The Second Killing of Jaswant Singh Khalra

On 3 July a feature film about the Sikh human rights investigator Jaswant Singh Khalra was released, uncut, on an Indian streaming platform. Within forty-eight hours it had been removed for viewers in India. It was withdrawn for international audiences the following week.

The film had by then been in existence for more than three years. It began as Ghallughara, became Punjab 95, and arrived under a third name, Satluj. India's Central Board of Film Certification cleared it in 2023 on condition of twenty-one cuts and a change of title. Its producers appealed. The demand then rose to 127 cuts. When the film-makers declined and gave up the theatrical release for which certification is compulsory, they published the film on a platform where no certificate is required. A ministry cited security concerns. The film disappeared.

The list of demanded cuts is, on inspection, a peculiar document. It required the removal of the word Punjab, and proposed that the state be indicated instead by the words for here and there. It objected to the phrase Punjab Police, notwithstanding that the officers depicted are visibly Sikh and could belong to no other force. It prohibited Delhi, desh, system and state. It prohibited law, judiciary and extrajudicial. It forbade the naming of Indira Gandhi, including in a line of dialogue that does no more than establish the year in which the events take place.

What the list does not contain is any requirement to remove the political programme with which Khalra was associated during his lifetime, namely the claim that Sikh sovereignty offered the only durable answer to the violence of the Indian state. It does not contain such a requirement because there was nothing to remove. The film had already declined to represent that claim. Its director has described the completed work as a demonstration of the achievements of the Indian courts, and has said that it shows the constitution in a favourable light and strengthens faith in it.

Two propositions follow from this, and this essay argues both.

The first is that the anticipatory accommodation did not work, and that the manner of its failure is evidence. A film that had removed the sovereignty claim, that had located its resolution in the constitution, and that rested its factual case on judgments of the Indian courts was nonetheless refused certification, and was then suppressed by other means once certification had been circumvented. Whatever the state was protecting itself against, it was not a separatist argument, because the separatist argument was not there. Read carefully, the censor's list identifies its own object: not the sovereignty claim, but the nameability of the state as an agent, in a particular place, subject to a particular law. The vocabulary prohibited is the vocabulary without which responsibility cannot be attributed to anyone.

The second proposition is that the accommodation was itself an erasure, and that it is the more consequential of the two. Censorship is a hostile act, and hostile acts announce themselves. The 127 cuts have been reported, contested, litigated and discussed, and the fact of their demand is now part of the public record of this case. The film's own removal of Khalra's politics announces nothing. It presents itself as a portrait. It will be received as the authoritative popular account of a man about whom no other popular account exists, and its silences will be read not as silences but as the shape of the life. What the state suppressed remains visible as a suppression. What the film declined to say leaves no such trace.

It is worth noticing at the outset how many separate points there are in this story at which something was removed from the record, because they are usually treated as unrelated matters. The Punjab Police created a document when they wrote the word unidentified beside a body. The municipality kept a second one when it filed the receipt for the wood. Khalra found both and put them together. The censor then worked on the conditions under which what he found could be spoken about in public. And the question of what the whole thing will finally be understood to have been about is still open, and is being settled now, in the argument over this film. A cremation register and a feature film do not belong to different orders of enquiry. They are stages in one process.

This essay is therefore concerned with a question about representation that is also a question about history. It does not ask whether Satluj is a good film, and it will concede that in several respects it is. It asks what is transmitted, and what is lost, when a man who understood his own work through a religious and political tradition that he stated explicitly is rendered for a mass audience as a human rights defender in the generic sense. And it asks why that rendering, once performed, did not secure the safety it was calculated to secure.

One prefatory remark. It would be possible to write this as a complaint, and a version of the argument circulates already in Sikh diasporic commentary in exactly that register. The complaint has force, but it has a structural weakness: it treats the film-makers as adversaries, which they are not, and it thereby misidentifies the mechanism it means to expose. The more troubling possibility, and the one pursued here, is that the erasure required no adversary at all.

Murdered again

The charge that has been laid against this film in Sikh commentary since its release is not a complaint about emphasis. It is that the film kills Khalra a second time.

The accusation has three parts, and they should be set out in the terms in which they are actually made rather than in a softened paraphrase.

The first is the erasure of the sangharsh, the struggle itself. Khalra, on this account, was not a bystander who stumbled into a human rights investigation. He was a participant in an armed movement for Sikh sovereignty. He is said to have founded the Khalistan Liberation Organization during his period in Britain and to have edited a magazine advocating Khalistan. He described the struggle as a righteous war and did not disown the methods of those who fought it, and when he spoke of the disappeared he named them not as anonymous victims but as the mothers and uncles and brothers of particular fighters. What the film puts on screen instead, the charge runs, is a watered-down and acceptable human rights activist, shaped for Western and mainstream Indian audiences who will applaud a man documenting abuses and would not applaud a man arguing for secession.

The second is the reduction of a genocide to the misconduct of individuals. What Khalra uncovered, on this account, was not a rogue element within a police force but a systematic and state-sponsored programme of extermination, with a payment schedule and a disposal procedure. What the film delivers is a grievance against corrupt officers.

The third is the omission on which the other two rest. Khalra held that Sikh sovereignty was not one possible response to state violence among several but the only durable answer to it. Remove that conviction and you have removed the thing for which he was actually killed. To honour the man while deleting the belief that cost him his life is, in the phrase that recurs throughout this commentary, to murder him again.

There is an irony in that phrase which nobody in the argument seems to have noticed. It is also the film's director's own phrase. Speaking about the certification process, Honey Trehan said that what the Punjab Police did in 1995 the censor board was now repeating, and that Khalra was being abducted a second time. The film's defenders and its critics have reached independently for the same figure of a second killing. They disagree only about who is doing it.

This essay takes the charge seriously and finds most of it sound. But one of its three parts is stated imprecisely, and the imprecision matters, because the accurate version is considerably more damaging to the film than the inaccurate one.

The film does not reduce the killings to the misconduct of individuals. Anyone who watches it will see that it does not. Its account of the Punjab Police is systemic and unsparing: promotion earned by a body count, an officer released from prison and installed in a district precisely because brutality was wanted there, a constable who cannot resign because he knows too much to be allowed to live, senior men discussing mass killing as a difficulty of public relations. This is not a story about bad apples. It is a serious argument about what an institution does when it holds a monopoly of violence and answers to nobody.

Which is exactly the problem, and the reason the sharper accusation is available.

Satluj does not fail to think structurally. It relocates the structure. It offers a rigorous account of a police force and no account whatever of the relationship between the Indian state and the Sikh community that force was killing. The genocide is not reduced to corrupt individuals. It is reduced to a policing crisis, which is a subtler substitution and a more effective one, because it satisfies the audience's appetite for systemic explanation while leaving the political question untouched. A viewer leaves the film understanding that Punjab's police were out of control. Nothing in the film requires that viewer to wonder why the people they were killing were Sikhs.

That distinction runs through everything that follows. The first two sections of this essay establish what Khalra actually argued, because the charge cannot be assessed until that is on the record. The next examine what the film shows and how the shape of its story constrained what it could show. The fourth reads the censor's list, which turns out to identify its own target more precisely than the public argument has. The fifth sets out the cost. The last states the strongest case against everything argued here, including the case made by Khalra's own widow, who has publicly approved the film.

One further caution before any of that. Two of the claims in the first limb of the charge, the founding of an organisation in Slough and the editorship of a magazine, circulate widely in community sources and have not to my knowledge been independently corroborated. They are load-bearing for the version of the argument made in that commentary. They are not load-bearing here, and the case set out below is built so that it stands without them, on the speeches and writings that can be verified.

