Two quotes before we begin:
The state of exception has today reached its maximum worldwide development. The normative aspect of law can thus be obliterated and contradicted with impunity by a governmental violence that, while ignoring international law externally and producing a permanent state of exception internally, nevertheless still claims to be applying the law.
— Giorgio Agamben, State of Exception (2005)
Justice is an experience of the impossible.
— Jacques Derrida, Force of Law (1990)
I. The Room with No CCTV
On the evening of 30 March 2026, a group of filmmakers and cultural workers sat in the office of the Director General of the Department of Film and Publications (DFP) in Dhaka, waiting. They had been there since two in the afternoon. They were owed money: back payments for documentaries commissioned by the Ministry of Information on the songs of the July uprising, the revolution that had toppled fifteen years of authoritarian rule just eighteen months earlier. They had done the work correctly. They had followed the rules. They had come, as summoned, to collect what was theirs.
What happened next has the quality of a parable, except that parables do not leave people with burst eardrums.
A DFP cameraman, Md Moshiur Rahman, walked into the Director General’s office uninvited. He declared himself a big man of the department. In front of the Director General, he demanded a percentage, a cut, from the filmmakers’ payments. They refused. He made a phone call. Within minutes, fifty to sixty men, who identified themselves as members of the ruling party’s student organisation, were inside a government building, beating artists, photographers, and human rights workers. Heads were split open. Blood pooled on the government floor. One filmmaker described shielding a colleague, turning to find another with blood streaming from his face. He screamed for the Director General to call the police. When the police arrived and asked for CCTV footage, they found the hard disk had been formatted. Wiped. The record, deleted.
Among the beaten was Mosfiqur Rahman Johan, a photographer whose archive contains thousands of images documenting enforced disappearances under the Hasina government. The faces of the missing. The waiting families. The government’s own leaders had wept at his exhibitions less than a year before. He was now beaten in a government office by people who claimed to represent the revolution he had documented.
One of the filmmakers described this afterward with five Bengali words that need no academic translation: “ক্ষমতা পাইয়া দানব হইয়া” (having got power, they became monsters). The Bengali phrase captures an ancient moral intuition about corruption’s speed.
II. The Deletion Is Not a Detail
The formatted hard disk is the essay’s first and most important image. It is more than evidence of a cover up. It is the emblem of a method. A method that does not change with the party in power. A method that transfers from one government to the next, because it is not a property of the government. It is a property of the state.
Consider what the deletion required: a DFP employee who knew where the DVR was; a supervisor who permitted the access; and a system of mutual protection that made the formatting rational, because no one expects consequences. That expectation of no consequences is not born in a day. It is institutional. It is structural. It is the operating assumption of every arm of the Bangladeshi state that touches coercive power.
This essay’s argument is that the events of 30 March 2026 at the DFP are not an isolated incident of party thuggery. They are a diagram: a small, legible version of something much larger — the systematic dismantling, in the weeks following Bangladesh’s first elected parliament under the new government, of every institutional mechanism that might have made the state accountable to the people who bled for it in July 2024.
The French philosopher Alain Badiou calls this a restoration period: the moment when the revolution’s own children begin to manage its energy instead of transforming it. The July Uprising was, in his terms, an Event: a genuine rupture that opened new possibilities for justice, institutional redesign, and human rights accountability. An Event requires fidelity, a sustained commitment to what it made possible. What follows an Event without fidelity is not transformation. It is Thermidor. It is the restoration of the old relations of power, dressed in the new government’s clothes.
The 133 ordinances of the Dr. Younus’ interim government were, taken together, an attempt at fidelity: an effort to formalize the July Uprising’s consequences in law. The BNP government’s special parliamentary committee reviewed all 133. Its report, presented to parliament on 3 April 2026, recommended converting 98 into law, amending 15, and allowing 20 to lapse or be repealed. The pattern of this selection — which ones were kept and which ones dropped — is the political theory of this moment in Bangladesh made visible.
