The Civilian Shield 

Punjab ’95 and the Silence of KBS Sidhu

The Censored Film, the Cremation Registers, and the Administrative Silence that Made Both Possible

A Forensic Historical Audit of Punjab, 1975–2026

Investigative Edition

By Dr. Kanwar Partap Singh Gill

Based on judicial records, NHRC proceedings, official government rosters, human-rights commission documentation, primary statutory sources, and archival research spanning five decades.

Author’s Note

The author is a practicing family medicine physician and a native of Amritsar. This article is a public-education research synthesis based on publicly available records, judicial materials, official institutional reporting, human-rights investigations, and analytical frameworks in the working documents from which it is drawn. It does not purport to establish criminal liability against any person beyond what competent courts or official bodies have adjudicated. Where analysis goes beyond adjudicated fact, it is identified as historical interpretation, institutional critique, or archival inquiry.

As a physician, I view the absence of magisterial inquests as a failure of the state’s forensic pulse. If the police are the muscle of the state, the District Magistrate is the nervous system—tasked with sensing and reporting trauma. In Amritsar between 1984 and 1996, that nervous system was not merely neglected. It was deliberately numbed. This manuscript is my clinical report on that numbing.

This manuscript functions simultaneously as historical investigation, forensic administrative audit, human-rights research synthesis, and narrative history accessible to any reader who wishes to understand how a democratic state can destroy thousands of its own citizens—not despite its bureaucracy, but through it.

The full research archive is published at kpsgill.com. A companion manuscript tracing the history of crimes against the Sikh nation from 1900 through 2025 is also available there, as is a rebuttal to the Hindu American Foundation’s selective framing of the Khalistan question.

Executive Summary: The Civilian Shield Thesis

This audit examines a structural phenomenon in the administration of Amritsar district between 1975 and 1996: the systematic failure of the office of the Deputy Commissioner and District Magistrate to exercise its non-delegable statutory obligations under the Code of Criminal Procedure, 1973, while simultaneously exercising broad powers of land acquisition, preventive detention, annual confidential reporting, and public-order certification. That combination—the active deployment of coercive administrative powers alongside the passive non-deployment of oversight obligations—constitutes what this audit terms the Civilian Shield.

The Civilian Shield did not require any individual officer to order extrajudicial violence. It required only that the office refrain from generating the documentary friction the law placed it in position to create. The result was that 2,097 persons—categorized by the Central Bureau of Investigation as 585 identified, 274 partially identified, and 1,238 entirely unidentified—were cremated by Punjab Police in Amritsar district between 1984 and 1994 without the civilian oversight architecture of the district producing a publicly documented, proportionate audit response.

The Supreme Court of India, by order dated 12 December 1996, remitted this matter to the National Human Rights Commission. The Commission’s annual report for 2007–08 characterized the proceedings as involving “flagrant violations of human rights on a mass scale.” The CBI report, which remains partially sealed, is described in Human Rights Watch’s 2007 report Protecting the Killers as listing 2,097 illegal cremations at three cremation grounds in Amritsar district alone. Jaswant Singh Khalra, the human-rights investigator who first surfaced the municipal record of these cremations, was abducted by Punjab Police on 6 September 1995 and murdered. Six police officers were convicted of his murder in 2005. He was abducted during the tenure of Karanbir Singh Sidhu as Deputy Commissioner and District Magistrate of Amritsar.

This audit proceeds through five principal findings. First, the statutory architecture of the CrPC created non-delegable custodial oversight duties that no emergency legislation, including AFSPA, TADA, or the NSA, eliminated. Second, the three Deputy Commissioners of Amritsar between 1984 and 1996—Ramesh Inder Singh (IAS, West Bengal cadre, 4 June 1984–6 July 1987), Sarabjit Singh (7 July 1987–10 May 1992), and Karanbir Singh Sidhu (11 May 1992–11 August 1996)—each held both coercive administrative powers and oversight obligations simultaneously. Third, the public record shows active exercise of the former and the near-total absence of documented exercise of the latter. Fourth, the political and administrative conditions from 1975 onward each individually and cumulatively degraded the civilian oversight culture of the district. Fifth, the institutional afterlife—Padma Shri awards, published memoir, career continuation, and geographic distance from accountability proceedings—represents what the companion research describes as administrative survival: the ability of civil authority to emerge from a morally disfigured district without proportionate public reckoning.

The film Punjab ’95, based on the life and work of Jaswant Singh Khalra, has been in censorial paralysis since December 2022. The Central Board of Film Certification has demanded more than one hundred cuts from the film, including the removal of the protagonist’s name, references to the Punjab Police, and the word “massacre.” This audit argues that the censorship of Punjab ’95 is not a cinematographic dispute. It is the most recent act in the same institutional reflex—the management of public memory by the management of public record—that produced the silence in Amritsar’s district office between 1984 and 1996.

PART ONE THE SYMPTOM: THE CENSORSHIP OF PUNJAB ’95 AND THE MODERN MECHANICS OF MEMORY CONTROL

The Film That Threatened National Order

In January 2025, something happened that almost never happens in Bollywood’s carefully managed public-relations ecosystem. A film’s director made an announcement. Three days later, he made an apology.

The announcement, posted to Instagram by director Honey Trehan, was that his film Punjab ’95 would have an international theatrical release on 7 February 2025. The lead actor, Diljit Dosanjh—at that moment one of the most commercially bankable stars in Indian popular music and cinema, a man who had just completed a sold-out stadium tour across India—confirmed the date on his own platform. The trailer went up. It was gut-wrenching, precise, and unmistakably about real things that happened to real people in a real district of a real state. Then, three days later, came the apology: “We are very sorry and it pains us to inform you that the movie Punjab ’95 will not release on 7th February due to circumstances beyond our control.” No elaboration. No named authority. No explanation of what those circumstances were. Just the polite, hollow language of a man who has been told by a powerful institution that he cannot tell you what he has been told.

The circumstances, as they emerged in reporting by The Caravan, New Lines Magazine, and Deadline, were the following. As the Bombay High Court case challenging the Central Board of Film Certification’s demands was approaching its final hearing, the film’s producer Ronnie Screwvala was summoned to New Delhi for an emergency meeting with officials of the Ministry of Information and Broadcasting. He called Trehan from that meeting. “He told me we have to do an out-of-court settlement,” Trehan recounted to The Caravan’s Jatinder Kaur Tur. “We cannot fight the government.”

“Ronnie got a call from a higher authority that it will not be taken in good taste.” — Director Honey Trehan, describing the intervention that stopped the international release of Punjab ’95

Dosanjh had already received his own communication. The film that was supposed to release internationally on 7 February 2025 did not release. It has still not received theatrical certification in India. It has been in regulatory limbo since December 2022, when it was first submitted to the CBFC. In that time, the demands made of the film have grown from 21 cuts to 45 to 65 to 85 to 120 to what multiple sources describe as 130 or more. Among those demands: remove the name of the protagonist, Jaswant Singh Khalra. Remove references to the Punjab Police as an institution. Remove the name of the state of Punjab. Change the original title, “Ghallughara”—a Punjabi word that means, simply, massacre.

This is where any honest account of the censorship of Punjab ’95 must begin: not with the film’s content, but with the state’s escalating panic at that content. Because the panic is itself the story. A state does not demand that a film remove the name of a state, the name of its police force, and the name of a man a Supreme Court judgment has already confirmed was abducted and murdered by that police force—a man whose killers have been convicted—unless the state understands, at some level, that the film is touching something far larger than one man’s biography.

What the Film Is Actually About

Punjab ’95 is, at its core, about a method. It is about a human-rights investigator named Jaswant Singh Khalra who, in 1994 and 1995, used the state’s own administrative paperwork to expose the state’s own crimes. He did this not by obtaining secret documents or bribing officials or cultivating informants within the security apparatus. He did this by reading records that the state had left in plain sight inside municipal cremation grounds, in the ordinary ledgers that a functioning administrative system generates as a matter of bureaucratic routine.

What those records showed was that the Punjab Police had been cremating human beings in industrial quantities in Amritsar district between 1984 and 1994—persons who had been in police custody, persons whose deaths had been classified as “encounter killings,” persons whose families had never been notified, persons whose bodies had been brought to cremation grounds by police officers who signed the fuel vouchers and watched the fire. Khalra calculated the number at over six thousand in Amritsar district alone. The Central Bureau of Investigation, when it later conducted its own examination, confirmed 2,097 illegal cremations at three cremation grounds in the district, of whom 1,238 were recorded as entirely unidentified.

This is why the film is threatening. It is not threatening because it depicts violence. Indian cinema has depicted violence for a century without triggering emergency ministerial interventions. It is threatening because it depicts accountability. It depicts the process by which a democratic state’s own records can be turned against the state’s own impunity. That process—Khalra’s method of reading the registers—is the most dangerous thing in the film. The rest is biography, emotion, and cinematography. The registers are evidence.

The History of Censorship in the Punjab Context

Punjab ’95 did not become the first film about the Punjab counterinsurgency to encounter the CBFC’s institutional hostility. In 2016, the Punjabi biographical film Toofan Singh, directed by Baghal Singh and starring Ranjit Bawa, depicted the torture methods of Punjab Police during the militancy period. The CBFC refused to certify it and banned it in India. It was released internationally in 2017. It was eventually allowed limited release in Punjab following legal challenges, but the ban itself, and its duration, represented a pattern that Punjab ’95 has reproduced at greater scale and with greater official intervention.

That pattern is not incidental. The CBFC’s institutional hostility to films about the Punjab counterinsurgency is not a bureaucratic quirk. It is a policy position. The state that committed the abuses depicted in these films is the same state that certifies the films. The institutional reflex—to manage what can be shown, to limit what can be named, to insist that suffering be sufficiently generic to be emotionally bearable but not sufficiently specific to be administratively dangerous—is the same reflex that operated in Amritsar’s district office between 1984 and 1996. It is the second disappearance that follows the first.

To understand why the censorship of Punjab ’95 is not merely a dispute about cinema, we must travel backwards through the decades that the film is about. We must understand the history that produced Khalra, the administrative architecture that permitted the crimes Khalra documented, and the institutional culture that then killed Khalra for documenting them. That journey begins not in 1994, when Khalra opened the registers, but in 1975, when the Emergency began to teach Punjab what its relationship to the Indian state actually was.

PART TWO METHODOLOGY OF EVIDENCE: HOW THIS INVESTIGATION IS CONSTRUCTED

Archival Research Methods and the Evidentiary Standard

This investigation applies a documentary standard consistent with institutional accountability inquiry rather than criminal prosecution. The operative question throughout is not whether any individual officer is criminally liable—a standard that requires judicial adjudication beyond the scope of this document—but whether the office of the Deputy Commissioner generated the statutory documentary outputs the law required of it. Absence of documentation where documentation was legally mandated is itself a material finding. This approach is consistent with the methodology employed by transitional justice commissions, parliamentary select committee inquiries, truth and reconciliation processes, and institutional audit frameworks internationally.

The investigation draws on five categories of source material, ranked in a formal evidentiary hierarchy to ensure that factual claims are distinguished from interpretive inferences. Tier One comprises judicial and statutory sources: the Supreme Court of India’s order of 12 December 1996 remitting the Punjab Mass Cremation case to the NHRC; the judgment in Prithipal Singh & Others v. State of Punjab (2012), which records the judicial history of the Khalra murder case and the conviction of six police officers; the primary text of the Armed Forces (Punjab and Chandigarh) Special Powers Act, 1983; the Code of Criminal Procedure, 1973, with particular attention to Sections 47, 57, 58, 107, 109, 131, 132, 167, 174, 176, and 218; and the Terrorist and Disruptive Activities (Prevention) Act, 1987.

Tier Two comprises official institutional sources: the National Human Rights Commission’s Annual Report for 2007–08, which contains the most authoritative public summary of the Punjab Mass Cremation proceedings and the 2,097-body figure; the Ministry of Home Affairs Annual Report for 1991–92; the Government of Punjab’s official Amritsar Deputy Commissioners Roster; and the SGPC’s White Paper on the Operation Blue Star events. These are documents the state produced about itself. Their significance lies partly in what they say and partly in what they conspicuously omit.

Tier Three comprises corroborated human-rights documentation: Human Rights Watch’s 2007 report Protecting the Killers: A Policy of Impunity in Punjab, India; the Ensaaf–Human Rights Data Analysis Group’s 2009 quantitative analysis Calculating Disappearances in Punjab, India; Human Rights Watch and Physicians for Human Rights’ 1994 report Dead Silence; Amnesty International’s various country reports on India’s Punjab between 1988 and 1996; and the Ensaaf accountability database, which documents individual perpetrators and case counts with transparent methodology.

Tier Four comprises community archive, press, and memoir sources: Tribune reporting on the Galiara project, the Punjab Police promotion records in Babushahi.com’s archive, Ramesh Inder Singh’s memoir Turmoil in Punjab: Before and After Blue Star (HarperCollins India, 2022), Sarabjit Singh’s Operation Black Thunder (SAGE Publications, 2002), community memorial archives documented in the endnotes [EN-1] through [EN-6], and the published reportage of The Caravan, New Lines Magazine, and Deadline on the Punjab ’95 censorship proceedings. Tier Five comprises interpretive analysis: historical scholarship, institutional critique, and archival inference, which are clearly labelled as such throughout the text.

Reading Khalra’s Method: The Register as Evidence

Jaswant Singh Khalra did something that is both obvious in retrospect and remarkably rare in practice. He understood that the state’s bureaucratic need to process the dead—to account for fuel, to register cremations, to enter the unidentified into ledgers—would leave behind documentary seams that official silence could not fully erase. The cremation grounds of Amritsar, Majitha, and Tarn Taran were not classified facilities. They were municipal institutions, staffed by municipal employees, maintaining the ordinary registers of civic administration. The firewood was purchased through procurement channels. The costs were recorded in accounts. The entries were made in ledgers that were, in principle, open to inspection.

What Khalra grasped was that the gap between what the records said and what the law required those records to reflect was itself the evidence of the crime. A cremation register that records the burning of an “unidentified” body, in a district under active counterinsurgency, in which hundreds of persons were known to have disappeared in police custody, without any corresponding Section 176 magisterial inquiry into how that body came to be at the cremation ground: this is not merely a statistical anomaly. It is a documentary fingerprint. The register entry and the missing inquiry form are, together, the shape of the crime.

Working alongside Jaspal Singh Dhillon of the Committee for Information and Initiative on Punjab, Khalra cross-referenced multiple municipal cremation registers against police encounter reports, family disappearance accounts, and district administrative records. He found bodies that had been cremated under police authority without family notification. He found entries for persons described as “unclaimed” who had, in fact, been claimed by families who were never contacted. He found firewood vouchers that recorded the procurement of fuel for cremations that no magisterial inquest had authorized. By 1995, he had documented evidence of over six thousand such cremations in Amritsar district alone.

The Ensaaf–HRDAG analysis, published in 2009, later validated and extended Khalra’s method using quantitative analysis. Their researchers identified 482 firewood purchases by Punjab Police specifically for the cremation of bodies classified as “unidentified,” and 303 cremation records from the Durgiana Mandir cremation ground in Amritsar district alone. They characterized these figures as an almost certain undercount. Their methodology integrated multiple datasets—municipal cremation logbooks, police encounter reports, and family testimony archives—to produce what remains the most rigorous quantitative analysis of the mass-cremation period.

Limitations of Evidence

Any investigation of this kind operates under constraints that must be clearly stated. The most significant is the systematic unavailability of the state’s own internal records from the relevant period. The Section 58 warrantless-arrest registers for Amritsar district for the periods 1984–1987, 1987–1992, and 1992–1996 have not entered the public record. If they were maintained, their production through RTI proceedings would constitute the single most important archival step available to any future investigation. If they were not maintained, then the failure to maintain them is itself a statutory violation. If they were maintained and are being withheld, the withholding is a material fact in any accountability proceeding.

The full CBI report on the Punjab mass cremations remains partially sealed. The breakdown of the 2,097 bodies by identification status has been partially described in secondary sources including the NHRC proceedings and HRW’s 2007 report; the full primary document has not been publicly released. The ACR and APAR records for the Amritsar SSPs across the three DC tenures have not been retrieved. The Section 174 inquest forwarding logs have not entered the public record. The monthly law-and-order progress reports from the Amritsar DC’s office from 1984 to 1996—the district’s official account of itself during the worst twelve years of its modern history—have not been published or disclosed.

These absences are not methodologically neutral. In institutional accountability analysis, the systematic non-availability of records that the law required to be created and maintained is itself evidence of institutional failure. The burden of explanation belongs to the state, not to the investigator. This audit states, plainly, what the public record contains and what it does not, and treats both the presence and the absence of records as evidentiary facts.

A second limitation concerns the use of community memorial archives and family testimony. The events documented in community archives [EN-1] through [EN-4] have not been adjudicated in courts with the procedural rigor that formal evidence standards require. They are presented in this investigation as the public historical record they constitute—consistent, durable, and widely corroborated across multiple independent sources—with explicit acknowledgment that they derive from community and survivor sources rather than judicial findings. Where they are relied upon for specific claims, that reliance is stated and the source identified.