A note on sources

Three commitments govern what follows.
The version analysed here is the one released on 3 July, which is the uncut version, published without a certificate because streaming release did not require one. The point is not incidental. Because this is the text its makers intended, no absence identified below can be attributed to the censor. Whatever Satluj does not contain, it does not contain by authorial decision. That is the condition which makes the distinction between the two erasures sustainable rather than speculative, and it should be borne in mind at every point where the analysis might otherwise appear to hold the film-makers responsible for constraints imposed upon them.

Claims about what Khalra believed rest, in order of authority, on his own speeches and writings; on the record of the Indian courts, the Central Bureau of Investigation and the National Human Rights Commission; on the scholarship on the Punjab conflict; and on the documentation reports produced by human rights organisations. Community media, including the podcast commentary through which much of the present controversy has been conducted, is treated as indicative of a position rather than as evidence of fact.

And a word about the armed movement. Khalra's writings and speeches contain assessments of the armed struggle in Punjab that are not those of a pacifist, and the argument here depends on reporting them accurately. Two things follow. Those positions are reported as positions and are not endorsed; the historical and moral evaluation of the armed movement, and of the atrocities committed within it as well as against it, lies outside the scope of an essay about representation. And the argument does not require that Khalra was right. The claim advanced here is that a representation which removes a subject's stated convictions misdescribes him, and that claim holds with equal force whether those convictions were sound or mistaken. A reader who regards Khalra's politics as wholly wrong should still find the argument about Satluj intact.

Some limits should be stated plainly. A substantial part of Khalra's written output exists in Punjabi and in periodicals of restricted circulation, and its availability in English is uneven. The events remain contested in ways no essay can settle: the figure of 6,017 unidentified cremations in Amritsar district rests on Khalra's own documentation and subsequent investigation, while the wider figures of twenty-five thousand and above are estimates and are treated here as estimates. And the reception of the film is still unfolding. The informal screenings and the litigation seeking its restoration are live at the time of writing.

Inheritance and refusal

Satluj supplies its protagonist with a reason to begin. The investigation is triggered by the search for a missing relative, and from that private loss the public enquiry proceeds. The choice is defensible as screenwriting and it is not a falsification, since Khalra did in fact come to the cremation grounds by way of people he knew. It is nonetheless worth registering what the choice accomplishes. It establishes that the man began as a private person to whom something happened.

The record establishes something else: that by 1984 Khalra had been organising, and had been imprisoned for organising, for more than a decade, and that he had been raised in a household in which refusal of the state was settled practice rather than a response to injury. What the film presents as an origin was, on the evidence, a continuation.

The distinction matters more than it may appear to. If Khalra's politics were the consequence of the cremations, then a film that omits the politics has omitted a conclusion, which is a lesser fault. If the politics preceded and produced the investigation, then a film that omits them has omitted the cause and retained only the effect.

The inheritance is documented in outline and contested in detail. Khalra's grandfather, Harnam Singh, was among the passengers of the vessel chartered in 1914 to carry British Indian subjects to Canada in direct challenge to the exclusionary provisions then in force, a voyage known in the ship's registry as the Komagata Maru and in Sikh memory as the Guru Nanak Jahaz. He is said to have been among the five men elected by the passengers to represent them, and family accounts hold that he took an active part in resisting the attempt to board. He was detained on the ship's return and afterwards evaded further arrest by leaving India. The account most often repeated places him in Shanghai. These details reach us principally through oral and community-historical channels, and the argument here does not depend on their precise accuracy. What matters is not whether Harnam Singh fired a pistol but that his descendants were raised on the understanding that he had.

Two acts attributed to the family in the following generation are more significant than the voyage itself. The first is the conversion of family land to the establishment of a high school, the first in the locality, on the reasoning that the estate was better disposed as an institution than retained as property. The second is the refusal of the pension offered by the Government of India to descendants of the passengers. The refusal is attributed to Jaswant Singh, who instructed his father to decline it on the ground that the sacrifice in question was not of a kind that could be compensated, and that money accepted from the state would constitute an acknowledgement the family did not wish to make.

The refusal repays attention, because it is not a gesture of protest in the ordinary sense. Nothing was gained by it and nobody was inconvenienced. What it does is preserve a claim in an uncompensated condition. A pension accepted converts a historical grievance into a settled account. A pension refused keeps the account open. It is also an early instance of a disposition that runs through the whole of Khalra's public life: the refusal to accept from the state a form of recognition that would presuppose the state's standing to confer it.

The religious frame within which such refusals become intelligible is not incidental, and it is worth saying what martyrdom means within the Sikh tradition before the word is used analytically later. It is not a category imported from elsewhere to describe deaths that happen to be sacrificial. It is a developed tradition with its own long history, in which a death is made meaningful by the community that receives and claims it, and in which the story of a martyrdom does identifiable political work at particular moments. That tradition is not timeless. It was built, and rebuilt, and it has been put to different uses in different centuries. But it was available to Khalra as a resource, and he used it. A man raised on the Guru Nanak Jahaz and on the refusal of its compensation was raised within an account of what a life is for.

The public record proper begins with the student period. Khalra joined the Punjab Students Union while at college, and his involvement in campus agitation is said to have cost him the opportunity to sit his final examinations. In 1972, following the killing of student protesters by police at Moga, he organised a solidarity protest and was arrested and imprisoned at Amritsar for the first time. He was then twenty. The sequence is unremarkable in itself and remarkable only in its date. The pattern of organisation, confrontation and detention was established twelve years before June 1984.

His subsequent employment as a village council secretary is the most interesting phase of the early period, and the one least visible in any popular account. The post placed him inside the administrative machinery at the level where it meets the village, and he appears to have used it as a vantage point. He organised against traders hoarding fuel and fertiliser. When he resigned in 1985 he did so by letter, characterising the bureaucracy as a mechanism for the extraction of resources from the people it purported to serve, and declining to continue expending his effort within it. The resignation matters for a reason that becomes central later: the man who would reconstruct a genocide from municipal firewood receipts had spent years learning how such records are kept, and by whom, and for what purpose.

One further episode belongs here. During the Emergency declared in 1975, Khalra went underground. In November of that year, during the tercentenary commemorations of the martyrdom of Guru Tegh Bahadur, he organised a Nagar Kirtan in which the Panj Pyare halted before the police station and discharged firearms into the air. Whatever else it was, this was not a demand. It was a statement about jurisdiction, made in a religious idiom, before a building that represented the state's claim to hold it. That the village did not subsequently inform on him is the other half of the statement.

The programme

Khalra was not a systematic writer. What survives is occasional: articles composed for a magazine of restricted circulation, speeches delivered to diasporic audiences, a resignation letter, correspondence, and the testimony of those who worked with him. From this material a consistent position can be reconstructed, and the consistency is itself significant, but it is a position inferred from occasions rather than one set out in a treatise.

The unmasking. Khalra's own account of June 1984 does not describe a wound. It describes a disclosure. The attack on the Darbar Sahib complex, on his reading, did not alter the relationship between the Sikh community and the Indian state so much as reveal what that relationship had always been, and his language is consistently the language of something becoming visible rather than something being inflicted. This is a claim about knowledge, and it accounts for the shift that follows.

That shift was, in the first instance, a departure from the framework in which he had spent the preceding fifteen years. His early organising had proceeded within the vocabulary of the Indian left: the student union, the youth organisation, the campaigns against hoarding, the language of exploitation and of the common person looted by those who administer him. His resignation letter of 1985 is still recognisably written in that idiom. What he came to argue was that the idiom could not name the thing it confronted. His objection, as far as it can be reconstructed, was not that class analysis was false but that it was insufficiently specific: that a framework locating the antagonism in economic relations could not account for a structure he identified as Brahmanical, nor for a nationalism whose object was the assimilation of difference rather than its exploitation, and that his former colleagues on the left had accordingly been able to read the Sikh movement only through the category of terrorism.