III. The Referendum That Lapsed: Technical Legality vs. Sovereign Will
Before we go further, we need to address an argument that has been circulating online with some confidence, most loudly from government apologists. The argument goes like this: the referendum ordinance was event-specific. It was passed to enable the 12 February 2026 referendum on the July National Charter. That referendum happened. Yes won with approximately 68 percent of votes. The ordinance’s purpose is therefore complete. Letting it lapse is routine legal housekeeping, the equivalent of not renewing a single-use camera once the film has been developed. Nothing to see here.
Let us give this argument the serious treatment it deserves and then take it apart.
What the Argument Gets Right
It is technically correct that the referendum ordinance was structured around a specific event. It is correct that under Article 93 of the Bangladesh Constitution, ordinances issued during the absence of parliament lapse if not ratified within thirty days of parliament’s first sitting. It is correct that the BNP government did not need to “cancel” the referendum. The vote happened. The result stands. The July Charter has popular democratic legitimacy. A sunset clause is not a repeal. New bills will eventually follow from the referendum result. On these procedural facts, the apologists are right.
But procedural accuracy is not the same as political honesty. And political honesty is not the same as philosophical coherence.
What the Argument Gets Catastrophically Wrong
Here is the question the procedural defense refuses to ask: what happened to every other ordinance that lapsed alongside the referendum one?
The enforced disappearance prevention ordinance lapsed. The National Human Rights Commission independence ordinance lapsed. The Supreme Court Judges Appointment Ordinance, which would have removed the Prime Minister’s effective control over High Court appointments, was actively repealed, not allowed to lapse. The Supreme Court Secretariat Ordinance, which would have freed lower court judges from Ministry of Law control, was also repealed. The Anti-Corruption Commission independence ordinance was allowed to lapse. These are not single-use cameras. These are the foundational accountability mechanisms of a democratic state. They had no sunset clause built into their purpose. They were dropped because the government chose to drop them.
So, when the apologist says “the referendum ordinance lapsed because it was event-specific,” they are technically correct about one ordinance and completely silent about nineteen others. This is not analysis. It is misdirection, a conjuror’s trick that asks you to watch the right hand while the left hand is busy.
But let us go deeper than political dishonesty, because the philosophical problem is more serious than the political one.
Carl Schmitt, the German legal philosopher whose ideas should make every democrat uncomfortable, but whose analytical precision remains useful, argued that the sovereign is he who decides on the exception. The exception is not a breakdown of the legal order; it is its foundation. Every legal system has zones where ordinary rules do not apply. The sovereign is the one who decides where those zones are and who lives inside them. This is not a description of dictatorship. It describes every state, including democracies.
The July Charter and the referendum that endorsed it were the Bangladeshi people’s attempt to decide the exception for themselves: to say, these are the principles that will govern us, including the principle that no security force is above civilian oversight. The referendum ordinance was the legal vehicle for that declaration of popular sovereignty. Letting it lapse without immediate legislative follow-through is not housekeeping. It is the new sovereign, the BNP executive, reasserting its right to decide the exception. It is the sovereign saying: the people expressed a will in February, but I will determine what legislative consequences that will has.
Jacques Derrida argued that every constitution rests on a founding violence it cannot fully acknowledge, an act of power that precedes and exceeds the law it subsequently produces. The July Uprising was Bangladesh’s most recent founding violence. The referendum was its ratification. The BNP government’s selective implementation of the referendum’s mandate is the new founding act: a constitutional settlement that preserves executive power while eliminating the mechanisms of accountability that the founding violence demanded.
Barrister Nowshad Zamir, a BNP parliamentarian and a member of the same committee that voted to lapse these ordinances, put it with legal precision in his written dissent: “If we don’t do this, it will be correctly read as a retreat from human rights obligations by the BNP government in its very first month. This is not in our best interest.” He was outvoted. The retreat was made. And it was read exactly as he predicted.