PART THREE THE ROAD TO 1984: CIVIL ADMINISTRATION, POLITICAL BETRAYAL, AND THE UNRAVELING OF PUNJAB, 1975–1984

The Emergency and the Securitization of Punjab

The institutional crisis that culminated in Operation Blue Star did not begin with the first shot fired in 1984. It began, as institutional crises so often do, in a period of apparent stability, when the architecture of governance was quietly being rearranged in ways that would not reveal their consequences for years. The Emergency of 1975–1977, declared by Prime Minister Indira Gandhi on June 25, 1975, and the political lessons it taught about the relationship between democratic process, minority communities, and the central government’s willingness to use emergency powers against constitutional norms, was the first and most formative of these rearrangements.

Punjab’s Sikh community observed the Emergency with particular alarm. The apparatus that Indira Gandhi’s government deployed during those two years—preventive detention under the Maintenance of Internal Security Act, press censorship, the suspension of habeas corpus, the jailing of opposition politicians, the forced sterilization campaigns that fell heavily on rural communities—was used against the full spectrum of democratic opposition. But for Punjab’s Sikhs, who had a historically grounded awareness of their community’s structural vulnerability within the Indian state, the Emergency was something more than a general assault on democratic norms. It was evidence of a specific proposition: that the constitutional guarantees theoretically available to all Indian citizens could be suspended at the center’s discretion when the center judged that suspension to be politically convenient.

That lesson embedded itself in Sikh political consciousness in ways that neither the Congress party nor successive central governments fully understood or took seriously. The Anandpur Sahib Resolution, passed in 1973 and reaffirmed in 1978, must be read against this background. Its demands—for greater federal autonomy for Punjab, the right to river waters that belonged to Punjab under any reasonable reading of riparian law, the return of the city of Chandigarh to Punjab, and a more equitable distribution of central power between Delhi and the states—were not, in their original form, secessionist demands. They were federalist demands, consistent with the kind of regional assertion that other states, including Tamil Nadu under the DMK and later the AIADMK, had made without being described as dangerous. The decision by successive central governments, both Congress and coalition, to treat the Anandpur Sahib Resolution primarily through a security lens rather than a political one was itself a choice, with devastating consequences that neither side fully anticipated at the outset.

The Nirankari Massacre of 1978: The State’s First Failure of Even-Handedness

On April 13, 1978—Vaisakhi, the holiest day in the Sikh religious calendar, the anniversary of the founding of the Khalsa—a procession organized by the Sant Nirankari Mission under its leader Gurbachan Singh passed through the streets of Amritsar while shouting slogans that the Sikh community interpreted as deeply sacrilegious toward the Guru Granth Sahib and the human Sikh Gurus. A peaceful jatha of Sikhs, led by Bhai Fauja Singh of the Akhand Kirtani Jatha and including members of the Damdami Taksal, marched toward the Nirankari congregation to protest what they regarded as blasphemy against their scripture. At the Reego Bridge near Gobindgarh, police halted the Sikh jatha and promised to intervene. They did not.

What followed was a massacre. The Nirankari gathering, which survivor accounts and press documentation describe as having been armed in advance with rifles, guns, spears, acid bottles, and hand-made bombs, attacked the Sikh jatha. Thirteen Sikhs, including Bhai Fauja Singh himself, were killed. Over seventy were wounded. The investigation that followed produced a sequence of institutional failures so complete that it reads, in retrospect, as a blueprint for the selective administration of justice that would characterize the entire subsequent decade.

Gurbachan Singh and more than sixty of his followers were charged. But Gurbachan Singh was not arrested in Amritsar. He was permitted to leave the city and travel to Delhi, where he was arrested only after being permitted a personal audience with Prime Minister Morarji Desai. This was not a minor biographical detail. It was an institutional signal that the Nirankari leader who had presided over a gathering that killed thirteen Sikh worshippers on the holiest day of the Sikh calendar was entitled to political access before he faced the law. The case was then transferred to a court in Karnal in the neighboring state of Haryana. On January 4, 1980—two days before the Lok Sabha poll—all accused were acquitted on grounds of self-defense. Punjab authorities were unable, or unwilling, to ensure that prosecution witnesses remained uncompromised in the Haryana venue. Chief Minister Prakash Singh Badal’s government did not appeal the acquittal. The legal process was permitted to die.

The institutional consequences were severe and irreversible. Every stage of the sequence had represented a decision by some level of the state to limit accountability for the killing of thirteen Sikh worshippers. The Nirankaris were permitted to hold their congregation in Amritsar with state government approval. The police did not prevent the attack. The chief accused was permitted to reach the Prime Minister before arrest. The case was shifted out of Punjab jurisdiction. The acquittal was not appealed. Taken together, these decisions taught a generation of Sikh observers that the state would protect establishments with Congress connections and proximity to Delhi political circles before it would protect the ordinary Sikh devotee. The credibility of the state’s claim to impartiality in Punjab was, after 1978, structurally compromised.

Punjab Waters: The Arithmetic of Betrayal

Water is life in Punjab, and the progressive surrender of Punjab’s riparian rights to non-riparian states represents one of the longest and most consequential administrative betrayals of the Sikh farming community across the period this investigation covers. The dispute’s modern form begins in 1966 with the reorganization of Punjab into the states of Punjab, Haryana, and Himachal Pradesh. The water-allocation question was left deliberately unresolved at the reorganization, a political decision that created the structural conditions for the crisis that followed.

In 1976, the central government issued a notification allocating 3.5 million acre-feet each to Punjab and Haryana from the waters of the Ravi and Beas—rivers that, under the principle of riparian rights recognized in both international and Indian water law, belong to the states through which they flow. Haryana, which has no riparian claim on the Ravi or Beas, was thus allocated water from rivers in which it had no territorial stake. Punjab filed a suit in the Supreme Court in 1979 challenging this allocation. Then, in 1981, Prime Minister Indira Gandhi prevailed upon Chief Minister Darbara Singh to withdraw that case.

The December 31, 1981 agreement, signed in the presence of Indira Gandhi by Punjab, Haryana, and Rajasthan, was the critical document. Under it, of the 17.17 MAF of re-assessed available water from the Ravi and Beas, Punjab received only 4.22 MAF, while Haryana received 3.50 MAF and Rajasthan—which has no geographical relationship to these rivers—received 8.60 MAF. That 75 per cent of Punjab’s river water was allocated to non-riparian states is the central arithmetic of the betrayal. Akali Dal documentation records that the Badal government had received Rs. 1 crore from Haryana for SYL construction as far back as November 1976 and had asked for additional funds thereafter—suggesting that the Akali government, while publicly opposing the division, was simultaneously facilitating the project it claimed to oppose.

On April 8, 1982, Prime Minister Indira Gandhi formally launched the Sutlej-Yamuna Link canal construction at Kapoori village in Patiala district. The SYL was a 214-kilometre canal designed to carry Haryana’s allocated water from the Sutlej to the Yamuna, 122 km of which would pass through Punjab. The Rajiv–Longowal Accord of July 24, 1985, signed four months after the assassination of Indira Gandhi, committed to SYL canal completion by August 15, 1986. Longowal was assassinated by Sikh militants one month after signing the accord, in August 1985. The accord itself was widely criticized within the Sikh community as a capitulation that conceded water rights without securing any of Punjab’s other demands.

The KBS Sidhu connection to this water story surfaces decades later, in November 2016, when Sidhu, serving as Financial Commissioner of Revenue, signed the executive notification de-notifying the land designated for the SYL canal and returning it to original owners. The man who administered Amritsar at the height of the counterinsurgency later issued the administrative resolution of one of that period’s central political preconditions. That he did so without, at any point in the intervening decades, any formal public reckoning with what his district had witnessed between 1992 and 1996 is an illustration, with documentary precision, of the institutional continuity of the Punjab administrative cadre across the crisis and its aftermath.

The Sacrilege of Sacred Space: Arms in the Darbar Sahib and the State’s Complicity

One of the most contested questions about the period immediately preceding Operation Blue Star is the question of who allowed the fortification of the Harmandir Sahib. By early 1984, the complex was functioning as an armed facility. Weapons had been stockpiled. Bunkers had been constructed. Major General Shabeg Singh—a decorated but subsequently court-martialed and dismissed army officer who had trained several insurgent groups in the region, including the Mukti Bahini during the 1971 Bangladesh liberation war—had organized the defensive architecture of the complex with professional military expertise. The arms had to enter. Their entry required logistical cooperation, transportation, and the absence of police interdiction despite clear intelligence awareness.

The investigation by The Caravan magazine into the unanswered questions of 1984 documented that the police were aware arms were being carried into the sacred precinct. An army intelligence officer was posted in Amritsar. A cover story in Surya magazine, published shortly after the operation, quoted senior RAW sources alleging that most of the arms inside the complex had been smuggled under the supervision of a so-called Third Agency, controlled directly by a director in the Prime Minister’s secretariat, to create justification for an assault. DGP Pritam Singh Bhinder reportedly confirmed to the central government that weapons were being transferred into the complex. One week before the operation, Punjab Police intercepted two truckloads of weapons in Batala sub-division, Gurdaspur district. A Third Agency officer reportedly persuaded the DGP to release them and ensure their passage to the Darbar Sahib.

The district administration of Amritsar throughout this period sat at the center of a question it has never publicly answered: what did the DC’s office know about the weapons entering the most symbolically important site in the district? SSP Ajaypal Singh Mann, who served as SSP of Amritsar from October 1983 to March 1984, is associated in accounts of this period with having raised concerns about or attempted to challenge the ingress of weapons into the Darbar Sahib. Mann was removed from his position in March 1984, replaced by Sube Singh, who was himself replaced by Bua Singh in June 1984. The rapid churning of the Amritsar SSP position—six different officers in four years between 1981 and 1985—reflects both the impossible political pressure of the posting and, in some cases, the consequence of taking positions that the state’s political leadership found inconvenient.

DIG Atwal’s Assassination and the Twenty-Four-Hour End of Elected Government

On April 25, 1983, DIG A.S. Atwal of the Punjab Police visited the Darbar Sahib to pray. He was a senior police officer of the Jalandhar range. As he was leaving the complex, he was shot and killed by a lone gunman. Two others were critically injured. The killing took place in broad daylight, metres from his bodyguards and official car. Multiple press and political sources of the period record that at the precise moment Atwal was gunned down, Inspector General Jarnail Singh Chahal was alone with Harchand Singh Longowal inside the complex, while another police officer, Grewal, was also in the vicinity. Both Chahal and Grewal left immediately for Chandigarh after the killing, where they met Governor Anant Sharma and then Chief Minister Darbara Singh, rather than attending to Atwal’s body.

The murder was handed to the CBI, bypassing Punjab Police, on direct instructions from the Prime Minister. Longowal stated publicly: “Whenever the situation becomes ripe for settlement, some violent incident takes place. I think there is a government conspiracy behind the DIG’s murder.” When asked who could be responsible, Longowal implied Chief Minister Darbara Singh, who had been on the verge of replacement by the Congress high command. The CBI investigation found no evidence of Bhindranwale’s involvement—the person the government promptly blamed. Within one day of Atwal’s killing, President’s Rule was imposed in Punjab on April 26, 1983, placing the district administration directly under central government authority and removing the last layer of electoral accountability from the governance of the state.

DC Gurdev Singh and the Abrupt June 3 Transfer

On June 3, 1984—the day before Operation Blue Star commenced—Deputy Commissioner Gurdev Singh was abruptly removed from office. Several operational chronologies of the period record that the Army had grown wary that Gurdev Singh might harbor sympathies toward militant elements. He was replaced the following day, June 4, by Ramesh Inder Singh, a 1974-batch IAS officer of the West Bengal cadre who was brought into Punjab on deputation and installed in the first Deputy Commissioner–level posting of his career. Contemporary administrative accounts suggest that the choice was deliberate: an officer without prior district command experience in Punjab and without entrenched local political or institutional relationships—someone who could be relied upon to manage the civil administration of Amritsar in the immediate aftermath of the Army’s entry into the Golden Temple complex.

The institutional implication of Gurdev Singh’s removal is direct and cannot be escaped by citing the AFSPA’s formal provisions. Ramesh Inder Singh’s later memoir argues, in effect, that because the Armed Forces (Punjab and Chandigarh) Special Powers Act was already in force in the state, he did not need to sign any document requesting army assistance or authorizing military intervention. Under this argument, the army was already legally deployed in aid of civil power by virtue of the pre-existing AFSPA framework, and the DC’s office was therefore not the operative civilian signatory of the military action. This argument is technically narrow and historically misleading. The primary text of the AFSPA makes clear that Section 3 vests the power to declare disturbed areas in the Governor of the State or the Central Government; the District Magistrate is not mentioned. So in the strictest sense, Ramesh Inder Singh is correct that his signature was not required for the army to operate in Punjab. But if no civilian signature was required, if the AFSPA had already enabled the army’s deployment without reference to the District Magistrate, then why was the removal of the incumbent DC considered operationally necessary? Why was it not sufficient to simply proceed with Gurdev Singh in post, if his role was formally peripheral?

The answer implied by the operational record is that civilian cooperation or non-objection to specific operational elements was considered material to the operation’s execution—even if not formally required under the AFSPA’s text. Ramesh Inder Singh’s claim that he did not need to sign anything is therefore not an exculpation. It is a deflection from the more pointed question of what civilian cooperation was actually provided, in what form, and whether any document exists in the operational files that records the nature of the civil-military interface on June 4, 1984.

Harbans Lal Khanna and the Administrative Tolerance of Anti-Sikh Provocation

The political climate of Amritsar in the years leading up to 1984 cannot be understood without examining the role of actors who engaged in documented anti-Sikh sacrilege while operating within the city’s formal political life, and who did so with conspicuous impunity from the district’s civil administration. BJP MLA Harbans Lal Khanna is the most extensively documented of these actors. In May 1981, he led what became known as the “bidi-cigarette march” in Amritsar—a march organized explicitly to oppose the Sikh community’s demand for Amritsar to be granted holy-city status. The marchers openly brandished tobacco products and chanted slogans mocking Sikh devotion.

On May 30, 1981, Tribune reporting documents that Khanna led a procession during which slogans were raised calling for Sikhs to be “sent to Pakistan.” The slogan, translated: “We are not going to let any second or third group exist, we are not going to let a turban remain on any head; the shorts, the iron bangle, the sword—send these to Pakistan.” On February 14, 1984—four months before Operation Blue Star—mobs led by Khanna gathered at as many as fifty-six places in Amritsar to engage in anti-Sikh desecrations. At the Amritsar Railway Station, a model of the Golden Temple was destroyed. A picture of Guru Ram Das—the fourth Sikh Guru and the founder of Amritsar itself, displayed at the station for several years—was defaced beyond recognition: feces and lit cigarettes were rubbed into it. On April 2, 1984, Khanna was killed.

The institutional question is precise and unambiguous. Fifty-six locations in Amritsar. A model of the Harmandir Sahib destroyed. A photograph of Guru Ram Das defaced with feces and burning cigarettes. All of this occurred in the district of Amritsar, under the civil authority of the Deputy Commissioner. No formal administrative action against Khanna for inciting communal violence through these provocations appears in the public record. The civilian administration’s silence in response to documented anti-Sikh sacrilege is part of the same institutional pattern this audit documents across the entire twelve years: the civilian office’s selective activation of its powers, which operated with considerable force against Sikh political assertion and with conspicuous restraint against anti-Sikh provocation.

PART FOUR OPERATION BLUE STAR AND THE CIVILIAN CERTIFICATION OF MILITARY VIOLENCE

The Operation and Its Immediate Consequences

What the Indian Army did inside and in the immediate vicinity of the Harmandir Sahib complex between June 3 and June 10, 1984 goes far beyond what the government’s official account acknowledges. The government’s position, articulated in the official White Paper on the Punjab Agitation published shortly after the operation, is that the military action was a precise counter-terrorism operation directed at a specific group of armed militants who had occupied the complex and converted it into a weapons fortress. That position, while not entirely false, omits the scale and nature of the civilian toll in ways that subsequent judicial proceedings have partially corrected.

In June 2017, Amritsar District and Sessions Judge Gurbir Singh ruled, in a case concerning compensation for pilgrims trapped in the complex, that there was no evidence the army had provided any warnings for civilians to leave before launching its assault. His written ruling: “There is no evidence that army made any announcements asking ordinary civilians to leave Golden Temple complex before commencing their assault… There is no written record of any public announcement by the civil authorities requesting the people to come out of the complex. No log of vehicle used for making such announcements is there.” The court awarded ₹400,000 in compensation to forty Sikhs on this basis. That judicial finding—from the Amritsar district court, in the same district where the DC’s office governed—is one of the most direct official confirmations of civilian harm from the operation, and it came thirty-three years after the event.

Ramesh Inder Singh’s memoir records that his office prepared body counts from within the Golden Temple complex: 717 bodies recovered between June 3 and 9, 1984, of which 501 were civilians and 216 were armed militants based on records prepared by the civil administration and police before autopsy examinations. That figure is morally staggering. The DC’s office was not observing from a safe distance. It was counting the dead. The question this fact raises is institutional rather than personal: what did the office do with what it counted? The army conducted the operation; the civil administration certified the result. In the legal and moral architecture of the Indian state, that certification carries the civilian seal of the district.