The pair of terms that replaces the earlier vocabulary is ghulami and azadi, servitude and freedom. The substitution is not decorative. A relation of exploitation is remediable by redistribution. A relation of servitude is remediable only by its termination. Once the antagonism is stated in the second form, the question of sovereignty is no longer one option among several. It is the form the question takes.

There is a persistent habit of reading that must be resisted at this point, because it is the habit that produces most of the misdescription this essay is about. Faced with a movement that states its aims in religious terms, the temptation is to treat the religious language as a covering, and to assume that a rational political demand lies underneath it awaiting extraction. It does not. For Khalra, as for many of those he wrote about, the religious language was the political language. There is no secular version of his argument sitting inside it. A reading which treats his invocation of Gurmat and of Dharamyudh as the culturally inflected packaging of a universal human rights claim has not decoded him. It has replaced him.

Exile and the argument against participation. Khalra left for England in 1991 under conditions of surveillance that had made continued work in Punjab impossible. He had three stated objectives: to inform international opinion, to reach a diaspora he considered disengaged, and to provide political education. In the same year he renounced his Indian citizenship by letter to the President, and applied for asylum in the United Kingdom while stating that the application was a temporary expedient and that he did not regard separation from the Darbar Sahib as a condition he could accept permanently.

The most substantial evidence of his thinking from this period is the critique of electoral participation he developed in 1991 and 1992, which is the closest thing to a systematic political statement he produced.

It had five components. The first distinguishes democracy from its apparatus: the right of a community to determine the terms of its own existence is the substance, and the availability of ballot boxes is not equivalent to it. The second is comparative and, in 1992, prescient. Khalra observed that the Palestinians possessed international recognition and formal representation while continuing to live in camps, and concluded that recognition secures nothing in itself. The third is strategic: to enter a contest whose terms are set by an adversary is to concede the terms. The fourth is historical: if relief had been the objective, the matter would have been settled in 1947. The fifth resists translation most stubbornly. Khalra argued that to swear an oath to a constitution one does not accept is to adopt the practice of saying one thing while intending another, and that a community which adopts that practice has abandoned the values in whose name it claims to act.

The fifth point is where the argument's character becomes unmistakable. It is not a calculation about efficacy. It is a claim that a particular tactical accommodation would destroy the thing it was undertaken to preserve, and it is stated in a register of integrity rather than of advantage. The proposal Khalra advanced in its place, that the question of Punjab's status be determined by referendum conducted through the existing Sikh electoral structure of the Shiromani Gurdwara Parbandhak Committee, confirms the reading. He was not withdrawing from procedure. He was disputing which procedure had authority.

He returned to Punjab in May 1992, having been granted asylum, and having written to the Commonwealth seeking protection while restating his support for self-determination for the Panth. The interviews conducted with participants in the movement during those years record, consistently, that they described their own motives in terms of sovereignty and of religious obligation together, and did not experience these as two separate commitments between which they were choosing.

The administration of killing. What Khalra returned to is best described not as a counter-insurgency but as a system. The Punjab Police operated a scheme of cash rewards for killings, and the aggregate figure most frequently cited, in excess of forty thousand payments by 1994, derives from United States State Department reporting. The consequence of a bounty structure is not merely that killings occur but that they acquire a productive logic. Quotas generate demand, demand generates supply, and the identity of the supplied becomes immaterial.

This is not violence as excess. It is killing organised as administration: scheduled, budgeted and distributed across a territory, with a price attached. The bodies were killed under a payment schedule and disposed of under a procurement schedule. The state paid for the killing and then paid for the wood.

It was the second payment that undid it.

What the film shows

The inciting substitution. The film establishes its protagonist as a bank employee who comes upon the disappearances by accident. The mother of a friend already murdered by police goes missing; enquiries at police stations and hospitals yield nothing; and her name is eventually found in a register at the local crematorium. The mechanism of the paper trail is then explained with admirable clarity: the state pays for the wood only when a name and an address accompany the body, and so the record exists because the disposal had to be accounted for.

The sequence is efficient, humane and dramatically sound. It also performs a substitution that turns out to be characteristic. The historical record indicates that Khalra's entry into the investigation came through the disappearance of colleagues from the cooperative bank who were themselves related to men active in the armed movement, which is to say that the disappearances reached him along a line that was already political. The film retains the bank and discards the line. What arrives instead is a bereavement of the most universally legible kind: a friend's mother, unaffiliated, whose disappearance requires no knowledge of Punjab to be felt as wrong.

Nothing here is falsified. A man did lose people he knew and did go looking. But the choice of which people, made at the point where a film establishes what kind of story it is going to be, determines whether the protagonist enters the narrative as a citizen encountering an outrage or as a participant tracing a pattern he already suspected. Satluj chooses the first.

The perpetrators are given a system. It must be said plainly, because the argument depends on it, that the film's account of the police is structural and intelligent, and that any critique characterising it as a story about a few bad officers has not watched it.

The opening sequence stages an execution in which two of the killings exist solely to raise a junior officer's tally, promotion being awarded on that metric. The Director General later secures the release of an imprisoned officer specifically in order to install him where brutality is required, which is a depiction of policy rather than of deviance. A constable who takes no pleasure in the violence is shown to be unable to leave, since he knows too much to be permitted to resign, so that the institution's coercive capacity is demonstrated to operate inwards as readily as outwards. And the senior officers discuss mass killing as a difficulty of public presentation, the Director General assuring reporters that the missing are probably driving taxis in Canada and answering complaints about corpses in Rajasthan's water by observing that people fall into canals.

This is a serious argument about what happens when an institution enjoys a monopoly of violence without accountability, and the film makes it well. The observation to be drawn is not that Satluj fails to think structurally. It is that the structure it identifies is the police.

The victim is given a conscience. Against this systemically conceived antagonist the film places a protagonist conceived in moral terms alone. The performance has been widely and rightly praised for its ordinariness. The character is legible as a man without special capacity or position who declines to look away, and whose heroism is presented as available to anyone.

The asymmetry this produces is the film's central formal fact, and it is worth stating slowly. The film explains the police. It shows how their incentives worked, how promotion was earned, how an officer was moved from a cell to a command because brutality was wanted in a particular district, how a man who disliked the whole business could not get out of it. Every action taken by the perpetrators is placed inside a machine whose workings the audience is invited to understand.

Khalra is placed inside nothing. He acts from decency. He is not shown to hold a position, only to hold a line. There is no scene in which he explains why the killings took the form they took, no moment at which he sets out what he thinks the relationship between his community and the state actually is, and no proposal from him beyond exposure and legal remedy.

The film therefore distributes understanding unevenly. It grants analysis to the men who did the killing and integrity to the man who documented it. And the effect of that distribution is to make the perpetrators comprehensible as a system and the victim comprehensible only as a temperament.

This is a common shape in accounts of rebellion and dissent, and it is worth noticing that it is a shape which sympathy produces more reliably than hostility does. An enemy concedes that you had a programme, because a programme is the thing that must be defeated. A friend, wishing to defend you, concedes only that you were wronged. And a wrong can be put right without anyone having to engage with what you actually thought. The film's very seriousness about the police is what makes the omission on the other side visible. A work that had thought carelessly about both parties would prove nothing. A work that thinks rigorously about one and warmly about the other has made a choice.

Who counts as a loss. There is a further condition operating in the film, quieter than the others.

Satluj grieves with real generosity, and the families it attends to are rendered with particularity rather than as illustration. But it admits the dead on a specific ticket, which is innocence. Its moral force derives from the demonstration that those killed had nothing to do with the insurgency. This is true of very many of the dead and it is effective as advocacy.