The apologist’s answer to this is: wait for the bills. They will come. But Barrister Zamir’s dissent explains precisely why waiting is not neutral. The Human Rights Commission ordinance, if it lapses, means Bangladesh loses its only path to GANHRI Status A, international certification of an independent human rights institution. The enforced disappearance ordinance, if it lapses, means Bangladesh has ratified the International Convention for the Protection of All Persons from Enforced Disappearance and then declined to enact its domestic equivalent. And the Supreme Court appointments ordinance, once repealed, does not wait in a drawer for a better day. Its repeal is a constitutional act that reasserts executive dominance. That act cannot be “undone” by procedural patience. And let us be clear on something else: a sunset clause for a referendum is procedural housekeeping; a sunset clause for judicial independence and enforced disappearance prevention is the sovereign deciding the exception once again.
A single-use camera, once used, has served its purpose. The accountability architecture of a democratic state is not a single-use camera. It is the entire plumbing system of the house. Letting it rust is not maintenance. It is a decision.
IV. The Constitutional Architecture of Impunity
Let us be precise about what was built and what was unbuilt in March and April 2026. Because precision matters here. Outrage without precision is just noise.
The Bangladesh constitution, as it now stands after the BNP government’s decisions, contains at least five structural features that make the prosecution of security force members for systematic human rights violations functionally impossible. Not difficult. Not unlikely. Functionally impossible.
First: the executive controls judicial appointments. Article 48(3) means that the President acts on the Prime Minister’s advice for virtually every executive act. Article 95 means that High Court appointments ultimately go through the Prime Minister. The repealed judicial appointments ordinance would have created a seven-member council, chaired by the Chief Justice, to shortlist candidates, effectively breaking the Prime Minister’s control. It is gone. Courts whose judges are effectively appointed by the executive cannot be trusted to rule against the security forces that protect that executive.
Second: lower court judges remain under Ministry of Law control. The repealed Supreme Court Secretariat Ordinance would have transferred the appointment, transfer, and promotion of lower court judges from the Ministry of Law to the Supreme Court itself. It is gone. This means the judges most likely to encounter preliminary security force cases are directly dependent on the executive for their careers.
Third: the Human Rights Commission cannot independently investigate the security forces. The lapsed NHRC ordinance would have given the commission the power to investigate security forces on its own motion, without asking anyone’s permission. The Law Ministry’s objection was that “the commission is not under any ministry, which is an inconsistency.” This objection deserves a direct reply: a Human Rights Commission that is under a ministry is not a Human Rights Commission. It is a ministry sub-department with a logo. Barrister Zamir’s dissent states it plainly: “A watchdog that must ask the permission of the institution it is watching before it can investigate is not a watchdog. It is a shield for that institution.”
Fourth: there is no independent prosecution. The Attorney General’s office operates under the executive. Prosecutors who challenge the security forces face career consequences. Prosecutors who follow government preferences do not.
Fifth: the Home Ministry has formally declared that security force investigations are “sensitive.” The ministry’s own memorandum on the enforced disappearance ordinance states: “Enforced disappearance is a sensitive crime. Its investigation involves the security forces and law enforcement agencies.” Note what this sentence does. It does not say the security forces were wrongly accused. It says the involvement of the security forces makes the matter too sensitive for independent investigation. This is the Awami League’s logic, word for word, transplanted into a BNP Home Ministry memo. The current Home Minister, Salahuddin Ahmed, is himself a survivor of enforced disappearance. His ministry has produced an argument that his own disappearance was too sensitive to independently investigate.
This is not a political criticism dressed in philosophical language. This is the structure of the state, laid bare.
V. Necropolitics and the RAB Death Machine
On 13 January 2026, Bangladesh’s Commission of Inquiry on Enforced Disappearances submitted its final 229-page report, titled “Unfolding the Truth: A Structural Diagnosis of Enforced Disappearance in Bangladesh.” Its findings are worth sitting with, because they are not allegations. They are documented findings of a government-mandated commission.
Between 4,000 and 6,000 people were forcibly disappeared under the Hasina government. The commission found that disappearances were “coordinated, institutional and politically driven,” relying on “a hidden detention infrastructure, coordinated security forces, routine torture, and a compliant justice system.” The Rapid Action Battalion was responsible for approximately 25 percent of disappearances. The police for 23 percent. The DGFI, Detective Branch, and CTTC for substantial portions of the remainder. These were not isolated officers acting on individual impulse. They constituted what the commission called “a core instrument of governance.”
