) with opposition claims and survivor oral histories. His conclusion — that “SAVAK used torture systematically but killed relatively few” — represents the academic consensus. The torture was real; the death toll was lower than the mythology.

Amnesty International: While Amnesty’s numerical estimates of political prisoners proved unreliable (their “25,000 to 100,000” figure was based on exile testimony and later contradicted by ICRC inspection), their qualitative reporting on torture methods was corroborated by other sources. The descriptions of techniques matched what the ICRC found and what survivors consistently reported.

Parviz Sabeti: The director of SAVAK’s Third Directorate, the man most directly responsible for internal security operations, acknowledged after 1979 that “harsh interrogation methods” were used. While his post-revolution testimony sought to minimize the scope, his acknowledgment of the basic fact — that SAVAK tortured detainees — constituted an admission from the person in the best position to know.

The Training

SAVAK did not develop its methods in isolation. The agency was created in 1957 with direct assistance from the CIA, which provided basic intelligence tradecraft training to the initial cadre from 1957 to 1961.⁴ As the Shah grew distrustful of American intentions, the training relationship shifted toward Israel’s Mossad, which focused on counter-intelligence and internal security.

The extent of foreign involvement in SAVAK’s interrogation methods remains debated. Declassified CIA documents confirm the training relationship but provide limited detail on specific techniques transferred. What is clear is that SAVAK’s post-1971 “scientific training” — wherever it was obtained — produced a level of methodological sophistication that distinguished it from the improvised brutality of many contemporaneous security services.

This foreign involvement is a genuine stain on the records of the governments that provided it. Acknowledging it does not weaken the argument about inflated death tolls. It strengthens it — because the argument depends on confronting the full record, not a sanitized version.

The Scale

Here is where the forensic record and the popular mythology diverge most sharply.

SAVAK’s torture was systematic. This means it was policy, not the random cruelty of individual officers. It was applied to suspected guerrillas, political organizers, and members of armed opposition groups. It was concentrated at specific facilities — the Joint Committee foremost among them. It was documented by multiple independent sources.

SAVAK’s killing was limited. Over sixteen years, Baghi’s forensic audit for the Martyrs Foundation documented 3,164 deaths total, of which 42 resulted from torture during the guerrilla war period (1971-1977). The political prisoner population, verified by ICRC physical inspection, was approximately 3,200 at any given time.

These two facts coexist without contradiction. A security service can be systematically brutal in its methods while remaining limited in the number of people it kills. The Stasi tortured thousands through what it called Zersetzung — psychological decomposition — while directly killing very few. SAVAK operated in a similar mode: extreme cruelty applied to a relatively small number of high-value targets, amplified by reputation into the perception of omnipresent violence.

The Rebranding

When the Islamic Republic consolidated power in 1979, it executed SAVAK’s top leadership. General Nematollah Nassiri, the agency’s chief, was among the first put to death by revolutionary tribunals. General Moghadam followed.

But the technical personnel — the interrogators, the analysts, the specialists who understood the craft of intelligence work — were retained. SAVAK was renamed SAVAMA, which later became MOIS (the Ministry of Intelligence and Security, also known as VEVAK).

The Islamic Republic inherited SAVAK’s methods, its institutional knowledge, and its personnel.⁵ What it did not inherit were the Shah’s inhibitions.

The Shah, for all his authoritarianism, operated within constraints. He cared about Western opinion. He responded to Jimmy Carter’s human rights pressure by allowing the ICRC into his prisons. He ordered a halt to torture in 1977. These constraints were insufficient — the torture continued longer than they should have allowed, and the political space closed faster than reform could open it. But the constraints were real.

The Islamic Republic discarded them entirely. The same interrogation techniques were applied without the Shah’s concern for international legitimacy, without the ICRC inspections, and without the constraint of keeping prisoners alive for intelligence purposes. The result was a system that killed more political prisoners in its first four years (8,000-9,500 executed from 1981-1985) than SAVAK killed in its entire existence.

The tools were the same. The restraints were gone.

Why This Matters

This article exists because of a principle embedded in the persuasion research: costly signaling builds credibility.

Documenting SAVAK’s real abuses — the Apollo, the Hot Grill, the bastinado, the systematic application of psychological terror — is not a detour from the argument about inflated death tolls. It is the foundation of that argument. A source that acknowledges genuine crimes is trusted when it challenges inflated numbers. A source that whitewashes or minimizes real abuses forfeits the right to challenge anything.

The Shah’s Iran used torture. This is documented, verified, and beyond dispute. SAVAK’s methods were cruel, sophisticated, and applied to real human beings who suffered real harm.

And the number of people SAVAK killed — verified by the Islamic Republic’s own Martyrs Foundation at 3,164 over sixteen years — is a fraction of what the regime that replaced it has done.

Both facts are true. Both facts matter. And the ability to hold both simultaneously — without minimizing torture to inflate the Shah’s reputation, and without inflating the death toll to justify the revolution’s mythology — is what distinguishes forensic analysis from propaganda.

This article is a companion to The Number That Changed Everything. For the full methodology behind the 3,164 figure, see The Baghi Paradox. For the hour-by-hour reconstruction of Black Friday, see Black Friday: 88.