The Ensaaf photo essay and report on Operation Blue Star documents multiple categories of army conduct that violated both Indian law and international humanitarian norms. Army troops burned the Sikh Reference Library inside the complex after taking control of the building, destroying irreplaceable rare Sikh manuscripts, historical artefacts, and one of the most significant collections of Sikh historical documentation in existence. Associated Press reporter Brahma Chellaney, who was present in Amritsar during and after the operation, reported that truckloads of bodies were transported to nearby crematoria around the clock in the immediate aftermath. The army’s post-operation conduct included the immediate cremation of the dead—a practice that destroyed evidence, prevented autopsies, and denied families the possibility of identification. The Caravan Magazine’s detailed reconstruction of the operation records the post-mortem report of a young man shot through the chest with his hands tied behind his back.

Operation Woodrose: Sexual Violence, Collective Punishment, and the Silence of the Magistracy

Operation Woodrose, which ran from June to September 1984, is one of the least publicly acknowledged chapters of Punjab’s post-Blue Star history. The operation was designed to eliminate the organizational infrastructure of the Sikh militancy in Punjab’s villages following the removal of its urban center at the Darbar Sahib. Its systematic character was documented across multiple primary sources, including the Indian Army’s own internal publication Baat Cheet (Talking Points), which declared all Amritdhari—baptized—Sikhs to be extremists. The Army’s own literature branded all Amritdhari Sikhs as extremists; Indira Gandhi’s government endorsed this characterization; Punjab and Chandigarh were declared disturbed areas; and the Punjab and Chandigarh Disturbed Areas Act was passed to provide legal cover for indiscriminate arrests.

What the army did in Punjab’s villages during Operation Woodrose was extensive and has been documented across multiple sources. The army went to villages and pulled out Sikhs. According to multiple testimonial and documentation sources, the army publicly humiliated Sikhs, killed most males or detained them, and engaged in a program of collective punishment that included sexual violence. The Punjab Disappeared digital archive’s comprehensive account quotes KPS Gill himself as acknowledging that the Indian Army had acted “blindly” and that its conduct suffered from “all the classical defects of army intervention in civil strife.”

The sexual violence component of the army’s conduct in this period has been systematically documented by Punjab’s International Human Rights Organization in their report The Rape of Punjab, and by the World Sikh Organization of Canada, whose 2014 account of the case of Amandeep Kaur—a twenty-year-old woman tortured, raped, and killed by the Punjab Police because she was the sister of a suspected militant—records that she was able to give a pre-death statement to human rights workers. The World Sikh Organization’s account states that in addition to the brutal rape and murder of hundreds of Sikh women in November 1984, in the decade that followed, the Punjab Police and Indian security forces routinely used rape as an interrogation tool. This use of sexual violence as an interrogation tool is one of the most under-examined dimensions of the Punjab crisis, and one that falls most directly within the district magistracy’s purview: when sexual violence is used in a police station, the magistracy’s custody-visibility architecture is precisely the institutional mechanism through which the victim’s legal status should have been protected.

The DM’s statutory responsibility in relation to Operation Woodrose is not negated by the army’s AFSPA protection from prosecution. The AFSPA protected army personnel from prosecution for actions within its scope; it did not remove the DM’s obligations under the CrPC to receive information about deaths, disappearances, and custody events, to forward inquest material, and to initiate independent inquiry. The DM could not prosecute an army officer without Central Government sanction. But the DM could and should have documented, escalated, and preserved a record of what was being done in the district in the army’s name. The absence of such a record—the absence of anything that might represent the civilian administration’s registered alarm at the army’s conduct during Woodrose and Blue Star—is itself one of the most significant silences in the entire institutional history of this period.

The Padma Shri of 1986: What Was Being Rewarded

Ramesh Inder Singh received the Padma Shri in 1986 for his contributions to public administration—two years into a tenure that began the morning soldiers entered the Harmandir Sahib. The honor arrived during his watch. The district was not yet quiet. The aftermath of Blue Star was still generating its human consequences: Operation Woodrose, the mass detention of young Sikh men across Punjab’s villages, and the beginning of a counterinsurgency psychology that would persist for a decade. That the state moved from military operation to civil decoration within two years is itself a statement about how it was processing the moral weight of 1984.

PART FIVE THE MECHANICS OF DISAPPEARANCE: HOW A DEMOCRATIC STATE PROCESSES ITS CITIZENS INTO ASH

The Administrative Pipeline

The enforced disappearances that define the Punjab counterinsurgency period were not, at the operational level, chaotic. They were, in their essential mechanics, administrative. Understanding how they worked requires understanding the pipeline through which a living citizen became an unidentified entry in a municipal cremation register. That pipeline had five stages, each of which was governed by law and each of which, in practice, was systematically subverted.

The first stage was custody. A person was arrested—sometimes at a checkpoint, sometimes in a village raid, sometimes from their home or workplace—and taken to a police station or an unofficial detention site. The law required that this arrest be documented: under Section 57 of the CrPC, a person arrested without a warrant could be detained for no more than twenty-four hours before being produced before a magistrate; under Section 58, the officer in charge of the police station was required to report the warrantless arrest to the District Magistrate. These requirements were systematically ignored. Families who went to police stations to ask about missing relatives were routinely told that the person had never been detained—even when eyewitnesses had seen the arrest. The detention pipeline was, from its first stage, designed to be invisible.

The second stage was the encounter. The police’s preferred method of disposing of custody cases was the staged encounter: the detained person was killed by police, and the killing was reported as an encounter in which police officers had fired in self-defense against an armed militant. The encounter killing had a specific legal form. It required a police report, a post-mortem, and, in principle, a Section 174 CrPC inquest by which the magistracy was to formally determine the cause of death. In practice, the post-mortem was either not conducted or conducted in a manner designed to confirm the encounter narrative. The inquest was either not initiated or conducted as a formality that endorsed the police’s account. The Section 176 independent magisterial inquiry—which the law gave the District Magistrate the power to initiate whenever a death in custody appeared suspicious—was, across hundreds of cases in Amritsar district between 1984 and 1994, never initiated in the public record.

The third stage was classification. Once the person had been killed in a staged encounter, the body required administrative classification. If the body could be connected to a name—if a family member identified it, if the police chose to acknowledge the identity for operational or propagandistic reasons, or if the person was sufficiently prominent that denial was impossible—the body was classified as that of an identified militant killed in an encounter. If the family was not notified, or if the police chose not to acknowledge the identity, the body was classified as “unidentified” or “unclaimed.” This classification, once applied, triggered the municipal cremation process.

The fourth stage was cremation. Bodies classified as unidentified or unclaimed were transported to municipal cremation grounds by police officers. The cremation grounds in Amritsar district used for this purpose were at Amritsar city, Majitha, and Tarn Taran. At each ground, the cremation was recorded in the municipal register, which noted the date, the description of the body, the name of the police officer responsible, and the quantity of firewood procured for the cremation. It is these register entries that Jaswant Singh Khalra later read, and it is the discrepancy between the number of these entries and the number of Section 176 inquiries that constitutes the central statistical finding of this audit.

The fifth stage was administrative silence. Once the body had been cremated and the cremation recorded, the administrative machinery of the district was supposed to have generated a paper trail that reached the DM’s office: a Section 58 report recording the original warrantless arrest, a Section 174 inquest report forwarding the details of the death to civilian authority, and a Section 176 inquiry initiated by the magistracy into the suspicious circumstances of the death. None of these documents appear in the public record for the vast majority of the 2,097 cases later documented by the CBI. The administrative silence was not incidental to the process. It was its purpose. The bodies were cremated not merely to destroy the evidence of how the persons had died, but to prevent the state’s own administrative machinery from generating the records that would later make the crime visible.

The 482 firewood vouchers are not the evidence of murder. They are the evidence of a system—a system that budgeted its own erasure, accounted for its fuel, and processed its dead through the ordinary channels of municipal administration.

The Firewood Vouchers: Reading the State Against Itself

The Ensaaf–HRDAG analysis identified 482 firewood purchase vouchers issued by Punjab Police specifically for the cremation of bodies classified as unidentified. These vouchers are procurement documents: purchase orders, account entries, administrative authorizations. Each one records a financial transaction in which an arm of the Indian state paid for wood to burn a human body it had decided not to identify. They are not aberrations in the administrative record. They are a budget line. They are what happens when a state institutionalizes disappearance: the disappearance becomes a cost center, the bodies become a procurement category, and the fuel becomes a line item.

The Ensaaf–HRDAG analysis also identified 303 cremation records from the Durgiana Mandir cremation ground in Amritsar district alone. This is one site. One ledger. Three hundred and three entries for which, in the surviving public record, no family was formally notified through legal process, no independent inquest was completed, and no Section 176 inquiry by the District Magistrate appears. This is the ledger that Jaswant Singh Khalra began examining in 1994. The thread he pulled from this single ledger eventually brought 2,097 bodies into the Supreme Court’s field of vision.

Human Rights Watch, in Protecting the Killers, notes evidence that multiple bodies were sometimes burned with firewood normally sufficient for one—a detail that speaks to the scale of the operation and the industrial character of what was being conducted. It also notes that post-mortems were reportedly conducted on only a small fraction of the bodies brought by police as “unclaimed.” The result was a cremation economy in which the normal safeguards against the disappearance of citizens—the inquest, the post-mortem, the family notification, the magisterial inquiry—were systematically bypassed, and the bypassing was recorded in the same municipal ledgers that the state maintained as evidence of its own administrative regularity.

The 2,097: What the CBI Found

The Central Bureau of Investigation, when it later examined the cremation records at three grounds in Amritsar district—Amritsar city, Majitha, and Tarn Taran—confirmed 2,097 illegal cremations. The CBI’s categorization of these bodies, as described in subsequent NHRC proceedings and in Human Rights Watch’s 2007 report, is as follows: 585 were fully identified; 274 were partially identified; and 1,238 were recorded as entirely unidentified. That 1,238 of these persons—the single largest category—remain officially unidentified is not a statistical footnote. It is the central accusation.

In a functioning legal system, the classification of a body as “unidentified” is a temporary administrative condition: it describes the tragic but ordinary situation in which a body has not yet been linked to a known identity, and it triggers a set of obligations—inquiry, notification efforts, family tracing—designed to resolve that condition. In Punjab’s cremation economy between 1984 and 1994, the classification of a body as “unidentified” became a permanent administrative terminus: it was the final station in a process designed to prevent the body from ever being connected to a name, a family, or a legal consequence. The 1,238 unidentified persons did not fail to be identified because the state could not identify them. They remained unidentified because the state did not try. The Section 176 inquiries that would have generated identification efforts were never initiated. The family notification procedures that would have connected bodies to relatives were never followed. The administrative machinery of the district processed these 1,238 human beings into ash without making the effort the law required.

PART SIX THE DISTRICT POWER MAP: INSTITUTIONAL ARCHITECTURE AND THE FLOW OF AUTHORITY IN AMRITSAR

How Authority Flowed Between Civilian and Police Command

To understand how the Civilian Shield worked in practice, it is necessary to understand the formal architecture of authority in Amritsar district during the relevant period. The district’s administrative structure was not simply a hierarchy in which the Deputy Commissioner sat above the Superintendent of Police. It was a dual system in which the civilian and police arms of the state were bound to each other by overlapping and sometimes contradictory legal frameworks, and in which the DC held both supervisory authority and a set of independent statutory duties that existed independently of that supervisory relationship.

The Police Act, 1861—still in force during this period, a product of the colonial administration’s need to place police power under civilian oversight—vested in the District Magistrate “general control and direction” over district policing. Under this provision, the DM was empowered to give direction to the SSP, to review the SSP’s performance, and to supervise the conduct of the district’s police forces. The language was deliberately broad, reflecting the colonial administration’s preference for flexible authority rather than specific legal obligation. In practice, this broad language created a relationship that could function either as genuine oversight or as mutual reinforcement, depending on the institutional culture and political direction in which both officers were operating.

The CrPC, 1973, was different in kind. Its obligations were not supervisory options. They were statutory duties. When Section 58 required that the officer in charge of a police station report warrantless arrests to the District Magistrate, that was not a courtesy procedure that the DM could choose to enforce or ignore. It was a legal mandate whose non-compliance constituted a statutory violation. When Section 174 channeled inquest reports into the magistracy, that was not an administrative preference whose activation depended on the DM’s initiative or comfort. It was a statutory design. When Section 176 empowered the magistracy to initiate independent inquiry into deaths in custody, that power was not contingent on the state government’s comfort level, the SSP’s cooperation, or the political direction in which both were operating. It existed in law as a duty that the magistrate was expected to exercise regardless of political weather.

The Intelligence Architecture and the DC’s Position Within It

Beyond the formal civil-police relationship, the Amritsar DC’s office sat within a broader intelligence and security architecture that included the Intelligence Bureau, the Research and Analysis Wing, and the Punjab Police’s own Special Branch. During the counterinsurgency period, this architecture was supplemented by auxiliary structures—the “Alam Sena” or “Black Cats” associated with Mohammad Izhar Alam, the various informant networks that different SSPs maintained as operational tools—that operated in the penumbra between official policing and deniable violence.

The DC’s relationship to these auxiliary structures was ambiguous but consequential. The DC was not, in any formal sense, responsible for their operation; they operated outside the official command structure of district policing. But the DC was the senior civilian officer in the district, and the existence and operation of these structures within the district was, in principle, part of the information environment to which the DC was exposed. If auxiliary enforcement networks were operating in the district, generating disappearances and deaths that then appeared in the municipal cremation registers, the DC’s failure to initiate the Section 176 inquiries that the law required cannot be separated from the DC’s position within the broader intelligence architecture of the counterinsurgency.

The TADA Architecture and the DC as Originating Authority

The Terrorist and Disruptive Activities (Prevention) Act, 1987—TADA—fundamentally altered the constitutional position of the District Magistrate in relation to detention. Section 9(1) of TADA authorized a District Magistrate to order the detention of a person where the DM was “satisfied” that detention was necessary to prevent the person from acting in a manner prejudicial to the security of the state or maintenance of public order. This detention could be for up to twelve months, renewable. Under this provision, the DC was not a recipient of detention decisions made elsewhere. The DM was the originating decision-making authority for preventive detention at the district level.

This matters enormously to the moral ledger of all three post-1984 Deputy Commissioners of Amritsar. When TADA was used against persons who were subsequently alleged to have been tortured, disappeared, or illegally cremated, the DM’s detention order sits in the administrative record before the abuse. Human Rights Watch documented that TADA was systematically misused: the definition of “terrorist act” was so broad that it covered virtually any act of violence, and the provision making confessions before a Superintendent of Police admissible as evidence was used to extract confessions under torture. By signing TADA detention orders, the DM was not merely processing paperwork in a technical chain. The DM was the civilian officer who gave preventive detention the stamp of official justification. When those detentions were followed by torture, disappearance, or death, the DM’s order was part of the chain of events that preceded the abuse. The DM is not the sequence’s bystander. The DM is its first signature.

The ACR System: Bureaucracy’s Internal Language of Reward

Among the least-examined dimensions of the Civilian Shield is the Annual Confidential Report system—the ACR—through which senior civil servants evaluated each other’s performance and thereby shaped career trajectories within the IAS and IPS. The ACR was the bureaucracy’s internal language: how the state spoke to itself about its own officers when it believed the world was not listening. In that language, what was rewarded, what was excused, what was left untroubled, and what was praised became evidence of institutional mood.

A 1968 Haryana instruction—applicable in principle to Punjab’s administrative culture during this period—authorized the District Magistrate to initiate the confidential report of the Superintendent of Police in categories including reputation for honesty, dealings with the public, and overall efficiency in law and order. If the three Deputy Commissioners of Amritsar between 1984 and 1996 assessed their SSP counterparts—Izhar Alam, Gobind Ram, Samant Kumar Goel, Ajit Singh Sandhu—and if those assessments reflected the law-and-order efficiency logic of the counterinsurgency rather than the statutory oversight obligations the DMs also held, then the ACRs of that period are themselves part of the evidence of how institutional impunity was constructed from within. They reveal not only that punishment failed to arrive from outside the system, but that praise continued to flow from within it.

The Punjab-specific ACR and APAR rules governing the DM’s evaluative role vis-à-vis the SSP for the periods 1984–1987, 1987–1992, and 1992–1996 remain to be formally retrieved through RTI applications or archival research. The underlying ACR extracts for Amritsar across all three civilian tenures remain to be formally obtained. This is a Public Interest Declassification Demand: the Government of Punjab, the Department of Personnel and Administrative Reforms, and any repository currently holding the sealed Annual Confidential Reports for the Amritsar district administration between 1984 and 1996 should declassify and publish the ACR extracts relevant to the district’s SSP counterparts during all three civilian tenures. The families whose relatives appear in the “unidentified” column of the cremation registers are entitled to know whether the civilian authority that governed their district praised the police officers who produced that column.

PART SEVEN THE MORAL PIVOT: ANOKH SINGH AND THE VAIROVAAL TEST

A Name the District Office Should Have Preserved

On August 30, 1987—in only the eighth week of Sarabjit Singh’s tenure as Deputy Commissioner of Amritsar—a thirty-year-old man named Anokh Singh was tortured to death inside Vairovaal Police Station in Amritsar district. He was born on May 28, 1957, in village Waring Suba Singh, near Khadur Sahib, in what was then Amritsar district and is now part of Tarn Taran district, to Bapoo Makhan Singh and Mata Niranjan Kaur. He was the third of four brothers. The family was amritdhari and Gursikh.