The Cameroonian philosopher Achille Mbembe argues that in the postcolonial state, sovereign power operates not through the management of life (who gets healthcare, who gets education) but through the management of death: who can be killed, by whom, with what impunity. He calls this necropolitics. The ultimate expression of sovereignty is the power and capacity to dictate who may live and who must die. This is not metaphor. In Bangladesh, it was operational procedure.
The RAB’s “crossfire” methodology, documented by Netra News, Deutsche Welle, Human Rights Watch, and now the commission itself, was a standardized killing protocol. A target was identified. He was abducted at night. He was taken to a RAB facility. Hours or days later, he was found dead, with a gun placed nearby, in an “encounter.” RAB officers, according to the Netra News investigation, received what amounted to kill points, career incentives for each confirmed kill. This was not a rogue unit. It was a bureaucratically organized death machine, with standardized procedures, institutional incentives, and a protection system that ran from the Directorate General of Forces Intelligence to the law courts that rubber-stamped the “encounter” verdicts.
Giorgio Agamben’s concept of the homo sacer, the person who can be killed but whose death cannot be processed through law, describes the disappeared of Bangladesh with uncomfortable precision. They were taken from their families, held in facilities known as Aynaghar (House of Mirrors, named for the disorientation of the detainee who cannot locate himself), tortured, sometimes released, sometimes killed. Their detention was never officially acknowledged. Their deaths, where they occurred, were attributed to crossfire or natural causes. They existed in a zone that was neither legally inside nor legally outside the state: they were citizens whom the state could kill with impunity, and that is, historically, the most accurate definition of bare life.
The crucial point — the one that connects the Aynaghar to the DFP’s formatted hard disk to the lapsed ordinances — is this: the RAB was not created by the Awami League. It was created in 2004, under a BNP government, as a “crossfire” machine from day one. The BNP invented the methodology. The Awami League intensified and politicized it. The difference between BNP-era RAB and AL-era RAB is a difference of target, not of structure. The structure was and is bipartisan.
This is why the BNP’s reluctance to dismantle security force impunity is not hypocrisy, though it has hypocritical consequences. It is something more structurally honest: the BNP cannot dismantle the RAB’s culture of impunity without dismantling the institutional logic that the BNP itself installed. To prosecute the RAB officers who disappeared BNP members under the Awami League, the BNP would also have to face questions about what RAB did under its own watch. The institution is too deep in the state’s foundations to pull out without something crumbling.
VI. The Comparative Mirror: Peru, South Africa, and Why Bangladesh Cannot Follow
The question of whether a state can try its own security forces for crimes against humanity is not abstract. Other countries have tried, with varying success and instructive failures.
Argentina comes closest to success. After the 1976-1983 military junta disappeared an estimated 30,000 people, the return to democracy produced prosecutions that eventually, after pardons and reversals, convicted junta leaders for crimes against humanity. The Mothers of the Plaza de Mayo, who had marched with photographs of the disappeared for decades, became one of the most powerful civil society movements in modern history. What made Argentina’s partial success possible? The military was discredited beyond recovery by the catastrophic Malvinas/Falklands defeat of 1982. Civil society organizations maintained sustained, co-optable pressure for decades. Independent prosecutors existed who were genuinely willing to pursue cases against senior officers. None of these conditions currently exist in Bangladesh. The military has not experienced a comparable discrediting event. Civil society is being co-opted. There is no independent prosecution infrastructure.
Peru achieved a remarkable landmark: the 2009 conviction of Alberto Fujimori himself, a head of state, for crimes against humanity committed by security forces under his command. But Peru’s story has a bitter 2025 chapter. In June 2025, Peru’s Congress passed an amnesty law covering security forces for the 1980-2000 conflict. In August 2025, the President announced implementation regardless of an Inter-American Court order to the contrary. Three decades of accountability work, undone by ordinary legislation. If this could happen in Peru, where accountability went furthest, Bangladesh’s prospects deserve honest assessment.