It is also a condition Khalra himself declined to impose. His documentation did not sort the innocent dead from the others. The objection he raised was to the manner of the killing and to the authority claimed for it, not to the political affiliation of the killed. A record built on innocence can only ever ask that the state stop killing the wrong people. His record asked something else.

A crowd, not a people. The figure of twenty-five thousand arrives in this film as an accumulation of separate households, each with its own missing person, its own frustrated enquiry, its own grief. What it does not arrive as is a people to whom something is being done.

The distinction is not sentimental. A collection of individuals who happen to share a misfortune can be compensated one at a time, and compensation exhausts the matter. A community subjected to a policy can make a claim that no quantity of individual redress settles. The film's Punjab is a place where a great many people were wronged separately. It is not a place where something was done to Sikhs.

Knowing, and proceeding otherwise. The film is not ignorant of the period it depicts. It is meticulous about the documented record, to the point of reproducing the Director General's remarks about taxi drivers in Canada in something close to their attested form. Its research is not in question.

What is in question is a film that knows this much and proceeds as though the political claim at the centre of its subject's life were not there. The word does not occur. The resolution offered is constitutional, and the director has described the completed work as one that shows the courts and the constitution in a favourable light and strengthens faith in them.

The shape of the story

The preceding section described a set of choices. This one argues that they were not only choices. A story has a shape before it has contents, and a substantial part of what Satluj does to Khalra was done by the shape.

The engine of the procedural. The film is, in its first movement, a tragedy, and in its second a police procedural. Following the abduction the narrative reorganises itself around a Central Bureau of Investigation officer pursuing the case, with the widow sustaining the work, and the drive of the remaining film is the assembly of evidence sufficient to be presented in court against the Director General and those beneath him.

Consider what that form requires in order to function. A wrong is done. An investigator gathers what is needed. The material is brought before a tribunal, and the tribunal responds. Every step assumes a legal order that is fundamentally sound and locally corrupted. The form can accommodate a corrupt police force, an obstructive bureaucracy, a compromised chain of command, and it can accommodate defeat, because the investigator may lose. What it cannot accommodate is the proposition that the tribunal itself lacks standing, because at that point there is nowhere for the evidence to go and the story has no engine.

This is why the second movement is decisive in a way that no individual scene is. The agent of hope in the latter half of Satluj is an officer of the Indian state investigating the Indian state, and the horizon towards which the film moves is a hearing before Indian courts. The audience is invited to want that hearing, and to want it urgently, which means the audience is invited into the premise that those courts possess the authority to settle what was at issue.

Khalra disputed precisely that premise. He had renounced the citizenship those courts adjudicate. He had argued against swearing to the constitution they administer. He declined to seek protection from the state he was investigating. The film does not argue with any of this. It builds a machine that will not run unless the opposite is assumed.

What tragedy discharges. The first movement contributes its own effect, and it is worth being exact about what that effect is.

Tragedy works by recognition and pity, and it completes itself in the feeling it produces. The spectator leaves having felt fully what there was to feel, and the feeling is the resolution.

A political claim requires the opposite condition. It requires that something remain unsettled. And unsettlement is what the tragic form exists to relieve. This is not an argument that the film should have withheld emotion, which would be impossible and inhumane given the material. It is an observation about what a form completes.

It is also a very old procedure in the way this country's history has been written. Episodes that cannot be fitted into the national story are seldom denied outright. They are admitted as suffering, sometimes lavishly, on condition that the suffering produce compassion rather than a demand. Compassion is exactly what the tragic mode is built to produce, and this film produces it abundantly.

Combine the two movements and the result is a story that begins with a grievance, proceeds through an investigation and ends in a plea for adjudication. That story has already decided, before any scene is written, that its subject was a man who suffered a wrong, because that is the only kind of subject such a story can hold.

The evidence of the titles. The film's own production history supplies the clearest available demonstration, and it can be read as a compressed history of this essay's argument.

The work was first titled Ghallughara, then Punjab 95, and finally Satluj. The first is a term from Sikh historical memory denoting catastrophe, applied within the tradition to the massacres of 1746 and 1762 and, by many, to 1984. To have used it would have placed the disappearances within a Sikh sequence: not an atrocity that occurred in Punjab, but the latest instance of a recurring assault understood historically and theologically by those who suffered it. The second title retains the place and the year and drops the sequence. The third names a river, and the river is in the story because the police threw bodies into it, some of which drifted into Rajasthan and prompted complaints about the water supply.

The trajectory runs from a term available only within a tradition, to a neutral marker of place and date, to a proper noun that carries no interpretation at all. Only the second of these substitutions was demanded by the censor, who required the removal of the word Punjab and thereby produced the third. The first substitution, the one that matters most, was made without instruction.

The counter-case for the form. It can be argued, and has been argued at length by historians who work on film, that cinema possesses competences of its own rather than merely failing at the competences of written history, and that condensation, composite characters and invented incident may serve historical truth rather than betray it. The argument holds and is worth conceding without reservation. Satlujtransmits the ordinariness of administered killing in a way no monograph achieves, and the sequence in which promotion is settled by adding two bodies to a tally does more work in ninety seconds than a chapter of exposition.

But that defence concerns compression, and the operation described here is not compression. To render twenty thousand deaths through one family is compression. To render a man's political analysis as a moral disposition is not a condensation of that analysis. It is its replacement by something of a different kind.

The censor's list, read as a document

The certification file on this film has been discussed almost exclusively as an obstacle: a quantity of interference to be reported, deplored and litigated. It is also a document, produced by an organ of the state, recording in its own hand what that state judged intolerable. Read as evidence rather than as impediment, the list is among the most informative sources available on this subject, and it says something other than what the public controversy has assumed it says.

The record. The film was submitted for certification in late 2022. It was cleared in 2023 subject to twenty-one cuts and a mandated change of title, a decision against which the producers appealed. The number of alterations required then grew to 127, an escalation which occurred while the film itself remained unchanged and which followed the exercise of a legal remedy. That sequence should be registered precisely. The demands did not merely expand. They expanded after the makers declined a conditional clearance and went to court.

Two features of the resulting list deserve attention before any interpretation is attempted.

The first is that it does not mention Khalistan. It does not require the removal of any reference to Sikh sovereignty, to self-determination, to a referendum, or to the political programme with which Khalra was associated. The most economical explanation is the correct one: there was nothing there to remove. But the absence remains significant, because it means the state's objections must be read for what they positively identify rather than for what one might have expected them to identify.

The second is the escalation itself. A list that grows from twenty-one to 127 against an unchanged film is not the output of a process applying settled criteria to a fixed object. It is the output of a process searching for grounds.

What the prohibited words have in common. Set out together, the proscribed terms fall into three groups, and the grouping is the finding.

The first group names places: Punjab, Delhi, desh. The second names institutions: state, system, Punjab Police, and a head of government. The third names legal standards: law, judiciary, extrajudicial.

What is being removed is not an argument. It is the apparatus by which an agent can be identified, situated and held answerable. A killing that occurs in an unnamed place, committed by an unnamed force, in violation of an unnameable law, is not thereby denied. It is rendered unattributable. The events survive the censorship. What does not survive is the possibility of saying whose events they were.

The suggestion that Punjab be replaced by here and there is the most instructive item on the list, because it is the point at which the logic becomes literal. The demand is that a specific place become a word whose reference depends entirely on where the speaker is standing, and which therefore designates nothing in itself. The state does not ask the film to lie. It asks the film to lose the capacity to point.

The third group completes the operation. To forbid the word extrajudicial is not to deny that killings took place. It is to remove the term by which a killing is classified as being outside the law, and thus to remove the frame in which the state's own legality becomes the measure. What remains is death without jurisdiction: terrible, but nobody's responsibility, and not a violation of anything, since the standard violated may not be named.