His martyrdom is not a footnote to the administrative history of these years. It is its test. Every statutory obligation that this audit documents—the Section 58 report to the District Magistrate, the Section 174 inquest, the Section 176 independent magisterial inquiry—was triggered by what happened to him in that police station. The evidence for what happened is documented in multiple community memorial archives and the SikhiWiki account. It has not been adjudicated in a court. But it constitutes a consistent body of testimony that has circulated publicly for nearly four decades, and it is recorded here as the public historical record it represents.

The Biographical Record

Anokh Singh completed his primary education in village Waring Suba Singh and attended high school in Khadur Sahib. He became an Inspector in the Shiromani Gurdwara Parbandhak Committee. Community accounts across multiple published sources describe a man of unusual moral integrity: he refused to misuse a single rupee of Gurdwara funds, insisted on recounting the golak at Gurdwara Muktsar Sahib when he found discrepancies, collected his family’s dasvandh personally from the farm harvest, and told his brother’s sister-in-law at the birth of a nephew that even two rupees from his own pocket felt excessive—since those two rupees were the Panth’s property and not his to spend on family occasions.

He trained in gatka at Bhai Fauja Singh’s Khalsa Farm and was present with Bhai Fauja Singh on April 13, 1978, the day of the Sikh-Nirankari massacre in Amritsar. The martyrdom of Bhai Fauja Singh and twelve of his companions that day shaped the remainder of Anokh Singh’s life. He is also documented in community memorial archives [EN-1, EN-2] as one of the Singhs who helped ferry pilgrims and others out of the Harmandir Sahib complex during Operation Blue Star in June 1984. Some earlier community documents give his age as twenty-nine or his birth year as 1953. This investigation adopts the primary family record—born May 28, 1957; martyred August 30, 1987; age thirty at death—as confirmed by family testimony, National Sikh Youth Federation records, and the ShaheedKhalsa institutional archive.

The Arrest: Ordinariness as Institutional Implication

What is significant about the circumstances of Anokh Singh’s final arrest is precisely its ordinariness. In August 1987, he was stopped while riding a bicycle by Jalandhar Police Inspector Surjit Singh. This was not a targeted operation mounted because police had intelligence on his location. It was a routine stop. The police did not initially know who they had. When asked his name, Anokh Singh identified himself: he said his name was Anokh Singh, that his father was Guru Gobind Singh Ji, and that he resided in Anandpur Sahib. Only at that point did the Punjab Police Chief arrive on the spot. The state had stumbled upon a significant figure not through intelligence success but through chance. He was then transferred to Vairovaal Police Station in Amritsar district.

That distinction matters to the institutional argument. A man stopped at a routine checkpoint and taken to a local police station falls squarely within the ordinary custody-visibility architecture of the CrPC. This was not a covert special operation conducted outside any paper trail. It was an arrest that should have produced documentation: a Section 58 report to the District Magistrate, a remand production, a custody register entry. If those documents were created, they have not entered the public record. If they were not created, then the most basic anti-disappearance mechanism of Indian law was ignored at the very first step.

What Was Done at Vairovaal

Multiple community memorial archives [EN-1, EN-2, EN-3] and the SikhiWiki account corroborate each other in their essential details about what happened to Anokh Singh inside Vairovaal Police Station over the days that followed. According to these accounts, he was kept suspended upside down inside the station for extended periods. Heated metal rods were hammered into his legs from the heels upward. Electric shock was administered. Heated rods were passed through his chest. His eyes were removed. His tongue was cut out with a metal saw. The SikhiWiki account records that an inspector named Surjit Singh—the same officer who had arrested him—ultimately removed the top of Anokh Singh’s skull, invoking the historical martyrdom of Bhai Taru Singh, whose scalp had been removed by Mughal authority in the eighteenth century. Anokh Singh reportedly responded: “You are persecuting me as a Sikh, but you should also know that if Taru Singh’s skull had been dismembered, then Suba Zakariya Khan would not have survived.”

Through all of it, the accounts record no cry of pain. What the guards heard was Waheguru. After the torture sessions, Anokh Singh was left on the ground, unable to move, his eyes gone, his beard soaked in blood. When the time for Rehraas Sahib arrived, he asked the guard what time it was. He began to recite Rehraas Sahib lying on the ground in a clear voice. He then performed his daily Ardaas:

“Hay Akaal Purakh Sachay Patshah Guru Nanak dev Gareeb Nivaaj Satguru, the day which rose in your bhaanaa has passed in happiness, meditating on your Naam. Sachay Patshah, night has arrived, may it too pass while remembering your Naam and in your bhaanaa… Satguru, may I be a Shaheed… Sachay Patshah, save me from becoming a traitor… Satguru, may the Sikhi and Kesh you have given me last with me till my final breaths… Such Patshah, have mercy on all… Bolay So Nihaal… Sat Sri Akaal!!”

The jaikara echoed through the police station. A sentry, hearing it, ran to the station head, surrendered his rifle, and submitted his resignation on the spot, saying he could not do a job that forced people to kill saints. The senior officer reportedly cursed: “They are all magicians… another of our officers has left the service because of this paath.”

Different accounts name different senior police officers as present or responsible for the conduct at the station. Some accounts name SSP Azhar Alam, the officer associated in other documentation with the “Alam Sena” auxiliary network. Community martyrdom documentation [EN-4] attributes his killing to SSP Brar of Patiala—an attribution noted here for completeness, not relied upon as the exclusive account given conflicting attributions across sources. The SikhiWiki account names Inspector Surjit Singh as the direct instrument of the final act. These varying attributions are recorded not to resolve them—the underlying investigation that would allow resolution has never been publicly conducted—but to illustrate the very absence of official accountability that is this audit’s subject. In the end, Anokh Singh was shot inside the station. A motorcycle was started to mask the sound.

The Sentry’s Resignation: The Conscience the System Could Not Accommodate

The detail of the sentry’s resignation must not be allowed to pass as biographical color. It carries institutional weight that the entire apparatus of the counterinsurgency failed to match. A junior guard’s conscience generated what the district’s supreme civilian officer did not: a rupture. A recognition that what was happening in the police station was incompatible with continued institutional participation. That recognition, expressed through the surrender of the rifle and the resignation from the post, represents what the CrPC’s custody-visibility architecture was designed to produce at a higher level—and did not. The sentry exercised the moral reflex the system was designed to possess but had learned to suppress.

This is how the broken magistracy reveals itself most concretely. Not through the absence of an office, but through the gap between what a junior guard’s conscience produced and what the state’s civilian apex produced in the same moment. That gap is the Shield. That gap is where impunity lives. The law imagined that the civilian administration’s institutional alarm would function at a higher level than a sentry’s personal conscience. The record suggests it did not.

Tarn Taran Crematorium: The Kara and the Archive That Refused to Close

The Sikh memorial archive record [EN-2] establishes, correcting some earlier community accounts that placed the disposal of the body in the River Beas, that Anokh Singh’s body was cremated by the Punjab Police at the Tarn Taran city crematorium—the same city whose cremation ground would later become central to the NHRC’s mass-cremation proceedings. The family in village Waring Suba Singh received word that he had been killed in a police encounter and identified. The villagers and family went to the Tarn Taran crematorium. They collected his ashes. Among what they found was his kara—his steel bracelet, one of the five articles of Sikh faith, worn since the day of Amrit—in the cremation remains.

That steel bracelet survived the fire. It was the physical trace that the state could not fully erase, the residue of a person in the very place the state had chosen to process him into nothing. Giani Puran Singh, the head granthi of Sri Harmandir Sahib at the time, attended the martyrdom ceremony in the village. He remarked that Anokh Singh’s martyrdom recalled the sacrifice of Bhai Taru Singh Ji, the eighteenth-century martyr whose scalp was removed by Mughal authority—the same historical comparison that Anokh Singh himself had reportedly made to his torturer in his final moments. A massive gathering of the Sikh Sangat attended. The village organized an Akhand Paath of Sri Guru Granth Sahib Ji. The martyrdom ceremony became public, attended by figures of standing in the community.

The Sikh memorial archive record [EN-2] also documents that after Anokh Singh’s martyrdom, his father Bapoo Makhan Singh was abducted by police and subsequently killed. This second targeting—of the father after the son—occurred during the era of Chief Minister Beant Singh, placing it within or close to the tenure of KBS Sidhu as DC of Amritsar. Human Rights Watch documented collective family punishment as a systematic feature of Punjab counterinsurgency operations: the state targeted not only the individual but the household, extending the geography of reprisal to ensure that survival itself carried the message of what dissent cost.

What the CrPC Required: A Statutory Accounting

Under Section 57 of the CrPC, detention beyond twenty-four hours without magistrate authority was unlawful. Under Section 58, warrantless arrest was reportable to the District Magistrate. Under Section 167, remand required judicial production. Under Section 174, a suspicious death was to generate inquest material forwarded into civilian authority. Under Section 176, the magistracy held the power to initiate independent inquiry into deaths in custody. A man was arrested at a checkpoint. He was taken to a police station in the district. He was held for multiple days. He died there under torture. His body was transported to the Tarn Taran crematorium and burned under police authority. The family was notified of a “police encounter” after the fact.

Every stage of that sequence was supposed to leave a paper trail that reached the DM’s office. Was an inquest report created and forwarded? Was a Section 176 inquiry initiated into what was plainly a custodial death? Was any document generated in Sarabjit Singh’s office recording that a man named Anokh Singh had been held at Vairovaal Police Station, had died there under circumstances of torture, and had been cremated at Tarn Taran? Was the father’s subsequent abduction and killing documented as a second custodial event requiring civilian attention? The public record does not contain the answers. The silence where those answers should be is itself part of the story.

The Vairovaal case establishes, in a single human life, the full anatomy of the systemic failure this audit documents across twelve years and 2,097 deaths. With that anatomy established, we turn to the three administrations that governed Amritsar across the full arc of the crisis.

PART EIGHT THE ROSTER OF ADMINISTRATIVE RESPONSIBILITY: THE STATUTORY CHAIR AND ITS OCCUPANTS, 1984–1996

The Statutory Chair

The institutional argument of this audit is not primarily about any individual officer. It is about the Statutory Chair—the office of Deputy Commissioner of Amritsar, which was legally bound throughout the entire period 1984 to 1996 by the obligations of the Police Act, 1861 and the Code of Criminal Procedure, 1973. Individuals rotated through that chair. Their names, their personalities, their retrospective self-presentations, and their later career trajectories all varied. What did not vary was the legal weight of the seat itself. The Police Act, 1861 vested in the holder of that chair general control and direction over district policing. The CrPC vested in that chair specific, mandatory, non-discretionary duties in relation to custody, death, and disappearance. Every officer who sat in the Statutory Chair of the Amritsar DC from June 1984 to August 1996 inherited those obligations in full. The chair did not pause its duties when one officer departed and the next arrived.

What follows is the administrative anatomy of the period—the documented occupants of the Statutory Chair, mapped against their SSP counterparts, the statutory obligations active during each tenure, and the events that fell within each district administration’s institutional responsibility. This is not a personal indictment. It is the documented map of the Facade, the Furnace, and the Shield in the hands of those whom the law placed in charge of preventing exactly what occurred.

The Ramesh Inder Singh Era: June 1984 to July 1987

Ramesh Inder Singh assumed office on June 4, 1984—the day Operation Blue Star commenced in earnest. He was a 1974-batch IAS officer from the West Bengal cadre, placed in his first Punjab district posting, selected for the assignment at the onset of a military operation inside the most symbolically important Sikh site in the world. He replaced DC Gurdev Singh, who had been removed the previous day on grounds that the army was “suspicious” of him. The circumstances of that appointment deserve the attention this audit has already given them.

He held the Statutory Chair until July 1987. He subsequently served as Chief Secretary of Punjab and Chief Information Commissioner. In 1986, while still serving as Deputy Commissioner of Amritsar and while the counterinsurgency was still generating its human consequences, he was awarded the Padma Shri for contributions to public administration.

His principal police command counterparts during this tenure included  Bua Singh; and Mohammad Izhar Alam, who is associated in human-rights documentation with the “Alam Sena” or “Black Cats” auxiliary enforcement network—a structure of informants and irregular enforcers operating in the penumbra between official policing and deniable violence.

The key events sutured to this tenure include Operation Blue Star itself—717 bodies counted by the DC’s office, per Ramesh Inder Singh’s own published memoir; the burning of the Sikh Reference Library; the army’s immediate cremation of the dead without autopsy; Operation Woodrose from June to September 1984, with its documented mass detention, collective punishment, and sexual violence; and the extended President’s Rule that characterized the entire Punjab administration from April 1983 onward. The introduction of TADA in 1985 added to the portfolio of coercive powers available to the DM’s office while simultaneously—through its permissive evidential standards—degrading the institutional culture of custody and proof that Section 176 inquiries were designed to maintain.

The statutory obligations active throughout this tenure under the Statutory Chair included the full suite of CrPC provisions: Sections 57, 58, 107, 109, 131, 132, 167, 174, and 176. Every one of the 717 Blue Star bodies counted by the DC’s office triggered a Section 174 inquest obligation. Every Operation Woodrose death in custody triggered a Section 176 independent inquiry obligation. No public record documents the systematic exercise of either power for any of these deaths. Ramesh Inder Singh’s later memoir acknowledges his role in counting the dead but does not address what was done with those counts in terms of the district’s statutory oversight obligations.

The Sarabjit Singh Era: July 1987 to May 1992

Sarabjit Singh held the office of Deputy Commissioner from July 7, 1987 to May 10, 1992. His was the longest tenure of the three, spanning five years of the most intensive counterinsurgency activity in Punjab’s history and the entirety of the period that the NHRC would later characterize as the core mass-cremation years. His entire tenure fell within extended President’s Rule; the MHA’s own annual reporting confirms that Punjab elections were held on February 19, 1992 and a popular government assumed office on February 25 of that year.

He was awarded the Padma Shri in 1989—while still serving as Deputy Commissioner of Amritsar—specifically for, in the words of his book’s publisher, “dedication and courage in the fight against militancy.” His account of Operation Black Thunder II was later published as Operation Black Thunder: An Eyewitness Account of Terrorism in Punjab (SAGE Publications, 2002, ISBN 9780761995968). That book has been reviewed as “probably the only one by a key civilian administrator who was privy to many of the episodes and developments leading to the Operation,” but critics have noted that his measure of counterinsurgency success was defined by the number of militants killed, and that he does not adequately address the prevalence and impact of fake encounters. That self-description—eyewitness, key civilian, dedicated and courageous against militancy—is precisely what this audit contests as the operative frame. His office was not merely an eyewitness. It was an institutional participant whose decisions and omissions form part of the accountability ledger of these years.

His principal police command counterparts included Mohammad Izhar Alam, continuing from the Ramesh Inder Singh era; Gobind Ram, SSP Batala and Amritsar zone, documented in 38 cases of abduction, disappearance, or extrajudicial execution per the Ensaaf accountability database, who was killed on January 10, 1990 in a bomb blast at the headquarters of the 75th Battalion of the Punjab Armed Police in Jalandhar; Sanjeev Gupta; Hardeep Singh Dhillon, a 1985-batch IPS officer who later rose to the rank of Director General of Police Punjab; Samant Kumar Goel, who later served as Director General of RAW; and Ajit Singh Sandhu.

The key events sutured to this tenure are the most consequential in the entire twelve-year period. They include: the custodial death of Anokh Singh at Vairovaal Police Station on August 30, 1987—the eighth week of his tenure; Operations Black Thunder I and II; the peak mass-cremation accumulation period; the 303 cremation records from the Durgiana Mandir register traced by Khalra to records generated in this period; the bounty system documented by HRW in which police officers received financial rewards for killing known militants; and the systematic use of TADA to generate confessions under custodial duress.

The Karanbir Singh Sidhu Era: May 1992 to August 1996

Karanbir Singh Sidhu assumed the office of Deputy Commissioner on May 10, 1992—or May 11, depending on the source, with the discrepancy itself a minor illustration of the opacity of the administrative record—and held it until August 11, 1996. He departed, in due course, toward the University of Manchester. He subsequently served as Financial Commissioner of Revenue, Punjab, in which capacity he signed the November 15, 2016 notification de-notifying land designated for the SYL canal and returning it to original owners.

His principal police command counterparts included Hardeep Singh Dhillon, continuing into the earlier years of this tenure; Samant Kumar Goel, SSP Amritsar during the Khalra investigation period; and Ajit Singh Sandhu, who served as SSP Amritsar and is named by the CBI among the nine officers to be prosecuted for the abduction and murder of Jaswant Singh Khalra. Sandhu was arrested on Supreme Court direction, released on bail, and died on May 24, 1997 under circumstances that community and rights documentation characterize as a staged killing rather than a suicide. This audit records that characterization without adopting it as a settled factual finding.