South Africa’s Truth and Reconciliation Commission is frequently cited as a model. But its most important lesson for Bangladesh is the one least discussed: the senior leadership of the apartheid security forces did not participate. The top commanders, the architects of systematic murder, evaded the process. Only the foot soldiers bore the cost. Even Eugene de Kock, “Prime Evil,” was a middle-ranking official. The generals kept their pensions and their silence. This is the structural pattern: when institutions are complicit in atrocity, accountability tends to stop at the level where institutional survival interests kick in.
The regional picture deepens the pessimism. Pakistan’s ISI model, security forces as a state within a state, effectively immune from civilian judicial oversight, is not alien to Bangladesh’s security architecture. It is its genealogy. After 1971, Bangladesh’s intelligence and paramilitary structures inherited colonial and Pakistani templates of security force sovereignty. India’s Armed Forces Special Powers Act zones, areas where security personnel are immune from prosecution without central government sanction, are the explicit legal model for what the BNP Home Ministry is asking for when it demands prior government approval for security force investigations. These are not separate national phenomena. They are a shared South Asian architecture of competitive authoritarianism, as Lucan Ahmad Way has called it: maintaining the forms of democracy while systematically insulating security power from democratic accountability.
VII. Derrida’s Impossible Demand: The Question the State Cannot Ask Itself
We have arrived at the essay’s hardest question. Can the Bangladeshi state try the military and intelligence officers who carried out thousands of enforced disappearances, extrajudicial killings, and acts of systematic torture?
Jacques Derrida argued that justice and law are permanently in tension. Law is calculable: it applies rules to cases. Justice is incalculable: it is the infinite demand that the law always be held accountable to something beyond itself. The moment a judge applies the law, Derrida says, she faces what he calls an aporia: a demand that exceeds what the law can provide. Justice is what law must always be reaching toward but never fully grasps. It is the impossible demand that makes the demand worth making.
Here is the aporia for Bangladesh in plain terms. The prosecution of security force officers for crimes against humanity would require the state to use its legal machinery against the institutions that make its legal machinery possible. The courts that would try them are appointed by the executive that commands them. The prosecutors who would prosecute them serve the government that benefits from their impunity. The security forces whose records would be used as evidence are the same forces that format hard disks when evidence becomes inconvenient. This is not a circle that can be squared from within the existing constitutional order. It is a structural contradiction.
Derrida’s point is not that we should stop trying. It is the opposite. Justice is possible only as the infinite, unrelinquishable demand that the legal order be held to account for what it excludes and protects. The demand that the state try its own security forces is exactly this kind of demand: it asks the law to confront its own founding violence, the force on which its authority rests and which it cannot fully acknowledge without undermining itself.
Every state is founded on what Derrida calls a mystical foundation of authority: an original act of power that precedes law, that is itself neither legal nor illegal because legality had not yet been established. Bangladesh’s founding violence includes 1971, the coups of 1975 and 1982, Operation Clean Heart in 2002, and the RAB from 2004 onwards. The July Uprising was an attempt to interrupt this founding violence, to subject it retrospectively to the rule of law. The lapsed ordinances were the instruments of that interruption. Their lapsing is the founding violence reasserting itself: the sovereign deciding, once again, that there are zones of state action that law does not reach.
VIII. The Pessimism of Clarity: Why Naming Matters
The argument of this essay is structurally pessimistic. Let us be clear about what kind of pessimism this is.
It is not the pessimism of the government apologist who says: everything is fine, the ordinances will come back as bills, wait and see. That is not pessimism. That is the political utility of patience deployed in the service of inaction.
It is not the pessimism of the outrage-without-analysis that fills social media with furious memes and zero institutional understanding. That is not pessimism either. That is rage performing itself.
It is the pessimism that Romain Rolland calls “pessimism of the intellect, optimism of the will,” that begins with a clear-eyed account of structural constraints as the precondition for any real strategy of change. You cannot fix a plumbing problem you have misdescribed. You cannot resist a restoration you have not identified.
So let us name what is happening.