What certification is for. It is tempting to treat censorship as a purely negative business, a matter of subtracting material from works. It is not. A board that rules on what may be shown is also making a claim about who is watching: it presumes a viewer who would be inflamed, and in presuming him it participates in producing him.

The list under discussion presumes a citizen who can be trusted with atrocity but not with attribution. Nothing in the record suggests that the depiction of murder was itself the difficulty. What could not be permitted was the conjunction of the murder with a place, a force and a legal standard, because that conjunction constitutes an accusation, and an accusation implies a body competent to hear it.

Note what this means about the nature of the intervention. The board did not contest the film's account of the past. It legislated on the vocabulary in which a past may be spoken about at all, which is a more fundamental thing to do than suppressing any particular claim. The film had already recovered a record the state had buried. The certification process operated one level above that, on the conditions under which a recovery may enter circulation.

The convergence. It is at this point that the central proposition becomes visible in the documentary record rather than merely as an interpretation.

The film had reduced a conflict over the legitimacy of a state's authority over a community to a conflict over the conduct of a police force. This was an act of narrowing, performed for reasons that were prudential, well intentioned and, as will be conceded below, partly vindicated by events. The censor then required a further narrowing of the same kind and along the same axis, removing the terms by which even a police force could be located within a state and measured against a law.

The two operations are not identical in degree, in motive, or in moral character, and nothing here suggests otherwise. They are continuous in direction. Each removes a level at which responsibility might be attributed to a polity rather than to persons, and the film's own removal was the more fundamental of the two, since it went to the nature of the claim rather than to the vocabulary in which the claim could be voiced.

One item on the list does not fit the pattern: the reported demand for the removal of Gurbani, which names neither a place nor an institution nor a legal standard. If that report is accurate, it indicates that the objection extended beyond attribution to the presence of Sikh liturgical material as such.

After the platform

The certification process failed. It failed in the ordinary sense that it did not prevent the film from reaching an audience, and in a more interesting sense: by the time the board's demands were final, the instrument itself had been overtaken.

The makers declined the required alterations and abandoned the theatrical release for which certification is compulsory. The film was published instead on a streaming platform on 3 July in the form its makers had completed, without cuts. Streaming release in India requires no certificate, and so the film became available by the straightforward route of declining to enter the process at all. Whatever the board had spent three and a half years securing, it secured nothing.

Within forty-eight hours the film had been withdrawn for viewers in India, following a citation of security concerns by the Ministry of Information and Broadcasting. It was withdrawn internationally the following week. A public interest litigation seeking its restoration was filed in the Punjab and Haryana High Court.

The sequence is worth stating in this bare form because the contrast between its two halves is the substance of what follows.

Two modes of prohibition. The certification process, for all that it was arbitrary in operation, possessed four characteristics which taken together made it a legal instrument rather than merely an exercise of power.

It was adjudicative: a body applied stated criteria to a work and issued a determination. It was enumerated: its demands took written form and could be counted, which is why this essay has been able to analyse them at all. It was contestable by the maker: refusal carried a defined consequence and the determination could be appealed, as it was. And it was accountable in the minimal but real sense that its outputs entered the public record and could be held against it.

The withdrawal possesses none of these in the same form. A ministerial citation of security concerns addressed to a platform is not an adjudication of a work. It states a conclusion without applying criteria to content, and it produces no enumerable list against which anything can be argued. The maker is not the respondent, and has no standing in the transaction at all, which is why the challenge has had to be mounted as public interest litigation by third parties rather than as an appeal by those whose film it is. And what enters the record is a ground so general that it can neither be tested nor rebutted.

The film's history therefore contains something close to a natural experiment. The same work was subjected first to formal censorship and then to administrative suppression. The first was survived. The second was not.

This is the material point about the regulatory transition through which Indian film has passed since intermediary regulation was extended to streaming services in 2021. That transition is habitually discussed as a question of scope: whether a given category of content falls within a given rulebook. The more consequential change is in kind. Pressure applied to a platform, rather than adjudication applied to a work, produces the same outcome while generating no enumerated reasons and no respondent with standing. It is faster, cheaper and cleaner, and it leaves considerably less for anyone to read.

There is a loss of information here that ought to trouble people who have no interest in this film at all. A censor's list, for all its absurdity, tells you what a state is frightened of. A security citation tells you nothing. It removes the object and explains nothing about itself.

The consequence. Two conclusions follow.

The first concerns the film's accommodations. The narrowing did not purchase circulation. A film that had removed the sovereignty claim, that had located its resolution in the courts, and whose director had described it as an affirmation of constitutional faith, was suppressed within two days of reaching an audience, by a means against which its own compliance offered no protection whatever. Whatever the state was defending, the film's concessions did not address it.

The second concerns what the state was in fact defending. The censorship record, taken with the withdrawal, points consistently away from the interpretation that has dominated public discussion. The demands did not target a separatist argument. The withdrawal did not follow the appearance of one. What was intolerable, on the evidence, was the visibility of the Sikh grievance as such: not the claim that Punjab should be sovereign, but the claim that the Indian state, named as such, killed a great many Sikhs, named as such, in a place that may be named.

One grammar, two erasures

The two removals are now specific enough to be set against one another.

The film relocates the antagonism. It offers a rigorous account of an institution that kills without accountability, and it declines to offer any account of the relationship between the state that maintains that institution and the community the institution was killing. Its Punjab is a place where policing went uncontrolled. It is not a place where a polity acted upon a people.

The censor removes attribution. The proscribed vocabulary named places, institutions and legal standards, and its effect was to leave the killings intact while dissolving the capacity to say whose killings they were, in what place, in breach of what standard.

Stated together, the two operations disclose their relation. The proposition that neither will carry is this: the Indian state killed a great many Sikhs, as Sikhs, in Punjab. The censor forbade the first half of it. The film declined the second. Between them the sentence cannot be assembled, and it is the sentence around which Khalra's own writing was organised.

A necessary qualification. That two operations produce a compatible result is not evidence that they were undertaken for compatible reasons, and nothing here imputes to Honey Trehan or his collaborators any sympathy with the apparatus that suppressed their work. The evidence runs emphatically the other way. The film was completed across four years of institutional obstruction, at considerable professional cost. Its makers declined the alterations demanded of them and forfeited a theatrical release rather than comply. The director has characterised the certification process as itself a continuation of the violence the film depicts.

The argument here is not about motive. It concerns the conditions under which a claim becomes tellable, and the fact that a story built to survive those conditions carries their impress whatever its author intends. A film-maker deciding what can be filmed about Punjab in 1995 is not choosing freely among available accounts. He is choosing among the accounts that remain sayable, and the boundaries of that field were drawn by the same power that later refused him a certificate. This is a convergence of effect produced by a shared constraint, not a collusion.

A rival account of authority. What Khalra was doing needs a description that neither the film's vocabulary nor that of human rights advocacy can supply.

He was not asking for better treatment within an order he accepted. His distinction between democracy as a community's right to determine the terms of its own existence and democracy as the availability of ballot boxes is a claim about who constitutes the polity in the first place. His objection to swearing oaths to a constitution one does not accept is a claim about the integrity of a different account of authority and its capacity to survive tactical accommodation. His proposal that Punjab's status be settled by referendum through the existing Sikh electoral structure was not a withdrawal from procedure but a dispute about which procedure had the right to decide.

That is a claim about jurisdiction. And courts do not adjudicate claims about jurisdiction advanced against themselves. They cannot, because to hear such a claim would be to concede that it might be well founded. What courts do with rival accounts of authority is end them, and the ending is not accomplished by argument but by the force that stands behind a judgment.