The key events sutured to this tenure include the Galiara Heritage Corridor project—the displacement of 859 families, the demolition of 500 houses and more than 1,150 shops, and the creation of an open security perimeter around the Harmandir Sahib complex; Jaswant Singh Khalra’s register research from 1994 to 1995; the CIIP petition to the Supreme Court in April 1995; Khalra’s abduction on September 6, 1995; the CBI investigation transferred by the Supreme Court in November 1995; KBS Sidhu’s departure for Manchester on August 11, 1996; and the Supreme Court’s formal remittal of the Punjab Mass Cremation case to the NHRC on December 12, 1996—exactly four months and one day after Sidhu left office.

The Khalra abduction alone—within this district, during this tenure, within the geographic jurisdiction of the Statutory Chair—triggered every statutory mechanism this audit has documented: a Section 176 independent inquiry into the custodial disappearance of a person already before the Supreme Court; the duty to preserve district records relevant to the CBI investigation transferred in November 1995; and the duty, as the district’s supreme civilian officer, to generate the documentary friction the law placed the DC in position to create. None of these obligations appear to have been discharged in any publicly available record.

Continuity Note

The statutory obligations set out above for each tenure were not suspended between tenures, were not waived by counterinsurgency legislation, and were not negated by the concurrent operation of AFSPA, TADA, or NSA powers in the district. Each DC who sat in the Statutory Chair inherited those obligations in full on their first day and carried them until their last. The Padma Shri awards to Ramesh Inder Singh in 1986 and to Sarabjit Singh in 1989, awarded while they served as DC Amritsar during the peak counterinsurgency years, represent the state’s internal evaluation of their performance. The question this audit places before the archival record is whether that evaluation reflected the statutory oversight obligations those officers held, or only the counterinsurgency’s operational logic.

PART NINE THE POLICE ARCHITECTURE OF AMRITSAR, 1984–1996: A COMMAND MAP OF IMPUNITY

Mohammad Izhar Alam and the Auxiliary Architecture of Fear

Mohammad Izhar Alam features prominently in human-rights accounts of Amritsar’s policing during the crisis years. He is associated in survivor testimony and rights documentation with the auxiliary structure known variously as the “Alam Sena” or “Black Cats”—a network of informants and irregular enforcers who operated in the penumbra between official policing and deniable violence. The significance of such structures is not merely operational. They are institutional. They reveal that the policing apparatus of Punjab had developed methods that went beyond any ordinary reading of the Police Act—methods that created a zone of action in which force could be applied without the full paper trail of official conduct. That zone was dangerous precisely because it touched, but officially denied touching, the district’s formal civilian architecture.

The counterinsurgency logic that Alam embodied required the district office to either rupture publicly over the methods being used in its name or to absorb those methods into administrative normalcy. The public record suggests the latter. The district continued to function. Reviews were held. Orders were issued. The silence around the auxiliary methods was not broken. That silence is part of the Civilian Shield’s operation: by refusing to generate the documentary friction that would have made the auxiliary methods visible, the DC’s office allowed them to continue.

Gobind Ram: Documented Misconduct, Institutional Protection, and Retaliatory Death

Gobind Ram served as SSP of Batala during the height of the counterinsurgency and became one of the most extensively documented perpetrators of custodial abuse in the Punjab rights archive. The Ensaaf accountability database documents 38 cases directly implicating him in abduction, disappearance, or extrajudicial execution. Human Rights Watch and Physicians for Human Rights’ Dead Silence (1994) contains detailed testimony about his methods, including the torture of the wives of suspected militants.

Among the most extensively documented incidents: on August 21, 1989, armed men abducted Bibi Gurdev Kaur and Bibi Gurmeet Kaur—wives of two figures associated with the Babbar Khalsa, both underground—from their workplace in Amritsar. They were taken to the Beco Torture Centre in Batala, where Gobind Ram personally beat them with a rod on arrival. The torture continued over subsequent days. Human rights organizations condemned Gobind Ram by name. DGP KPS Gill publicly declared the reports false and described the officers involved as “honest and hardworking.” No action was taken. Amnesty International’s 1991 report on India noted that Gobind Ram was described in Punjabi press reporting as having been involved in the torture of villagers and the wives of suspects. His name appeared repeatedly in petitions and complaints filed with the Punjab and Haryana High Court. No official inquiry sustained any findings against him while he held command.

Gobind Ram was killed on January 10, 1990, in a bomb blast at the headquarters of the 75th Battalion of the Punjab Armed Police in Jalandhar. He held the rank of Commandant by that time. Community martyrdom archive documentation records that the bombing was carried out using explosives planted in a cooler in Gobind Ram’s office. The blast killed three people, including Sub-Inspector Prem Kumar, and injured at least four others. This audit records his death in its proper context: after a career of documented custodial abuse during which no official accountability process was initiated against him, and during which the civilian administration of the district in which much of his command operated failed to generate a visible statutory response to his documented conduct. His death was not justice. It was retaliatory violence from outside the legal system precisely because the legal system—including the District Magistrate’s oversight architecture—had never engaged with his conduct.

Hardeep Singh Dhillon: Career Ascent Through the Counterinsurgency Years

Hardeep Singh Dhillon is a 1985-batch IPS officer who rose through the ranks of the Punjab Police to eventually attain the rank of Director General of Police, one of the highest positions in state policing. His career arc spans the crucial period from the mid-1980s through the mid-1990s—the years of most intensive counterinsurgency activity—and the question his career raises is the same question this audit raises about all who advanced through this period: what was the counterinsurgency curriculum through which he was trained, and what methods were rewarded during the years when the Punjab Police’s internal culture was shaped by encounter logic, the bounty system, and TADA’s evidentiary shortcuts?

These questions are not personal accusations. They are structural questions about what it means to advance through an institution during years when that institution was later officially described as having committed flagrant human-rights violations at scale. The specific ACR and evaluation records that would illuminate what supervisors said about Dhillon’s performance during his early field postings remain to be retrieved. Until they are, the question of how the civilian evaluation architecture interacted with the police career structure—and whose advancement was endorsed, by whom, in what terms—remains an outstanding archival demand.

Samant Kumar Goel: The Intelligence Career That Followed

Samant Kumar Goel occupied the police command in Amritsar during the latter phase of the counterinsurgency, including the period overlapping with KBS Sidhu’s DC tenure and the year of Jaswant Singh Khalra’s abduction. He subsequently rose to serve as Director General of RAW, India’s external intelligence service—a career elevation representing one of the most striking examples of institutional advancement from within the Punjab counterinsurgency apparatus to the apex of India’s national security architecture. The question this audit places before the archival record is not about his later career; it is about what evaluation the civilian administrative authority of Amritsar may or may not have provided to his service record as SSP of a district whose conduct during this period the NHRC would later characterize as involving flagrant human-rights violations.

Ajit Singh Sandhu: 512 Cases, Command Responsibility, and a Death Before Justice

Ajit Singh Sandhu, IPS, is the counterinsurgency officer whose career most comprehensively concentrates the institutional failures this audit examines. The Ensaaf accountability database documents that Sandhu bears command responsibility for a minimum of 512 cases of enforced disappearances and extrajudicial executions across his postings as SSP in Kapurthala, Sangrur, Ropar, and Tarn Taran. Ensaaf explicitly states this figure is an undercount.

The Kuljit Singh Dhatt case is among the most legally substantiated. On July 23, 1989, Sandhu led a police party that abducted Kuljit Singh Dhatt and others from a village in Gurdaspur district. Other detainees were later released. When Kuljit’s brother Harbhajan went to the police station to inquire, Sandhu reportedly told him directly: “We have let the rest go. We have done with Kuljit Singh what we wanted to do. We aren’t going to return the body. Do what you want.” SP (Operations) SPS Basra confirmed this position. A habeas corpus petition was filed with the Supreme Court in September 1989. In October 1993, the report of retired Sessions Judge H.L. Randev—appointed as Commissioner by the Supreme Court—implicated five police officials: Ajit Singh Sandhu, Jaspal Singh, Sardool Singh, SPS Basra, and Sita Ram. Three serving officers, including Sandhu, were suspended.

The CBI recommended Sandhu among the nine officers to be prosecuted for the abduction and murder of Jaswant Singh Khalra. Jaspal Singh—documented in both the Dhatt disappearance and the Khalra case—was one of the six officers convicted of Khalra’s murder by the court in November 2005. Sandhu was promoted from the Punjab Police Service to the Indian Police Service by DGP KPS Gill specifically in recognition of his Tarn Taran operations—a promotion that is one of the most direct pieces of evidence available for the proposition that the counterinsurgency rewarded exactly the conduct the law required to be investigated and punished.

Sandhu was arrested on Supreme Court direction for his alleged role in Khalra’s disappearance, released on bail, and died on May 24, 1997. The Ensaaf documentation and materials from the Khalra case characterize his death as a staged killing rather than a suicide. This audit records this characterization without adopting it as a settled factual finding; the operative point is that the man who bore documented command responsibility for 512 minimum cases, who reportedly threatened witnesses outside a courtroom in January 1997, and who was among the CBI’s recommended prosecution targets died before standing full trial, for reasons that have never been officially or judicially resolved.

PART TEN THE FACADE AND THE FURNACE: THE GALIARA PROJECT AND THE MUNICIPAL LEDGER OF DEATH

The Galiara: Heritage as Military Architecture

The Galiara—the Heritage Corridor cleared around the Harmandir Sahib complex during Sidhu’s tenure as Deputy Commissioner—is the most visible physical legacy of the administrative period this audit examines. It is also the most instructive illustration of the Facade: the use of administrative power for a purpose it publicly denied while pursuing another.

The public rationale for the Galiara was heritage preservation and beautification. The creation of an open corridor around the Harmandir Sahib complex would enhance the pilgrimage experience, provide space for the management of large crowds, and create a heritage precinct appropriate to the site’s status as one of the most visited religious sites in the world. This rationale was not entirely false. The complex does attract enormous numbers of pilgrims. Crowd management is a genuine administrative challenge. Heritage preservation is a legitimate governmental objective.

But one operational reconstruction of Operation Blue Star, available in multiple military and academic analyses of the 1984 operation, notes with blunt directness that the operation had been “hardly worth the game, as the bulk of militants escaped through the many lanes and by-lanes around the complex, remaining unsecured due to poor reconnaissance.” The Galiara was the state’s answer to that operational lesson. The lanes that had protected civilians and provided concealment for escape would be removed. The perimeter would be opened. Sightlines would be cleared. The next military operation at the Harmandir Sahib, if it came, would begin from a tactical geometry designed to eliminate the urban complexity that had made Blue Star so costly.

This is not a speculative interpretation. It is the statement—explicit and documented—of the operational logic that the Galiara served. The DC’s office administered this project. Eight hundred and fifty-nine families were displaced to achieve it. Five hundred houses and more than eleven hundred and fifty shops were demolished, per Tribune reporting. The city of Amritsar was founded in 1577 by Guru Ram Das on land purchased and consecrated as the property of the Sikh Panth, held in perpetuity as the Guru Ka Chakk. The revenue administration of Punjab therefore faced a title structure in which the “owner” of significant portions of the urban fabric adjacent to the complex was, in Sikh theological and legal understanding, the Guru Granth Sahib itself, with the SGPC as temporal custodian. The Land Acquisition Act’s standard framework was not designed for this situation. Concerns about the legitimacy of the acquisition proceedings were raised at both the state level and in representations to central government bodies. Those concerns do not appear to have substantially delayed or modified the project. The Galiara proceeded. Sacred urban intimacy was replaced by state-managed open space. The displaced families carried the human cost of a project that served military geometry while wearing the clothes of heritage improvement.

That visible capacity is the central counter-argument to any alibi of bureaucratic helplessness. An office capable of reshaping the sacred urban fabric of the most symbolically significant Sikh site in the world—over legal objections, over title complexity, over community resistance—cannot plausibly be imagined as institutionally powerless in the face of municipal irregularities, custody anomalies, or suspicious death paperwork. Power existed. The ethical question is what it was used for, and why the same civil authority that could impose its will on stone and street appears so comparatively thin in the surviving public record of civilian rupture around the district’s deeper rights collapse.

The Municipal Ledger: Budgeted Erasure

The most devastating finding of the mass-cremation proceedings is not simply that people were killed. It is that a democratic state found ways to continue appearing governed while the dead were being converted into administrative residue. Bodies were not merely hidden; they were entered, categorized, cremated under public municipal authority, and dissolved into the district’s ordinary civic paperwork. The NHRC annual report states that the Punjab Mass Cremation case, remitted by the Supreme Court on December 12, 1996, involved alleged disappearances due to police abductions culminating in the alleged secret cremation of 2,097 bodies by Punjab Police in Amritsar, Majitha, and Tarn Taran between 1984 and 1994.

AUDIT FINDING: THE MUNICIPAL PAPER-TRAIL — FIREWOOD VOUCHERS

Primary Source: Ensaaf – Human Rights Data Analysis Group, Calculating Disappearances in Punjab, India (2009). Corroborated by HRW, Protecting the Killers (2007); NHRC Annual Report 2007–08.

Four hundred and eighty-two firewood purchase vouchers were issued by Punjab Police for the cremation of bodies classified as “unidentified.” These are procurement documents: purchase orders, account-ledger entries, administrative authorizations. Each records a financial transaction in which an arm of the Indian state paid for wood to burn a human body it had decided not to identify. They are not aberrations in the administrative record. They are a budget line—the accounting signature of what this investigation calls budgeted erasure.

Three hundred and three cremation records are from the Durgiana Mandir cremation ground in Amritsar district alone. One site. One ledger. Three hundred and three entries for which no family was formally notified through legal process, no independent inquest was completed, and no Section 176 inquiry by the District Magistrate appears in the surviving public record.

The Ensaaf–HRDAG methodology note: these 482 vouchers and 303 records represent a documented minimum. For each of the 482 purchases: someone signed off on the fuel. Someone recorded the cost. Nobody recorded the name. Administrative Lapse: zero corresponding Section 176 CrPC Magisterial Inquests on file.

AUDIT FINDING: THE MUNICIPAL PAPER-TRAIL — THE 2,097 CREMATION RECORD

Primary Source: NHRC Annual Report 2007–08; CBI findings as described in NHRC proceedings; HRW, Protecting the Killers (2007). Supreme Court remittal: 12 December 1996.

Two thousand and ninety-seven: the total number of illegal cremations confirmed in the sealed CBI report at three cremation grounds in Amritsar district, covering the period 1984–1994. This figure entered the Supreme Court record when the Punjab Mass Cremation case was formally remitted to the NHRC on 12 December 1996—four months and one day after KBS Sidhu vacated the office of DC Amritsar.

CBI breakdown: five hundred and eighty-five were fully identified; two hundred and seventy-four were partially identified; one thousand two hundred and thirty-eight were entirely unidentified. The single largest category—by a wide margin—is “entirely unidentified.” In a functioning legal system, “unidentified” is a temporary investigative condition pending inquiry. In Punjab’s cremation economy between 1984 and 1994, it became a permanent administrative terminus.

Jaswant Singh Khalra estimated from his own examination of the registers that more than six thousand cremations had occurred in Amritsar district alone. The CBI’s 2,097 is the state’s own conservative floor.

Criminal Court Violation triggered across all 2,097 cases: Every one of these cremations was, by the statutory definition of Section 174 CrPC, a death in circumstances raising a reasonable suspicion that some other person had committed an offence. Every such death required an inquest. Section 176 gave the District Magistrate independent power to order a separate inquiry in all such cases. No Section 176 inquiry into any of these deaths appears in the public record from the Amritsar district magistracy across any of the three tenures. This is not a gap. It is a Willful Omission across twelve years and three administrations. Administrative Lapse: zero Section 176 CrPC Magisterial Inquests on file.

PART ELEVEN THE BROKEN MAGISTRACY: WHAT THE LAW DEMANDED AND WHAT THE RECORD SHOWS

A Systemic Diagnostic Failure

As a physician, I view the absence of magisterial inquests as a failure of the state’s “forensic pulse.” If the police are the muscle of the state, the District Magistrate is the nervous system—tasked with sensing and reporting trauma. In Amritsar between 1984 and 1996, the nervous system was intentionally numbed.

In medicine, a systemic diagnostic failure is distinguished from a single missed diagnosis. A single missed diagnosis is a tragedy of individual judgment. A systemic failure is a pattern—a structured suppression of the feedback loop that would ordinarily carry alarm from the point of injury to the point of intervention. Peripheral nerves report pain upward. The spinal cord relays the signal. The brain responds. Remove that relay capacity at any point in the chain, and the organism continues to function in its other systems while the injury deepens undetected. The patient continues to walk, eat, speak, and appear governed—until the damage becomes irreversible.

The CrPC was designed precisely as the state’s diagnostic infrastructure. Section 58 was the peripheral nerve: every warrantless arrest was to generate a signal upward to the District Magistrate within twenty-four hours. Section 174 was the spinal relay: every suspicious death was to enter the formal inquest channel that reached civilian authority. Section 176 was the brain’s intervention function: the magistracy’s independent power to order inquiry when the body below was failing to report accurately. When all three fail together—systematically, across twelve years, across thousands of deaths and disappearances—what has occurred is not individual negligence. It is an administered anesthesia of the state’s own sensing capacity. In Amritsar between 1984 and 1996, that anesthesia was not an accident. It was a feature.