The lapsing of the enforced disappearance ordinance is a decision to protect the institutional structures that carried out disappearances. It is not a procedural delay.
The deletion of the DFP CCTV footage goes beyond evidence destruction. It is the demonstration, in miniature, on a government hard disk, that the mechanisms of impunity are intact and operational under the new government.
The co-optation of human rights organisations, the marginalisation of independent civil society, the beating of a photographer whose archive contains the faces of the disappeared — these are not isolated incidents of post-revolutionary disorder. They are the systematic elimination of the infrastructure of accountability.
Badiou says that every genuine political Event produces what he calls “faithful subjects”: people who maintain fidelity to what the Event made possible, against the forces of restoration. The faithful subjects of the July Uprising are identifiable. They are the human rights workers who continue to document disappearances despite the new government’s indifference. They are the photographers, Johan among them, whose archives preserve the faces of the missing even when the state prefers they be forgotten. They are the parliamentarians like Barrister Nowshad Zamir who record their dissent knowing it will be outvoted, because the record matters. They are the families of the 330 people still missing, who have not stopped asking where their sons and daughters are.
These people are keeping the Event alive. They are doing so at personal cost. The least the rest of us can do is name, with precision, what is happening around them.
IX. Conclusion: The Sovereignty Question and Its Answer
Can the Bangladeshi state try the military and intelligence officers who carried out enforced disappearances, extrajudicial killings, and systematic torture?
The answer, as of April 2026, is no.
Not because the evidence does not exist. The Commission of Inquiry produced a 229-page report documenting coordinated institutional crime at the highest levels of the security apparatus.
Not because there is no legal framework. The International Crimes (Tribunals) Act could, in principle, be used for prosecutions of security force members.
Not because there are no brave lawyers, honest judges, or committed human rights workers. Bangladesh has all of these, though their institutional space is shrinking.
But because the constitutional architecture that would support such prosecution has been systematically dismantled in the weeks since the 13th parliament sat. Judicial independence has been re-confirmed as executive prerogative. The human rights commission has been prevented from becoming genuinely independent. The enforced disappearance law has been allowed to lapse. The security forces retain the structural immunity of institutions that the state, as currently constituted, cannot function without.
Schmitt tells us: the sovereign decides the exception. The BNP government has decided. The security forces are the exception.
Derrida tells us: law rests on a founding violence it cannot fully acknowledge. Bangladesh’s founding violence is still operational, still producing bodies, still formatting hard disks.
Agamben tells us: the state of exception, once installed, tends to become the permanent paradigm of government. Bangladesh’s security forces have been in a permanent state of exception since at least 2004.
Mbembe tells us: necropower is not a deviation from sovereignty. It is its purest expression. The RAB’s kill protocols were not a failure of the state. They were a feature of it.
And Badiou tells us: restorations succeed when the political formation born of an Event becomes the manager of that Event’s consequences instead of its faithful executor. The BNP government is managing the July Uprising’s consequences. It is not serving its mandate.
This is the hard answer to the sovereignty question. The state cannot try itself because it is constituted, in part, by the practices that would need to be tried. To change this requires not a better ordinance or a cleverer parliamentary maneuver. It requires what Derrida calls a new founding act, a genuine reconstitution of the relationship between state power and the rule of law. It requires what Argentina’s accountability history required: a discrediting of the security apparatus so complete that institutional self-protection becomes impossible. It requires sustained civil society pressure that cannot be co-opted because it is too distributed, too numerous, too embedded in the families of the disappeared. It requires international mechanisms with real enforcement power, which Bangladesh currently lacks.
None of this is available in April 2026.
But here is what Derrida’s aporia insists: the impossibility of justice does not release us from its demand. The experience of the impossible is not the reason to stop. It is the reason the demand is worth making, because it is the only demand that is equal to the scale of what happened.
The families of the 330 people still missing are making exactly this demand. They are asking the impossible from a state that has demonstrated it cannot provide it. They have been asking for years. They were at Tarique Rahman’s exhibition, faces still waiting, before the election. They are still waiting now.
The room with no CCTV is not a metaphor.
It is a method.
And the method works only if we let it.