The Indian state's response was of the ordinary kind. It killed the man who carried the rival account, and it has resisted every subsequent effort to establish the scale of what was done.

The film's response is of the same kind by other means and with contrary intentions. By constructing a narrative whose engine is the assembly of evidence for presentation before Indian courts, and whose horizon of hope is adjudication by them, it does not argue against Khalra's account of authority. It renders his story in a form within which that account cannot be stated, and then supplies a resolution in which the very order he disputed appears as the guarantor of his significance. The disagreement is not defeated. It is made unavailable.

The fragment and the inherited state. The national narrative has a long-established procedure for material of this kind. The fragment that cannot be integrated is not usually denied. It is admitted in the register of pathos, as tragedy rather than as claim, and thereby absorbed. Khalra as a man killed for uncovering an atrocity is admissible on those terms. Khalra as the bearer of a competing account of legitimate authority is not, and the film does not attempt to make him so.

And there is a further irony. The order into which the film's Khalra is finally absorbed is one that inherited its coercive apparatus from the colonial state substantially intact. The Punjab Police of 1995 were not an aberration within a new republic; they were an old instrument in continuous use. A narrative that offers the constitution and the courts as the resolution of this story is not merely accommodating a censor. It affirms exactly the continuity that Khalra's own writing identified as the problem, and it does so in the name of the man who identified it.

The limits of the claim. The claim established here is structural, and its limits should be stated. It is not that the film's makers share the censor's purposes, which they plainly do not. It is not that a different film was available to them, which will be doubted below. It is that two operations of very different character removed adjacent levels of the same attribution, that the removals were complementary rather than merely simultaneous, and that the film's removal was the more fundamental of the two.

The cost of the acceptable victim

That a substitution has occurred is now established. What was exchanged for what?

The temptation is to treat the matter as one of accuracy, and to say that the film gives an incomplete portrait of a man whose fuller portrait is available elsewhere. That understates the case. What has been altered is not the quantity of information conveyed about Khalra but the kind of claim his life is made to advance. The film's Khalra asks the Indian state for justice. The Khalra of the record questioned the state's title to grant it. These are not two versions of one demand, differing in emphasis. They are demands of different orders, and the second cannot be reached by adding detail to the first.

Defended into nothing. There is a way of defending a persecuted man that finishes the work of those who persecuted him, and it is worth setting out because it is the operation this film performs.

The defence proceeds by insisting on what is general. He was a husband. He was a decent man. He was a citizen entitled to the protection of the law like anyone else. Every one of these things is true, and every one of them is a description under which he is interchangeable with any other person.

What the defence cannot accommodate is the particular thing he was: a Sikh who held that the relation between his community and the Indian state was one of subjection rather than of citizenship, and who acted on that view for twenty years. The persecutor wished to destroy the man. The defence, offered in good faith, dissolves the thing the man actually was and preserves the citizen. Between them very little survives.

Satluj does not deny any of this about Khalra. It offers, in the register of sympathy, a Khalra from whom it has been subtracted. And the subtraction is performed by his admirers, which is why it is harder to answer than the censorship.

The condition of being understood. Why should the particular have been unavailable? Not, on the evidence, because the censor forbade it. The answer lies in a prior condition governing what kind of person can appear as a political subject in a national story at all.

The condition is that religion be treated as a private matter, held internally, and admitted to public argument only once it has been so treated. A man whose political claim cannot be separated from his religious tradition is not, on this arrangement, someone advancing an unusual claim. He is someone who has failed to advance a claim of the recognised kind, and who must be translated before he can be heard at all.

In the Sikh case that translation has a documented history. The rendering of Sikh thought into the categories used for religions elsewhere, begun under colonial administration and consolidated since, produced a Sikhism recognisable as a religion in the modern sense precisely by detaching it from the political claims with which it had been continuous. What was presented as description was actually a change to the thing described. Within the tradition itself, the devotional and the political are not two spheres between which Sikhs mediate. They are aspects of a single account of where authority lies. To read Sikh militancy as religion overspilling its proper bounds is already to assume bounds the tradition does not recognise.

Satluj does not argue any of this. It inherits it. A film that wishes its protagonist to be intelligible to a national audience must present him in the form in which political subjects are intelligible there, and that form requires his Sikhi to appear as background, as culture, as the texture of a life rather than as the substance of an argument. The result is a Khalra who is incidentally Sikh, and the word incidentally is where the whole difficulty lies.

What Shahadat is not. The clearest instance is the treatment of his death.

A shahid is not simply a person who died for a good cause. He is a person whose death is claimed by a community as a deposit within a continuing account of what that community owes and is owed, and the claiming is what makes the death a martyrdom rather than a killing. Khalra's own speech and writing place him within that account explicitly, and the record indicates that he anticipated his death in those terms in the days before it.

Rendered as sacrifice in the general sense, the death becomes a loss: grievous, unjust and finished. Rendered as Shahadat it is a transmission, and what it transmits is an unsettled obligation. The difference is not decoration. It is the difference between a death that concludes and a death that is owed something, and only the second generates a claim in the present.

What replaced it. If the sovereignty claim was removed, something was put in its place, and it is worth being precise about what.

The language of human rights is often treated as though it were the neutral container within which any grievance may be lodged. It is not neutral, and it is not general. It acquired its present authority relatively recently, and it did so as an alternative to the politics of self-determination that had previously organised claims of this kind, not as the wider category to which those claims belong.

The substitution has a cost. Rights language addresses the injured individual and secures protection for him. It does not address the arrangement that produced the injury, and by offering redress it tends to displace the question of that arrangement altogether. It also has entry requirements. The victim qualifies as a victim by being innocent and passive. The perpetrator is individuated, named, and prosecuted. And a rescuer arrives from within the system to put things right. Each of those three is discernible in this film, and the third is the officer from the Central Bureau of Investigation.

So the exchange can be stated exactly. A claim that the Indian state's authority over Sikhs was illegitimate, advanced in a religious idiom by a man who had renounced its citizenship, has been converted into a claim that the Indian state failed in its obligations to a citizen. The second claim is true. It is also the claim Khalra spent the last years of his life declining to make.

Speaking for. There is a difference between portraying a man and standing in his place, and a film of this kind inevitably does both. It will be, for most of its audience, the only account of Khalra they ever encounter, which means it does not merely describe him. It speaks as him, and on his behalf, to people who have nothing else to go on. That second function is exercised whether or not it is claimed, and it is most consequential when those exercising it believe themselves to be transparent.

Recognition granted. Recognition conferred is not the same thing as recognition won, and it leaves untouched the asymmetry by which one party remains in a position to confer it. The freed man who did not free himself is recognised at the pleasure of the power that recognises him, and that power may withdraw its recognition whenever it likes, as this one duly did within forty-eight hours.

Satluj confers recognition on Khalra. It confers it magnificently, and the affection in it is unmistakable. But the recognition is conferred by the national story on a figure who has first been made admissible to it, and admissibility was purchased by removing the ground he stood on. He is honoured as a man the nation can be proud of having produced, which is precisely the relation he spent his life disputing.

The audit

The preceding sections have concerned what Khalra believed and how a film disposed of it. This one concerns what he did, and it is offered partly in correction of this essay's own emphasis. A critique preoccupied with the removal of a man's convictions risks reproducing, in a different key, the reduction it objects to: treating Khalra as a holder of opinions rather than as the author of a method. The method was the achievement. It has been described in outline by everyone who has written about him and analysed by almost no one, and Satluj, which reproduces it accurately, reproduces it as plot.

He did not testify. Testimony is the mode in which the survivor speaks. It carries the authority of having been present and the vulnerability of resting on the person who bears it. Documentation of atrocity conventionally proceeds by accumulating such accounts. The family says the son was taken. The neighbour says he saw the vehicle. The record is built from the aggregate of what people are prepared to say.