The Section 176 Audit: 2,097 Deaths, Zero Inquiries

Under Section 176 of the CrPC as in force during the relevant period, every person who dies while in the custody of the police generates a mandatory magisterial inquiry. The provision uses mandatory language: the Magistrate shall hold an inquiry. The obligation is not conditioned on the DM’s satisfaction that the death was suspicious. It is triggered by the fact of custodial death itself. Under this provision, 2,097 custodial deaths in Amritsar district between 1984 and 1994 should have generated 2,097 Section 176 inquiries, or their documented equivalent. This audit finds no publicly available evidence that any number approaching 2,097—or any substantial fraction of that number—was produced.

The absence of a corresponding Section 176 inquiry record is not explained by any emergency legislation in effect during the period. AFSPA Section 7 immunizes army personnel from prosecution; it does not remove the Magistrate’s obligation to inquire. TADA confers enhanced police powers; it does not suspend Section 176. NSA permits extended preventive detention; it does not eliminate the mandatory inquest obligation. The statutory obligation existed. The institutional record shows it was not discharged. This is the central finding of the CrPC audit.

Section 58: The Missing Arrest Register

Section 58 requires police station officers to report all warrantless arrests to the District Magistrate. In a counterinsurgency environment characterized by widespread use of warrantless arrests under NSA and TADA provisions, this register—if maintained—would represent the primary documentary record of the detention pipeline. The Section 58 registers for Amritsar district for the periods 1984–1987, 1987–1992, and 1992–1996 have not entered the public record. If they were maintained and are intact, their production through RTI proceedings would constitute the single most important first step in reconstructing the civilian administration’s knowledge of the custody-visibility failures documented in the NHRC proceedings. If they were not maintained, the failure to maintain them constitutes a separate finding of institutional failure. If they were maintained and are being withheld, the withholding itself constitutes a material fact in any accountability proceeding.

The DM as Certifier of Encounter Deaths: The Inquest Corruption

When Punjab Police killed persons in alleged encounters and presented the bodies for official processing, the chain of certification that followed was a civilian chain. The Section 174 inquest was supposed to be an independent process by which the cause of death was formally determined. In practice, across hundreds and then thousands of deaths in Punjab during this period, the inquest process became a rubber-stamp: deaths were certified as occurring in armed confrontations, bodies were moved into municipal cremation channels, and the encounter narrative was preserved in the official record without the independent forensic and witness scrutiny that would be necessary to test it.

The systematic non-use of Section 176 in the face of hundreds of ‘encounter’ deaths in Amritsar district across these years is not explained by the absence of suspicious circumstances. It is explained by the presence of an institutional culture in which the magistracy did not see it as its role to challenge the encounter narrative. That non-use was itself a certification. By not challenging the police’s self-serving account of deaths, the DM’s office was, in functional effect, endorsing them. That endorsement is an act of active institutional complicity, not mere passive failure. The consequences were material: an encounter death endorsed by the civilian inquest process produced a paper record that was far harder to subsequently challenge. Years later, when families sought to establish that a son or husband had been murdered rather than killed in combat, they faced not only the police’s narrative but the district’s own official record, which the DM’s institutional silence had helped create and stabilize.

District Progress Reports and the Production of False Normalcy

Every district administration in India produces regular reports for the state government: monthly or quarterly assessments of law and order, public-health conditions, revenue matters, development progress, and the general state of civic life. These reports are the district’s official voice. In a district where disappearances, custodial torture, illegal cremations, and encounter killings were occurring on the scale later documented by the NHRC proceedings, what did the Amritsar district’s progress reports say?

The public record does not contain these reports. They have not been published, disclosed, or sought through RTI proceedings in any publicly documented form. But their content is a material question. If the monthly law-and-order reports filed by the Amritsar DC’s office during 1984–1996 certified normal or improving conditions without reference to the patterns of custodial abuse, disappearance, and irregular cremation that were occurring in the district, then those reports were false official documents. They certified a reality that did not exist. Under IPC Section 218, a public servant who, with intent to save a person from punishment, frames an incorrect record or writing, commits an offence. An official who produces a false picture of district reality, at scale, over years, while fellow citizens are being tortured and their bodies cremated in the ‘unidentified’ register, is building the documentary foundation on which impunity rests.

PART TWELVE BEYOND THE SHIELD: THE DC’S OFFICE AS ACTIVE INSTRUMENT OF THE COUNTERINSURGENCY STATE

The preceding sections have examined the Deputy Commissioner’s office primarily through the frame of the civilian shield: the failure to activate statutory oversight powers, the absence of proportionate civilian rupture, the institutional composure that allowed police violence to become administratively survivable. That is an important argument. But it is incomplete. It remains possible, when reading the shield argument alone, to imagine the DC as a passive figure: negligent perhaps, morally inert at best, but not directly complicit. That reading is too comfortable.

The Deputy Commissioner in Punjab between 1984 and 1996 was not merely the officer who failed to stop what the police were doing. In significant, documented ways, the DC’s office was the instrument through which some of the most serious human-rights violations were given legal form. The DM signed detention orders under TADA and the NSA. The DM’s office received and processed encounter certifications. The DM exercised land-acquisition powers over Sikh populations. The DM’s office produced the district reports that certified normalcy to the state government. The DM sat inside the TADA and NSA administrative apparatus that enabled indefinite detention without meaningful judicial oversight. These are not passive failures. They are active participations.

Externment: The Administrative Silencing of Witnesses

Section 47 CrPC and parallel provisions of the Punjab Security of the State Act empowered the District Magistrate to issue externment orders directing persons to leave the district. In counterinsurgency administration, externment was deployed against a range of persons extending beyond accused militants: relatives of suspects, lawyers representing TADA accused, journalists documenting rights violations, activists documenting disappearances, and witnesses to specific acts of police violence. An externment order imposed by the DC’s office in the wake of a disappearance or a rights complaint is not administrative housekeeping. It is an act of evidence management directed against a potential witness. Families of the disappeared who were externed from Amritsar district could not visit the cremation ground, could not attend hearings, could not speak to neighbours who might have seen what happened.

Compensation Gatekeeping and the Coercive Architecture of Official Recognition

A family whose son or father had disappeared in police custody could not simply access compensation, obtain a death certificate, or establish the legal presumption of death without interacting with the district administration. The DC’s office was the primary point of this interaction. In practice, families approaching the DC’s office for death certificates or compensation were implicitly required to accept the state’s framing of what had happened—encounter death, militant killed—as the price of any official recognition. Families who contested that framing risked receiving nothing and potentially drawing further attention to themselves from the police structures that had already targeted their family. The DC’s office was not the last institution to fail these families. It was often among the first.

The Full Institutional Ledger

When the full picture is assembled—TADA detention orders, NSA signatures, inquest certification, externment of witnesses, false progress reports, displacement through land acquisition, and compensation gatekeeping—it becomes impossible to describe the Deputy Commissioner’s office in Punjab between 1984 and 1996 as merely a passive shield. The office was an active instrument of the counterinsurgency state. Its powers were regularly deployed not to protect citizens from police abuse but to give that abuse legal form, official cover, and administrative durability. The distinction between the shield and the instrument is not merely semantic. It changes what accountability demands. A passive shield can be reformed by stronger oversight rules. An active instrument requires a reckoning with the specific acts it performed and a full accounting of the records it created, certified, filed, and withheld.

PART THIRTEEN JASWANT SINGH KHALRA: THE MAN WHO READ THE REGISTERS

The Genius of the Method

There are many ways to challenge a state’s preferred version of itself. You can march. You can petition. You can litigate. You can testify. You can write. Jaswant Singh Khalra chose a method so simple, and so devastating in its simplicity, that it has not yet been fully absorbed into the language of what he accomplished: he read the registers. He understood something that the state had overlooked in its confidence in the bureaucratic management of death. A murder may be deniable. A disappearance may be deniable. An encounter killing may be narrated as a legitimate police action. But the systemic process by which bodies are converted into ash cannot be entirely hidden—because it requires resources.

Khalra found the state’s confession in its own procurement records. The body count was hidden not in whispers but in firewood vouchers.

Working alongside Jaspal Singh Dhillon of the Committee for Information and Initiative on Punjab, Khalra used government crematoria records to establish that the Punjab Police had illegally cremated over six thousand persons across Amritsar district alone. As Human Rights Watch documented in Protecting the Killers, Khalra’s research became the foundation on which the CIIP moved the Supreme Court of India in April 1995 to demand a comprehensive inquiry. A state can survive accusation more easily than it can survive being read against its own records. That is why Khalra became intolerable.

The brilliance of this method is also its cruelty—or rather, it reveals the cruelty of what the state had done. The cremation grounds were not black sites. They were civic institutions. The firewood purchases were not black-budget items. They were municipal expenditures. The registers were not classified. They were the routine paperwork of public administration. The state had gambled that no one would read them carefully enough. Khalra called that gamble. He changed the terms of the struggle from allegation to trace, from trace to pattern, and from pattern to institutional implication.

The Abduction and the District’s Response

Jaswant Singh Khalra was abducted by Punjab Police on September 6, 1995. He was secretly detained and tortured for approximately two months. He was murdered in late October 1995. His body was disposed of in a canal. The Supreme Court’s judgment in Prithipal Singh and Others v. State of Punjab, decided in 2012, records the judicial history of the case. Six police officers were convicted of his murder in November 2005. Despite eyewitness testimony implicating then Director General of Police KPS Gill in Khalra’s illegal detention and murder, no charges were brought against Gill.

He was abducted during the tenure of Karanbir Singh Sidhu as Deputy Commissioner and District Magistrate of Amritsar. He was, at the time of his abduction, not an obscure activist in a remote rural setting. He was a nationally prominent human-rights investigator, already the subject of Supreme Court attention following the CIIP’s April 1995 petition. He had been publicly named, publicly located, and publicly acknowledged as the source of the evidence that had moved the Supreme Court. He was abducted in daylight in Amritsar itself.

The question this audit asks of the archival record is specific, institutional, and unanswered: in the period between Khalra’s abduction on September 6, 1995 and Sidhu’s departure from Amritsar on August 11, 1996, what did the office of the Deputy Commissioner and District Magistrate of Amritsar do? What communications passed between that office and the CBI, whose investigation had been mandated by the Supreme Court in November 1995? What records were preserved? What escalations were made? What Section 176 inquiry was initiated into the custodial disappearance of a man whose abduction had taken place within the district’s own jurisdiction?

The public record contains none of these documents. Their absence is not evidence of specific wrongdoing beyond what the absence itself establishes. It is evidence of a systemic institutional failure whose dimensions this audit has been tracing across twelve years. The district continued to function. Law-and-order reports were filed. The Galiara was proceeding. And the man who had done more to expose the district’s hidden archive than anyone before or since had been abducted and murdered within its borders, without generating a visible civilian rupture commensurate with what had occurred.

PART FOURTEEN WIDOWS, WITNESSES, AND THE CONSTITUTIONAL BURDEN OF REPAIR

The dead did not remain alone in the archive. The burden of preserving interrupted truth fell onto those least equipped to carry it. Widows became unwilling archivists. Families carried dates, rumors, fragments, and faded documents for years because the state had failed to preserve the citizen’s legal trace. Fear did not end with the killing. It moved into testimony, into village life, into the long corridor between grief and process. A widow need not be physically threatened every day to remain under the command of fear. It is enough that she understands who still holds power, who still visits the police station, who can still make village life difficult.

This burden cannot be romanticized. It was a cruelty layered on top of the first cruelty. The republic that should have preserved the trace of its citizens instead required grieving households to carry for years what institutions had abandoned. Kitchens became archives. Family recollection became evidentiary scaffolding. Widows learned legal vocabulary because the state had failed to preserve the person in legal form. A republic whose bereaved must become its archivists has already failed the test the law was designed to prevent.

Human Rights Watch has documented that six years after the Supreme Court directed the NHRC to investigate the mass-cremation cases, the Commission had not heard testimony in a single case. Amnesty International criticized the Commission’s interpretation of its own mandate as excessively narrow. Delay was not neutral procedure. It allowed fear to sediment into the witness, memory to be burdened by survival, and truth to cool into category before it was fully heard. This is why the widows are constitutional figures in this history, not peripheral mourners. They reveal what happens when the formal custodians of legality fail and private life is forced to perform the labor of public memory.

PART FIFTEEN THE MANCHESTER VACUUM: A BREAK IN ADMINISTRATIVE CONTINUITY

The Documentation Vacuum

Sidhu’s departure for the University of Manchester in August 1996 created a Documentation Vacuum at the precise moment the Supreme Court began looking for the paper-trail of the dead. By the time the NHRC took the case in December, the primary civilian witness to the Amritsar magistracy’s conduct was thousands of miles away and outside the ordinary reach of domestic accountability processes.

The Chronological Record

The following chronology is presented as a factual sequence. No intent beyond what the documentary record establishes is attributed to any individual in this sequence.

On August 11, 1996, Karanbir Singh Sidhu, Deputy Commissioner and District Magistrate of Amritsar, ended his tenure of four years and three months and departed on academic leave, subsequently proceeding to the University of Manchester—an institution of considerable academic distinction in a country that placed him beyond the ordinary reach of India’s domestic accountability processes for the duration of his academic study.

In November 1995—nine months before his departure—the Supreme Court of India transferred the investigation into the abduction and murder of Jaswant Singh Khalra from Punjab Police to the Central Bureau of Investigation. The CBI investigation was therefore active, at an early stage, and focused on events within Sidhu’s district at the time of his departure.

On September 6, 1995—eleven months before his departure—Jaswant Singh Khalra, the human-rights investigator who had surfaced the municipal cremation records from within Sidhu’s district, was abducted by Punjab Police from Amritsar. He was subsequently murdered.

On December 12, 1996—four months and one day after Sidhu’s departure on August 11—the Supreme Court of India formally remitted the Punjab Mass Cremation case to the National Human Rights Commission. The NHRC proceeded to conduct proceedings for over a decade. The Commission’s 2007–08 annual report records findings characterized as “flagrantly violating human rights on a mass scale.”

The departure in August 1996 and the Supreme Court order in December 1996 are not abstractly proximate. They are precisely, documentably, four months and one day apart.

The Asymmetry of Mobility

The governed populations of Amritsar, Majitha, and Tarn Taran—the families of the 2,097 documented cremations, the widows who became unofficial archivists of the state’s custodial record—remained in place throughout this period and the decades of proceedings that followed. They navigated intimidation, evidentiary attrition, compensation processes that required accepting the state’s framing of events, and the progressive narrowing of institutional attention. The administrative officer who held the district’s supreme civilian authority throughout the decisive accountability-generation phase was, from August 1996, resident in England and thereafter in resumed domestic administrative service.

This audit characterizes this as administrative structural asymmetry rather than individual design. The governed remain with consequences while the governing class retains institutional mobility. That asymmetry acquires particular weight here because the institutional mobility occurred precisely at the temporal hinge between the active accountability period—the CBI investigation, the Khalra murder proceedings—and the formal national institutional engagement that began with the December 1996 Supreme Court order. Courts and commissions have long recognized that the timing of administrative departures in relation to active or imminent proceedings is material to any assessment of cooperation and candor.

The Outstanding Institutional Questions

In the months between Khalra’s abduction in September 1995 and Sidhu’s departure in August 1996, the office of the Deputy Commissioner held the following concurrent positions: active oversight authority under the Police Act over the SSP whose command is associated with Khalra’s abduction; Section 176 obligations in relation to Khalra’s custodial disappearance and death; Section 58 visibility obligations in relation to the police operations that produced that disappearance; and general documentation and escalation responsibilities as the district’s supreme civilian authority. The questions this audit directs at the archival record are institutional, not biographical: What communications passed between the DC’s office and the CBI, the state Home Department, and the central government between September 1995 and August 1996? Were these communications documented? Are they archived? Are they available?

These are not accusations. They are archival demands of exactly the kind a serious institutional inquiry would put to the record. None of this constitutes proof of design. Administrative transfers are routine. Academic pursuits are legitimate. The responsible analyst must say so plainly and mean it. And yet the same responsible analyst cannot pretend that timing is a neutral matter in a case where the accumulation of accountability pressure was visibly gathering throughout 1995 and 1996.

The 2016 SYL Notification: Administrative Continuity

In November 2016, as part of the Punjab government’s legislative response to the SYL dispute, an executive notification denotifying the land designated for the SYL canal and returning it to original owners was issued by the Financial Commissioner of Revenue: Karanbir Singh Sidhu. The SYL dispute had been one of the three principal political grievances driving the Sikh community’s alienation in the period leading to 1984. The same IAS officer who administered Amritsar during the counterinsurgency’s decisive accountability period later issued the administrative notification resolving a central dimension of its political precondition. This is an institutional data point, not a culpability finding. It illustrates, with documentary precision, the institutional continuity of the Punjab administrative cadre across the crisis and its aftermath—and the absence of any formal public reckoning with the crisis period as a condition of continued service.

PART SIXTEEN CULTURAL MEMORY AND NARRATIVE CONTROL: CINEMA, CENSORSHIP, AND THE SECOND DISAPPEARANCE

The Second Disappearance

After official truth comes a subtler struggle. The state no longer asks whether it can deny the wound altogether. It asks whether it can curate what the wound means. This is the second disappearance: not the disappearance of the body, but the disappearance of full narrative and institutional specificity from public memory. It occurs through censorship, de-indexing, selective commemoration, respectable vagueness, and the elevation of officials whose districts remain morally unresolved.