This method is indispensable, and under the conditions of Punjab in the early 1990s it was almost useless. A witness may be killed. A family may be induced to stop speaking. And the state's standard reply, that the missing had gone abroad, was constructed precisely so that no testimony could disprove it: a family's insistence that their son is dead can always be met with the observation that the family cannot know.

What Khalra did instead was to identify a point at which the state was obliged to write something down.

The obligation arose from cost. A body must be disposed of. Disposal consumes wood. Wood must be purchased. Purchase must be accounted for. And the account requires an entry naming what was burned. The cremation registers recording bodies as unidentified were not evidence that anyone intended to create. They were the residue of a payment. Khalra obtained them and cross-checked them against the municipal receipts for the firewood, and the second series is the more remarkable of the two, because it was kept by officials with no interest in the first and no notion that they were maintaining a record of anything but expenditure. Two independent paper trails for the same corpse, generated for unrelated reasons by people who could not have colluded.

The result does not depend on any witness. It cannot be retracted under pressure, because nobody said it. And it cannot be met by the claim that the missing went abroad, since the state's own paperwork records that they were burned.

He documented 6,017 such cremations in Amritsar district alone, and extrapolated to figures for the state which he stated as estimates. In January 1995 he made the finding public and challenged the head of the Punjab police to a debate. In September he was abducted from outside his house while watering his plants. In October he was killed.

The receipt nobody was guarding. The insight underneath the method is worth drawing out, because it is more general than the case.

States keep two kinds of record about the same event, and they do not usually keep them in the same place. There is the record of what happened to a person, which is guarded, curated and, where necessary, falsified. And there is the record of what something cost, which is kept by clerks, filed by accountants, and guarded by nobody, because it is not understood to be about persons at all.

The Punjab Police controlled the first. They did not control the second, and had no reason to think they needed to. A receipt for firewood is not evidence of a killing until somebody puts it next to a register, and the two functions of the administration, accounting for people and accounting for money, were not coordinated.

Khalra coordinated them.
There is a further point about the register entry itself. To record a body as unidentified is an administrative act with a legal effect. It produces a corpse to which no claim attaches, and therefore a death that generates no obligation in anyone. What the registers record is not merely the fact of killing. It is the precise moment at which the killed were placed beyond the reach of anyone entitled to ask after them. And what Khalra performed on those entries was the reversal of that act. To restore a name to an unidentified body is not to add information. It is to reconstitute a person to whom something is owed.

Why the method matters. There is a difficulty at the heart of atrocity documentation which no amount of diligence resolves: the events that most require witnesses are the events that destroy the conditions under which witnessing is possible.

Khalra did not resolve that difficulty. He went around it. He built a case that requires no one to be believed. This is, in the strict sense, a contribution to the practice of documenting state violence, and it has been treated as biography rather than as method. Investigators working on disappearances elsewhere have arrived independently at comparable insights, and it is a small scandal that the Punjab case is cited for its scale and almost never for its technique.

The silence he broke had been created at the very beginning, by a clerk writing a word beside a body. It was undone at the same level. He did not oppose the state's account with a counter-narrative. He opposed it with the state's own paperwork, which is why the finding has proved so difficult to dislodge.

The film's version. Satluj renders all of this accurately and, in the ordinary sense, well. The register is found. The payment mechanism is explained. The significance is conveyed. The audience understands what has been discovered and why it matters.

What the audience does not encounter is the discovery as an act of thought. The sequence functions within the film as the moment at which the protagonist obtains what he needs in order to proceed, which is the structural role such a scene occupies in an investigative story. Evidence is located. The plot advances. Rendered thus, the registers are a find rather than an argument, and the man who assembled them is a person who looked in the right place rather than one who understood something about how states account for what they destroy.

The pattern is by now familiar. Analysis is once again converted into disposition, and a mind is once again rendered as a conscience.

The strongest case against this essay

An argument of this kind incurs an obligation to state the case against itself at full strength rather than in a form convenient to answer. Five objections are available, and the fourth is serious enough that the conclusions must be qualified in light of it.

The film exists. No comparable account of the disappearances has reached a mass audience, and none was likely to. It was made across four years of obstruction, at professional cost, by people who declined the alterations demanded of them and forfeited a theatrical release rather than comply. To dwell on what such a work omits, from the safety of a British desk, is to apply a standard that costs the critic nothing and cost its objects a great deal.

It worked. Following the withdrawal the film was screened informally across Punjab, and the director's account is that it functioned as a balm and drew Hindu and Sikh audiences together to watch and discuss a shared history. Whatever the text withholds, the event did political work. A critique of the script's silences is beside the point if the screening supplied what the script did not.

The alternative was nothing. A film that named the sovereignty claim would not have been financed, certified, distributed or, plausibly, completed. The choice was never between this film and a fuller one. It was between this film and silence, and silence serves the families still searching for their dead considerably less well.

The family approved it. Paramjit Kaur Khalra stated publicly that the Khalra family had approved the version released, that it was the same version originally screened for them, and she thanked the director for refusing to compromise the film's integrity under pressure. She said it authentically portrayed both the truth of the unclaimed bodies and her husband's struggle to bring the truth of the Sikh struggle before the world. The person with the greatest standing to object has instead endorsed. An argument that the film erases Khalra must reckon with the fact that his widow does not think so.

The critic's position is not innocent. The Sikh diaspora has in part constituted itself around the circulating image of the tortured Sikh body, so that its political work and its traffic in such images are difficult to separate. A diasporic critique demanding a more militant Khalra participates in that economy. Worse, it demands from film-makers working under an Indian censorship regime a risk that the critic does not bear. And there is a sharper version still. This essay has argued that the film claims Khalra for a project by rendering him in a form its purposes require. The reconstruction offered above performs an operation of the same species. It selects from an incomplete and partly partisan record, and it produces a Khalra congenial to the argument being made. If the film murders him again, the critique may only be conducting a rival funeral.

The reply. The first, third and fifth objections should be conceded, the third almost entirely.

There is no version of this argument in which the film ought not to have been made, and none in which its makers acted badly. The convergence identified above is a convergence of effect under shared constraint, not of purpose. If the counterfactual is silence, the film is a gain, and this essay does not dispute it.

The fifth objection is conceded in substance and answered only in scope. This essay does not claim to possess the true Khalra, and its argument does not require that it should. The claim is narrower and consequently more robust: that a set of positions which he stated in his own words, in identifiable documents, is absent from the film, and that their absence changes the kind of claim his life advances. That is a verifiable proposition about a text, and it survives whatever one concludes about who is entitled to claim him. The reflexive objection is nonetheless a good reason to hold the argument at the level of documents rather than of essence.

The second objection requires a distinction. That a film provokes a political conversation is not evidence that it contains one, and the informal screenings tend to establish the case rather than to unsettle it. Audiences arrived at those screenings already possessing the frame the film withholds, and supplied it. What the event demonstrates is the vitality of the surrounding memory culture, not the completeness of the text, and the distinction matters because the text will outlast the occasion.

The fourth objection requires this essay to qualify itself, and it should do so plainly.

Paramjit Kaur Khalra's endorsement establishes something that cannot be argued past: that the film did not wrong the people with the strongest claim to be wronged by it, and that its makers dealt honourably with them. Any implication to the contrary would be both false and offensive, and none is made here.

What her endorsement does not settle is the separate question of what a representation transmits to an audience possessing no other source. The bereaved may approve a portrait for reasons that have nothing to do with its adequacy as a historical account, and their approval is not a finding about the work. It is worth noting, too, that her own public framing exceeds the film's at several points. She has described what occurred as Sikh genocide, called upon the Akal Takht to convene a people's commission to determine the true numbers, and demanded accountability, the revocation of honours and social boycott for those responsible. That is a Sikh institutional and openly political register, and it is not the register in which the film resolves.