The second disappearance operates differently from the first. The first disappearance—the arrest, the torture, the cremation—required police, firewood, and municipal authority. The second disappearance requires only institutional inertia, bureaucratic procedure, and the willingness of cultural gatekeepers to prevent the first disappearance from becoming fully legible in the public record. The censor board performs the second disappearance in exactly the same register that the District Magistrate performed the first: by declining to generate the documentation that would make the thing visible.

The Films That Challenged the Official Account

Punjab ’95 is not the first film about the Punjab counterinsurgency to have encountered the CBFC’s institutional hostility. The pattern stretches back across years of engagement between cinema and the Punjab’s contested history. Saada Haq, released in 2013, addressed the Sikh political movement and encountered significant resistance to distribution. Toofan Singh, a 2017 Punjabi biographical film directed by Baghal Singh, depicted the torture methods of Punjab Police during the militancy period. The CBFC refused to certify it in 2016 and banned it in India. It was released internationally in 2017 and eventually allowed limited release in Punjab following legal challenges, but the ban itself, and its duration, represent the cultural-censorship reflex that runs from the cremation ground to the censor board.

What distinguishes Punjab ’95 from these earlier films is the scale of the state’s intervention and the specificity of what the state found threatening. Toofan Singh depicted torture methods. Punjab ’95 depicts a method of accountability: the reading of registers, the tracing of a pattern, the construction of an evidentiary case from administrative paperwork. If what the state found intolerable in Toofan Singh was the depiction of what police did to individuals, what it finds intolerable in Punjab ’95 is the depiction of what a citizen can do with the state’s own paperwork. The former is a grievance. The latter is a threat to the architecture of impunity.

Why This Film Specifically

The demands made by the CBFC of Punjab ’95—remove the protagonist’s name, remove references to the Punjab Police as an institution, remove the name of the state of Punjab, change the title from “Ghallughara” to something less historically specific—reveal with precision what the state finds threatening. Each demanded cut is directed at the specific chain of accountability that Khalra’s investigation established: name, institution, state, and the word that describes what happened. The film is threatening not because it depicts suffering but because it names the institutional sources of suffering with the specificity that Khalra’s method made possible.

The state’s position, revealed in the censorship demands, is that suffering may be depicted if it is sufficiently anonymous, sufficiently generic, and sufficiently unmoored from the administrative record of how it was produced. A film about a nameless man who was abducted by nameless forces in a nameless state would be, presumably, acceptable. A film about Jaswant Singh Khalra, who was abducted by the Punjab Police in the state of Punjab, during the tenure of specific officials, based on a method of reading specific administrative records: this is intolerable. The intolerance is, itself, the evidence of what the film is about.

Public Honours, Administrative Afterlife, and the Management of Memory

Public honors, memoir, spiritual or intellectual afterlife, and the selective memory of the administrative class all perform the same function in the management of the second disappearance. The Padma Shri awarded to Ramesh Inder Singh in 1986, while his district was still processing the aftermath of Blue Star, tells the administrative class that one may pass through a morally damaged district and still emerge dignified. The Padma Shri awarded to Sarabjit Singh in 1989, while his district was generating the mass-cremation archive, tells the same class the same thing in the same language. The district remains wounded; the official becomes refined. The municipal ledger still accuses; the later public voice speaks in calmer, higher tones.

This is not an argument that any individual’s later intellectual or spiritual work is without value. People change. Careers evolve. Retrospective self-presentation is a human universal. The issue is asymmetry—the structural gap between the district’s enduring historical wound and the official’s continuing public dignity—and the institutional mechanism through which that gap is maintained. The ACR system rewarded the counterinsurgency’s operational logic during the crisis. The honor system rewarded administrative service after it. The memoir genre permits retrospective self-exculpation. The sum of these mechanisms is the management of memory: the production of a public record in which the officials who administered the Furnace appear as the authors of their own competence, without confronting the full institutional weight of what they governed.

PART SEVENTEEN FLAGRANT VIOLATION: OFFICIAL TRUTH AND ITS CEILING

The true nature of the Civilian Shield lies in the sealed ACRs. Did the DCs of Amritsar grade their SSPs “Outstanding” while the Durgiana Mandir registers were filling with unidentified bodies? This is the central archival question of the 1990s.

The Supreme Court’s routing of the Punjab Mass Cremation matter into the NHRC, and the CBI’s formal investigatory role, marked a profound threshold. Official truth had arrived. The phrase “flagrant violation of human rights on a mass scale” entered the institutional vocabulary. That was not a small concession. The state had admitted enough to wound its own claimed innocence.

But official truth arrived bounded. It recognized enough to establish enormity without restoring the full ecology of how the district worked: how the municipality processed bodies, how the magistracy failed to produce proportionate friction, how police command cultures were incentivized, how families were intimidated, and how administrators were later absorbed into respectability. The CBI report remained sealed even as its importance was described. The NHRC process, though eventually producing compensation categories, was criticized for delay and procedural narrowing. The state admitted enough to prevent total disbelief. It did not admit enough to surrender narrative control.

This is why official truth must be treated as a floor, not a ceiling. The republic’s partial acknowledgment should have opened the archive wider, not closed it. It should have demanded the full civilian file alongside the police record. It should have asked: where are the Section 58 registers? Where are the Section 174 forwarding logs? Where are the district-level inspection notes, the anomaly flags, the escalations to the Home Department, the adverse remarks that a functioning magistracy should have produced? The absence of these materials from the public record is itself a question that the republic has not yet adequately answered.

PART EIGHTEEN THE LONG AFTERMATH: NHRC PROCEEDINGS, UNRESOLVED ACCOUNTABILITY, AND THE RESURGENCE OF MEMORY, 1996–2026

The NHRC Proceedings and Their Limits

The formal beginning of the institutional accountability process came with the Supreme Court’s December 12, 1996 order remitting the Punjab Mass Cremation case to the National Human Rights Commission. The NHRC’s acceptance of the case was itself a significant threshold: it represented the first formal national institutional acknowledgment that the mass cremations in Amritsar district were a matter requiring independent investigation rather than a matter settled by the police’s own encounter narratives.

The NHRC’s proceedings over the subsequent decade produced important findings. The Commission’s annual report for 2007–08 records the Punjab Mass Cremation case as involving alleged disappearances due to police abductions culminating in the alleged secret cremation of 2,097 bodies in Amritsar, Majitha, and Tarn Taran between 1984 and 1994. The Commission characterized the violations as “flagrant violations of human rights on a mass scale.” These findings gave the families of the disappeared a partial institutional recognition that had been denied to them for decades.

But the NHRC process had significant limitations, documented by both Human Rights Watch and Amnesty International. Human Rights Watch reported that six years after the Supreme Court directed the NHRC to investigate, the Commission had not heard testimony in a single case. Amnesty International criticized the Commission’s interpretation of its own mandate as excessively narrow. The compensation frameworks eventually developed by the NHRC required families to engage with an administrative process that was administered by the same state apparatus whose conduct was under investigation, and the terms of compensation required accepting official characterizations of events that families often disputed.

The Khalra Murder Case: Six Convictions and One Acquittal

The conviction of six police officers for the murder of Jaswant Singh Khalra, delivered in November 2005, was the most significant moment of judicial accountability in the entire Punjab counterinsurgency period. The Supreme Court’s judgment in Prithipal Singh & Others v. State of Punjab, decided in 2012, records the judicial history of the case and upholds the convictions. The convicted officers were Prithipal Singh, Jasbir Singh, Satnam Singh, Hardyal Singh, Jaspal Singh, and Surjit Singh.

The judgment also records that eyewitness testimony implicating then Director General of Police KPS Gill in Khalra’s illegal detention and murder was presented in the proceedings but did not result in charges being brought against Gill. This outcome—six convictions at the operational level, no accountability at the command level—is consistent with the pattern documented throughout this audit: the counterinsurgency’s operational methods were carried out by junior and mid-level officers whose actions were enabled, incentivized, and administratively protected by institutional structures that the accountability process could not or would not reach.

Punjab Water Politics and the SYL in the Post-Counterinsurgency Period

The Sutlej-Yamuna Link canal dispute did not end with the counterinsurgency. It continued to define Punjab’s political landscape through the 1990s, 2000s, and 2010s. Engineers and officials connected to the SYL canal construction project died under contested circumstances in the 1990s, contributing to a political environment in which the canal—already a symbol of Punjab’s marginalization within the Indian federation—became associated not only with water rights but with violence and state accountability.

The Punjab Termination of Agreements Act, 2004, through which the Punjab state government sought to unilaterally abrogate the 1981 water agreement, was struck down by the Supreme Court in 2004 as unconstitutional. The political battle over the canal continued into the 2010s, eventually producing the November 2016 executive notification—signed by KBS Sidhu as Financial Commissioner of Revenue—denotifying the SYL land. That notification resolved the land question without resolving the water question, and without addressing the historical relationship between the water dispute and the political crisis of the 1980s.

Diaspora Activism and the Resurgence of Historical Memory

The Sikh diaspora in the United Kingdom, Canada, the United States, and Australia has maintained an active engagement with the events of 1984 and the Punjab counterinsurgency throughout the post-crisis period. This engagement has taken multiple forms: parliamentary and legislative resolutions recognizing the 1984 events as genocide or mass atrocity; public memorials and commemorations; pressure on governments to declassify relevant intelligence and diplomatic records; and the creation of community memorial archives that have preserved testimonial evidence that the Indian state’s institutional machinery did not.

The UK government’s 2017 declassification of some records relating to Operation Blue Star, which revealed that British officials had been consulted about military plans in advance of the operation, was a significant example of archival disclosure produced by diaspora advocacy. The records revealed that British military advice had been sought and provided, raising questions about the complicity of allied governments in the planning of the operation and the management of its international presentation. The full scope of British governmental involvement remains a subject of ongoing parliamentary inquiry.

The resurgence of interest in the Punjab counterinsurgency in the 2020s, driven in part by the censorship battle around Punjab ’95 and in part by the broader global conversation about colonial and post-colonial state violence, has created new audiences for the historical record that this audit documents. Diljit Dosanjh’s global visibility as a cultural figure—and his association with a film about Khalra—brought the Punjab ’95 censorship dispute to international attention in ways that earlier protests had not achieved. The state’s handling of the censorship battle, including the intervention of Ministry officials and the escalation of cut demands, has itself become part of the evidential record of the institutional reflex this audit documents.

Unresolved Accountability Questions

The core accountability questions that this audit has traced across twelve years of administrative history remain unresolved. The Section 58 registers for Amritsar district have not entered the public record. The Section 174 inquest forwarding logs have not been produced. The Section 176 inquiry records—if any exist—have not been disclosed. The ACR extracts for the Amritsar SSPs across the three DC tenures have not been obtained. The full CBI report on the Punjab mass cremations remains partially sealed. The compensation frameworks developed by the NHRC have addressed some families but have not resolved the broader question of how 1,238 entirely unidentified persons came to be cremated in Amritsar district between 1984 and 1994 without generating a single documented independent magisterial inquiry.

This audit closes with the same demand that Khalra opened with: read the registers. Not the cremation registers alone—those have been read, and their testimony has entered the national record, however incompletely. Read the administrative registers that the District Magistrate’s office was required to maintain. The Section 58 register. The Section 174 forwarding log. The Section 176 inquiry index. The monthly law-and-order progress reports. The ACR extracts. The TADA detention order ledger. These are the registers that would reveal, with the same documentary precision that Khalra applied to the cremation grounds, whether the Civilian Shield was maintained by commission or omission, by active participation or passive silence. Until they are produced, the full account of what happened in Amritsar district between 1984 and 1996 remains unfinished.

PART NINETEEN COUNTING THE UNCOUNTED: CONCLUSION

In the end this investigation has been about counting. Not counting in the state’s administrative sense, where the dead become identified, partially identified, unidentified, duplicate, compensable, or processed. Counting in the deeper sense means restoring personhood where the state substituted residue. It means refusing the category as the end of the story. It means reconnecting the dead to place, office, institution, witness, family, and the district that tried to absorb them.

To count rightly in Punjab is to count more than bodies. It is to count three Deputy Commissioners whose tenures spanned the full arc of the crisis. It is to count the police command culture—Izhar Alam, Gobind Ram, Hardeep Dhillon, Samant Goel, Ajit Singh Sandhu—and to ask what the civilian world above them knew, endorsed, evaluated, and preserved. It is to count the magistracy’s failures: the Section 58 registers that may never have been maintained, the Section 174 inquest material that may never have generated alarm, the Section 176 inquiries that appear nowhere in the public record. It is to count the municipal ledger that budgeted the erasure. It is to count the widows who became archivists. It is to count the official truth that arrived late and partial. It is to count the second disappearance: the censored film, the de-indexed name, the narrative managed into comfortable vagueness.

Punjab’s tragedy was not only that force was used. It was that force was made administratively survivable. That survivability required civilian participation: not necessarily at the level of ordering specific acts, but at the level of maintaining the forms of legality while allowing legality’s protective purpose to drain away. That is what the Civilian Shield is. It does not create the sword. It sheaths it. It allows the district to keep functioning, and the official to keep advancing, while the dead are being converted into paperwork.

The film Punjab ’95 is censored because it makes this process visible. The 127 cuts are not about aesthetics or narrative fairness or the protection of reputations. They are about the protection of impunity—the same institutional interest that declined to initiate Section 176 inquiries, that failed to maintain Section 58 registers, that awarded Padma Shris to the civilian administrators of a district that was processing its citizens into ash. The film is threatening because Khalra’s method—reading the state against its own records—is threatening. This investigation is conducted in that same spirit. Read the registers. Follow the paper trail. Count the uncounted. The count is not yet complete.

The final question this audit places before the historical record—and, by implication, before the former civilian heads of Amritsar who are still living, still public, and in some cases still publishing on ethics and governance—is not rhetorical. It is archival. During your tenure, what did the office do when the district’s own legal and municipal machinery was generating the record that later reached the Supreme Court, the NHRC, the CBI, and the Khalra murder judgment? The public record gives us dates, structures, and silence. It does not yet give us the full civilian file. This audit will not stop until the count is complete.

APPENDIX: FULL CHRONOLOGICAL TIMELINE, 1975–2026

The Emergency Period and Its Aftermath, 1975–1979

June 25, 1975: Emergency declared by Prime Minister Indira Gandhi; suspension of habeas corpus; preventive detention of political opponents; press censorship. Fundamental rights of Indian citizens formally suspended. For Punjab’s Sikh community, this period establishes the conditionality of constitutional guarantees.

April 13, 1978: Sikh-Nirankari clash in Amritsar on Vaisakhi. Thirteen Sikhs killed, including Bhai Fauja Singh of the Akhand Kirtani Jatha. Nirankari chief Gurbachan Singh permitted to travel to Delhi before arrest; given audience with Prime Minister Desai. Case transferred to Karnal, Haryana.

January 4, 1980: All Nirankari accused acquitted on grounds of self-defense. Chief Minister Prakash Singh Badal’s government does not appeal. Legal process ends without conviction.

The Political Crisis Deepens, 1981–1983

1976: Central government notification allocating Punjab river water to non-riparian states. Punjab files Supreme Court challenge, 1979.

December 31, 1981: Indira Gandhi prevails upon Chief Minister Darbara Singh to withdraw Punjab’s Supreme Court suit. New water-sharing agreement signed; Punjab allocated only 4.22 of 17.17 MAF.

May 1981: BJP MLA Harbans Lal Khanna leads ‘bidi-cigarette march’ in Amritsar opposing holy-city status. May 30, 1981: Khanna leads procession with anti-Sikh slogans calling for Sikhs to be sent to Pakistan.

April 8, 1982: Prime Minister Indira Gandhi formally launches Sutlej-Yamuna Link canal construction at Kapoori village, Patiala district. Akali Dal launches Kapoori Morcha in protest.

October 6, 1983: Armed Forces (Punjab and Chandigarh) Special Powers Act enacted. Enforced October 15, 1983. Eight months before Blue Star, army given shoot-to-kill, arrest-without-warrant, and search-without-warrant powers in Punjab.

February 14, 1984: Mobs led by Khanna desecrate Sikh religious imagery at fifty-six locations in Amritsar, including destroying a model of the Golden Temple at the railway station and defacing a portrait of Guru Ram Das with feces and lit cigarettes.

April 25, 1983: DIG A.S. Atwal shot and killed at the Darbar Sahib gate. April 26, 1983: President’s Rule imposed in Punjab.

Operation Blue Star and Woodrose, 1984

June 3, 1984: DC Gurdev Singh removed from office; replaced by Ramesh Inder Singh (1974-batch Bengal cadre IAS). Stated reason: army “suspicious” of Gurdev Singh.

June 3–10, 1984: Operation Blue Star. Army enters Harmandir Sahib complex. 717 bodies recovered per DC’s office count (501 civilians, 216 militants). Sikh Reference Library burned by army troops. Bodies cremated immediately without systematic autopsy. No civilian warning to pilgrims documented in court records (Amritsar District Court, 2017).

June–September 1984: Operation Woodrose. Mass detention of Amritdhari Sikhs across Punjab. Army’s internal publication Baat Cheet declares all Amritdhari Sikhs extremists. Documented sexual violence in police stations. KPS Gill later acknowledges army acted “blindly.”