The most accurate reading is therefore not that she disagrees with this essay's premise. It is that she appears to regard the film as a contribution to a struggle whose terms she states elsewhere and in her own voice, rather than as that struggle's full statement. The difficulty is that most of the film's audience will encounter the contribution and never the statement.

Afterlives

The film was removed from circulation and the argument it had declined to make began immediately to be made in its name.

This is the final piece of evidence, and it is the most difficult to assess, because it can be read as vindicating either party.

What returned. Satluj does not name Khalistan. Within days of its withdrawal, the question of Sikh sovereignty was being publicly debated in connection with it in a manner the film's release had not produced.

The withdrawal itself supplied the argument. A ministry citing security concerns to a platform, in respect of a film about killings the state has never fully acknowledged, demonstrated in a single administrative act what three hours of drama had been constructed to avoid asserting: that the relationship between the Indian state and this history remains something the state polices rather than something its courts have settled. Khalra's own analysis, which the film had removed, was performed by the government at the moment it removed the film.

Whatever is refused a place in the ordinary business of public speech does not thereby cease to exist. It comes back from outside, and it comes back less manageably than it would have arrived had it simply been allowed in.

What audiences supplied. The second afterlife is the more consequential. Following the takedown the film was screened publicly across Punjab by community organisations, political groups and religious bodies, and the director's account is that these screenings functioned as a balm and that Hindu and Sikh audiences watched and discussed together.

Two things should be said, and they point in opposite directions.

The first is that the screenings constitute a mode of circulation the state's instrument was not designed to reach. Prohibition addressed to platforms presupposes that distribution is centralised and therefore governable at a chokepoint. A film projected in a village hall by people who obtained it themselves lies outside that architecture entirely. A work that survives the machinery built to prevent it acquires a significance beyond its contents, because its continued existence becomes part of what it says. Satluj now carries that significance, and its makers can claim no credit for it. It was conferred by the state that suppressed the film and the audiences who ignored the suppression.

The second is that these audiences did not receive a frame. They brought one. The film was screened by Sikh institutions, by political organisations, and under the auspices of a memory culture that has never required a film's assistance to understand what happened in Punjab. The village screening was not the transmission of the film's meaning to a public. It was the encounter of a text with a public that had its own.

Which is why the event cannot settle the question. That audiences supplied the political register the film withholds demonstrates the vitality of the memory culture surrounding it. It does not demonstrate that the film contains that register, and the two will not remain together. The screenings are occasions. The film is a document. For a Sikh audience in Punjab in 2026 the film is one item within a dense inheritance. For a viewer in Toronto or Southall in 2046, encountering it alone, it will be the inheritance.

The undecidable. The strongest case for the film's makers is available at this point, and it should be stated because this essay's argument does not defeat it.

It may be that the accommodations were correct. A film that named the sovereignty claim would very likely not exist. The film that exists reached audiences the other would not have reached. And its removal generated a public argument about precisely the question it omitted, conducted by people better placed than any film to conduct it. On this reading, restraint in the text was the condition of provocation in the world, and the strategy succeeded in a form its authors could not have designed.

The case against is that this outcome was not produced by the film. It was produced by the state's decision to suppress it, which was not within the makers' control and which a differently constituted government might not have taken. Had the film simply streamed unremarked, the accommodations would have purchased a wide audience for an account of Khalra with the politics removed, and nothing would have returned at all.

There is no adjudicating between these readings yet, because the material for adjudication does not exist. What can be said is that the film's political effect to date has depended on a suppression it did not intend and could not have arranged, and that a strategy vindicated by its opponent's mistake is not thereby shown to have been sound.

The lamp

This essay began with two facts placed side by side, and it ends by proposing that their conjunction is the most informative thing about the case.

A film about Jaswant Singh Khalra was made which did not name the political claim for which he was known, which located its resolution in the Indian courts, and whose director described it as an affirmation of constitutional faith. That film was refused certification, then cleared subject to twenty-one cuts and a change of title, then, following its makers' appeal, faced a demand for 127. When they declined and published it without a certificate on a platform where none was required, it was withdrawn within two days on a ministry's citation of security concerns, and internationally within the week.

Nothing in that sequence is explicable by the presence of a separatist argument, because there was none to be found. The proscribed vocabulary named places, institutions and legal standards: the words by which a place is identified, a force located and a law invoked. What was intolerable was not the proposition that Punjab should be sovereign. It was the proposition that the Indian state, named as such, killed a great many Sikhs, named as such, in a place that may be named. The film had already declined the second half of that sentence. The censor forbade the first. Between them it could not be assembled, and it was the sentence around which Khalra organised his life.

The second claim concerns the price of the accommodation. It failed. It did not secure circulation, and the suppression it was calculated to avert arrived by a route against which compliance offered no protection. What it did achieve was a Khalra rendered fit for a national audience: a man of conscience rather than of analysis, whose objection was to the conduct of a police force rather than to the authority claimed over him, and whose story concludes in a courtroom he had himself declined to petition. The film's structural intelligence about policing is what makes this legible as a choice. A careless work would have proved nothing.

What the whole affair will finally be taken to have been about is still open, and the argument over this film is where it is being settled. The registers were written first and the receipts filed second. Khalra found them and put them together himself, and did it in a manner that has proved almost impossible to dislodge, because he built a case that requires nobody to be believed. What remains undecided is what that case will be understood to have concerned. On the present evidence it will be understood to have concerned impunity, which is true, and not sovereignty, which is what he said it concerned.

Two details are worth holding together at the close.

Khalra's most cited speech turns on a lamp lit against a darkness that has fallen for the first time, and the lamp's significance lies in its refusal to concede that the darkness is total. The figure has been received widely as an image of individual moral courage, and it will bear that reading. It also bears another, in which the light is not one man's integrity but the collective refusal in which he located himself, and which the tradition he belonged to had a name for. Which reading the coming decades take is precisely the question still open.

And the film is named for a river. The state's police used the Satluj to dispose of bodies, some of which drifted into Rajasthan and prompted complaints about the water. A work first titled with a word from Sikh historical memory for catastrophe, then retitled for a place and a year, and finally for the water into which the dead were thrown, records in its own title the trajectory described here: from a claim intelligible within a tradition, to a neutral marker, to a proper noun that interprets nothing. Only the second of those substitutions was ordered. The first, and the most consequential, was made freely, and by people who admired him.

That is the shape of the problem. A man may be honoured out of existence, and the honouring is harder to answer than the suppression, because there is nobody to accuse.

The commentary that accuses this film of murdering Khalra a second time is right about the outcome and wrong about the agent. Nobody killed him again. He was honoured, generously, expensively and at real personal risk, by people who admired him and who dealt fairly with his family, and the honouring removed him more completely than the censorship could have managed. That is a harder thing to say than an accusation, and a harder thing to remedy, because there is no one to hold responsible and nothing to appeal against. It can only be answered by saying, repeatedly and in public, what he actually thought.

William Gomes is a British Bangladeshi writer and human rights campaigner based in York. He has spent much of his working life documenting enforced disappearance, extrajudicial killing and custodial torture, first in Bangladesh and later with an international human rights organisation and an Italian news agency, and he continues to work with human rights bodies on accountability for state violence. He writes on racism, displacement and impunity, and his journalism has appeared in the Huffington Post, openDemocracy, the Tablet, the Big Issue, the Yorkshire Post, Yorkshire Bylines, Counter Currents and CounterPunch. He holds a master's degree in Refugee Care from the University of Essex and the Tavistock and Portman NHS Foundation Trust, and is currently completing a doctorate at the University of Essex.

   Comments0