October–November 1984: Anti-Sikh pogroms in Delhi and other Indian cities following assassination of Indira Gandhi on October 31. Thousands of Sikhs killed.

The Counterinsurgency Period, 1985–1992

1985 / 1987: Terrorist and Disruptive Activities (Prevention) Act enacted and renewed. DM becomes originating authority for preventive detention orders. Confessions before SP rank made admissible, enabling extraction under torture.

July 24, 1985: Rajiv–Longowal Accord signed. August 1985: Longowal assassinated by Sikh militants one month after signing accord.

July 7, 1987: Sarabjit Singh assumes office as Deputy Commissioner of Amritsar.

August 30, 1987: Anokh Singh arrested at routine bicycle checkpoint in August 1987, transferred to Vairovaal Police Station, Amritsar district. Tortured to death; body cremated at Tarn Taran crematorium. No Section 176 inquiry appears in public record.

1988: Operation Black Thunder I and II. Security forces retake Golden Temple complex from militants who had reoccupied it.

August 21, 1989: Bibi Gurdev Kaur and Bibi Gurmeet Kaur abducted from their workplace in Amritsar, taken to Beco Torture Centre, Batala. Gobind Ram personally beats them with a rod. No action taken; DGP Gill calls reports false.

January 10, 1990: Gobind Ram killed in bomb blast at 75th Battalion headquarters, Jalandhar.

February 19, 1992: Punjab elections held; popular government assumes office February 25. Sarabjit Singh’s entire tenure had fallen under President’s Rule.

May 10, 1992: Sarabjit Singh ends tenure as DC Amritsar.

The Sidhu Tenure and the Khalra Investigation, 1992–1996

May 1992: KBS Sidhu assumes office as Deputy Commissioner of Amritsar.

1992–1995: Galiara Heritage Corridor project administered; 859 families displaced, 500 houses and 1,150+ shops demolished. Project widely understood as security-geometry operation to clear sightlines around Harmandir Sahib.

1994–1995: Jaswant Singh Khalra, working with Jaspal Singh Dhillon of the CIIP, reads cremation registers at municipal grounds in Amritsar district. Documents 6,000+ illegal cremations.

April 1995: CIIP moves Supreme Court of India demanding comprehensive inquiry into Punjab mass cremations.

September 6, 1995: Punjab Police abducts Jaswant Singh Khalra in Amritsar. He is secretly detained and tortured for approximately two months. Murdered in late October 1995. Body dumped in a canal.

November 1995: Supreme Court transfers Khalra murder investigation from Punjab Police to CBI.

August 11, 1996: KBS Sidhu ends tenure as DC Amritsar; departs for University of Manchester.

December 12, 1996: Supreme Court of India formally remits Punjab Mass Cremation case to NHRC. Four months and one day after Sidhu’s departure.

The Long Aftermath, 1997–2026

May 24, 1997: Ajit Singh Sandhu dies under contested circumstances before standing full trial on Khalra murder charges. Ensaaf documentation and rights materials characterize his death as staged.

1996–2007: NHRC conducts proceedings on Punjab Mass Cremation case. HRW documents that Commission had not heard testimony in a single case six years after the Supreme Court referral.

November 2005: Six police officers convicted of Jaswant Singh Khalra’s murder. DGP KPS Gill not charged despite eyewitness testimony implicating him.

2007: NHRC Annual Report 2007–08 records Punjab Mass Cremation case findings as “flagrant violations of human rights on a mass scale.” CBI breakdown: 585 identified, 274 partially, 1,238 unidentified.

2007: Human Rights Watch publishes Protecting the Killers: A Policy of Impunity in Punjab, India, documenting the career trajectories of perpetrating officers and the systematic impunity of the counterinsurgency period.

2012: Supreme Court decides Prithipal Singh & Others v. State of Punjab, upholding the six murder convictions in the Khalra case.

November 15, 2016: KBS Sidhu, as Financial Commissioner of Revenue, Punjab, signs executive notification denotifying SYL canal land.

2017: Amritsar District and Sessions Court rules there is no evidence army warned civilians before launching Blue Star assault; awards compensation to forty Sikhs.

December 2022: Punjab ’95, biographical drama about Khalra directed by Honey Trehan, submitted to CBFC.

2022–2026: CBFC demands escalate from 21 cuts to 45 to 65 to 85 to 120 to 130+. Demands include removing protagonist’s name, references to Punjab Police as institution, the name of the state of Punjab, and the word “Ghallughara.”

February 7, 2025: Announced international theatrical release date. February 10, 2025: Release cancelled after Ministry of Information and Broadcasting intervention. Honey Trehan publicly reports: “Ronnie got a call from a higher authority that it will not be taken in good taste.”

March 2026: This audit published. Film still not certified for theatrical release in India.

ENDNOTES — DIGITAL ARCHIVE REFERENCES

The following digital archive sources are referenced in the text by endnote number [EN-N]. Because the content of web-based community archives may change over time, all URLs and access dates are recorded here. Where a digital archive no longer hosts the cited content, the author holds copies on file.

[EN-1] NeverForget84.net, “Shaheedi of Bhai Anokh Singh,” Sikh martyrdom digital archive. URL: https://neverforget84.net/shaheedi-of-bhai-anokh-singh-babbar [Accessed March 2026]. Used for: community testimony accounts of Anokh Singh’s martyrdom and the Vairovaal Police Station events. Corroborated against SikhiWiki and family testimony on file with the author.

[EN-2] June84.com, “Shaheed Bhai Anokh Singh,” Sikh memorial archive. URL: https://june84.com/shaheed-bhai-anokh-singh-babbar/ [Accessed March 2026]. Used for: the corrected account of the Tarn Taran cremation, the family’s recovery of ashes and kara, and Giani Puran Singh’s attendance at the martyrdom ceremony. Corroborated by family member video testimony on file with the author.

[EN-3] GurmatBibek.com, Sikh devotional and historical archive. URL: https://www.gurmatbibek.com [Accessed March 2026]. Used for: supplementary biographical details of Anokh Singh.

[EN-4] Sikh Martyrdom Documentation Archive (formerly hosted at 1984tribute.com). URL: https://www.1984tribute.com [Accessed March 2026]. Used for: the attribution of killing to SSP Brar of Patiala (noted as conflicting attribution, not relied upon as exclusive account); Gobind Ram death account, subject to corroboration.

[EN-5] Wikipedia, ‘1978 Sikh–Nirankari Clashes,’ Wikimedia Foundation. URL: https://en.wikipedia.org/wiki/1978_Sikh%E2%80%93Nirankari_clashes [Accessed March 2026]. Used as secondary aggregative source; institutional failure corroborated by primary press documentation and SGPC White Paper.

[EN-6] Wikipedia, ‘Toofan Singh (film),’ Wikimedia Foundation. URL: https://en.wikipedia.org/wiki/Toofan_Singh_(film) [Accessed March 2026]. Used for: CBFC ban history, international release 2017, subsequent Punjab release.

BIBLIOGRAPHY AND SOURCE NOTES

I. Judicial and Statutory Sources (Tier 1)

Supreme Court of India, Order dated 12 December 1996, remitting the Punjab Mass Cremation case to the National Human Rights Commission.

National Human Rights Commission, Annual Report 2007–08, Punjab Mass Cremation proceedings. Available at: https://nhrc.nic.in/assets/uploads/annual_reports/NHRC-AR-ENG07-08.pdf [Accessed March 2026].

Supreme Court of India, Prithipal Singh & Others v. State of Punjab (Khalra murder case), 2012. Available at: https://indiankanoon.org/doc/342168/ [Accessed March 2026].

Amritsar District and Sessions Court, 2017, judgment awarding compensation to pilgrims for failure to warn before Blue Star assault. Amritsar District and Sessions Judge Gurbir Singh.

Armed Forces (Punjab and Chandigarh) Special Powers Act, 1983, Act No. 34 of 1983. Primary text from India Code database. Available at: https://www.indiacode.nic.in/bitstream/123456789/1843/3/a1983-34.pdf [Accessed March 2026].

Code of Criminal Procedure, 1973 (Act 2 of 1974). Sections 47, 57, 58, 107, 109, 131, 132, 167, 174, 176, 218.

Terrorist and Disruptive Activities (Prevention) Act, 1985 and 1987. Section 9(1).

National Security Act, 1980. Section 3(2).

Indian Penal Code, Section 218.

Ministry of Home Affairs Annual Report 1991–92. Available at: https://www.mha.gov.in/sites/default/files/REPORT_1991_92_12022021.pdf [Accessed March 2026].

Government of Punjab, Official Amritsar Deputy Commissioners Roster. Available at: https://amritsar.nic.in/list-of-deputy-commissioner/ [Accessed March 2026].

Rajiv–Longowal Accord (Punjab Accord), 24 July 1985. Full text at: http://www.sikhtimes.com/doc_072485a.html [Accessed March 2026].

II. Human Rights Documentation (Tier 3)

Human Rights Watch, Protecting the Killers: A Policy of Impunity in Punjab, India (2007). Available at: https://www.hrw.org/report/2007/10/17/protecting-killers/policy-impunity-punjab-india [Accessed March 2026].

Ensaaf and Human Rights Data Analysis Group, Calculating Disappearances in Punjab, India (2009). Available at: https://hrdag.org/wp-content/uploads/2018/03/ensaaf-report_50pp_2009.pdf [Accessed March 2026]. Ensaaf accountability database available at: https://data.ensaaf.org/ [Accessed March 2026].

Human Rights Watch and Physicians for Human Rights, Dead Silence: The Legacy of Human Rights Abuses in Punjab (1994).

Amnesty International, India: Human Rights Violations in Punjab (1991) and subsequent annual reports on India through 1996.

World Sikh Organization of Canada, account of Amandeep Kaur case (2014). Available at: https://www.worldsikh.org/india_s_forgotten_rapes_the_story_of_amandeep_kaur [Accessed March 2026].

Ensaaf photo essay and report, Operation Blue Star: The Launch of a Decade of Systematic Abuse and Impunity. Available at: https://ensaaf.org/photo-essays/operation-blue-star/ [Accessed March 2026].

Punjab’s International Human Rights Organization, The Rape of Punjab. Used for documentation of sexual violence during Operation Woodrose.

III. Official Governmental and Institutional Sources

Government of India, White Paper on the Punjab Agitation (1984). Used for the official account of Operation Blue Star and as an example of official narrative management.

SGPC White Paper on the events of June 1984. Used for documentation of anti-Sikh sacrilege and institutional failures.

Reserve Bank of India and central government documentation on SYL canal construction financing, as documented in Aam Aadmi Party research on the SYL history.

IV. Published Memoir and Institutional Autobiography (Tier 4)

Ramesh Inder Singh, Turmoil in Punjab: Before and After Blue Star — An Insider’s Account (HarperCollins India, 2022). Used for: the 717-body count from the Golden Temple complex; the DC’s role in Operation Blue Star’s civilian aftermath; the AFSPA self-exculpation argument; and as an example of the genre of administrative retrospect as selective memory.

Sarabjit Singh, Operation Black Thunder: An Eyewitness Account of Terrorism in Punjab (SAGE Publications, 2002, ISBN 9780761995968). Used for characterization of counterinsurgency success metrics and for the genre of DC memoir as institutional self-presentation.

V. Media Sources (Tier 4)

The Caravan, various articles on Operation Blue Star, the Punjab counterinsurgency, and the Punjab ’95 censorship proceedings. Including reporting by Jatinder Kaur Tur on the February 2025 intervention.

New Lines Magazine, reporting on Punjab ’95 censorship and the “higher authority” intervention.

Deadline, reporting on the cancelled February 2025 international release of Punjab ’95.

Indian Express (2025–2026): reporting on Punjab ’95 censorship proceedings, the 127-cuts demand, and Director Honey Trehan’s public statements.

Tribune (Chandigarh): reporting on the Galiara project (859 families displaced, 500 houses and 1,150+ shops demolished); reporting on Punjab Police promotions including Hardeep Singh Dhillon’s DGP appointment; SYL canal history.

The Print: reporting on the 1978 Sikh-Nirankari clash and the SYL canal dispute history.

Babushahi.com: archive material on Punjab Police promotion records and district-level career documentation.

Surya magazine, post-Blue Star cover story quoting senior RAW sources on the Third Agency and arms smuggling into the Darbar Sahib.

VI. Community Testimony and Family Records

Family member video testimony cited by the author regarding the recovery of Anokh Singh’s ashes and kara at Tarn Taran crematorium. On file with the author. Corroborates the Tarn Taran cremation account against the River Beas account in earlier community documents.

Anokh Singh Martyrdom account circulated in community PDF format, multiple versions. Used for biographical details and the dasvandh and two-rupees episodes. Birth date discrepancy (1957 vs 1953) noted and resolved in favor of primary family record.

SikhiWiki entry on Anokh Singh. Used for the routine bicycle checkpoint arrest, the identification statement, the arrival of the Punjab Police Chief, and the Bhai Taru Singh comparison in custody.

VII. Films (Cultural Memory Chapter)

Punjab ’95 (2022–present, unreleased in India as of March 2026). Directed by Honey Trehan, starring Diljit Dosanjh.

Toofan Singh (2017). Directed by Baghal Singh, starring Ranjit Bawa. CBFC ban history documented at [EN-6].

Saada Haq (2013). Punjabi film addressing the Sikh political movement.

VIII. Author’s Published Research Archive

Punjab ’95 and the Silence of KBS Sidhu (primary article). Available at: https://www.kpsgill.com/KPSGILL/punjab-95-and-the-silence-of-kbs-sidhu [Accessed March 2026].

Crimes Against the Sikh Nation, 1900–2025 (companion manuscript). Available at: https://www.kpsgill.com/KPSGILL/Crimes-Against-Sikh-Nation [Accessed March 2026].

A Sikh Perspective on the Hindu American Foundation’s Khalistan Brief. Available at: https://www.kpsgill.com/KPSGILL/Crimes-Against-Sikh-Nation/a-sikh-perspective-on-the-hindu-american-foundations-khalistan-brief [Accessed March 2026].

IX. Outstanding Archival Demands

The following categories of records have not been obtained. Their absence from the public record is itself a material fact in the accountability ledger. Any institution, researcher, or RTI applicant seeking to complete the historical audit this investigation demands should prioritize these:

Section 58 warrantless-arrest registers for Amritsar district, 1984–1996. Section 174 inquest forwarding logs, 1984–1994. Section 176 inquiry initiation orders. TADA detention orders signed at the DM level, 1987–1995. NSA detention authorizations. Monthly and quarterly law-and-order progress reports from the Amritsar DC’s office, 1984–1996. ACR/APAR records for all SSPs during the three DC tenures. Galiara project planning files including security-rationale documentation. Handover notes between consecutive DCs. The full CBI report on Punjab mass cremations. TADA review board proceedings for Amritsar district, 1987–1995. Police station custody registers (roznamcha and malkhana registers) for Vairovaal Police Station, August–September 1987. Compensation and death-certificate files from the Amritsar DC’s office for families of persons killed in custody or fake encounters, 1984–1996. All communications between the DC’s office and the CBI between September 1995 and August 1996 relating to the Khalra investigation.

X. Sources Consulted but Not Directly Relied On

General internet summaries and secondary aggregations of Punjab counterinsurgency history that were less rigorously sourced than the NHRC, HRW, and HRDAG materials. Public-facing biographical and devotional materials about KBS Sidhu and his later intellectual work. Popular media accounts of the Punjab militancy period where superseded by more rigorous HRW, Amnesty, and NHRC primary documentation. Accounts of specific police encounters that could not be independently corroborated through multiple sources. The 1984tribute.com attribution of Anokh Singh’s killing to SSP Brar of Patiala—noted for completeness but not relied upon as the exclusive account given conflicting attributions across sources.

[Footnote] A continuously updated archive of K.B.S. Sidhu’s published writings on governance, constitutionalism, public administration, and statecraft is available at The KBS Chronicle Substack Archive (https://kbssidhu.substack.com/archive?sort=new). That body of post-retirement commentary is read, across this website, as a contemporaneous interpretive record against which the administrative history of Punjab—particularly Amritsar district between May 1992 and August 1996—is analytically tested.

Within the evidentiary framework developed at KPSGill.com, the office of the District Magistrate is examined not merely as an administrative post but as a civilian shield—a statutory locus of “general control and direction” over the police under the Punjab Police Act, and of mandatory judicial inquiry under Code of Criminal Procedure §176. The record attributed to that office during the relevant period is therefore assessed in light of both documented actions—including high-visibility interventions such as hijacking negotiations and the Galliara (Darbar Sahib precinct) project—and documented inactions, including the absence of magisterial inquiry and civil-administrative intervention in relation to custodial deaths, disappearances, and the pattern of illegal cremations later confirmed in judicial and quasi-judicial proceedings.

The analytical method applied throughout KPSGill.com situates such writings not as abstract reflections on governance, but as part of a dual record: one declarative (post-retirement articulation of constitutional norms), and the other administrative (the statutory, documentary, and evidentiary footprint of the same office when those norms were operationally engaged). The relevance of the archive, therefore, lies in this juxtaposition—between stated principles and the historical exercise, or non-exercise, of authority in a district subsequently defined by litigation, inquiry, and unresolved questions of accountability.