Counting the dead of 1971 and a new book reopening Bangladesh’s most sensitive historical question

For more than half a century, one particular number has occupied a powerful place in Bangladesh’s account of its birth…three million.

It is the official figure for those killed during the 1971 Liberation War, repeated in state documents, textbooks, speeches and commemorations. Over time, it has become part of the vocabulary through which Bangladesh explains itself; the scale of its suffering, the legitimacy of its independence and the brutality of the Pakistani military campaign that preceded it.

Yet there has never been a comprehensive national enumeration of those three million dead. That gap between commemoration and documentation is the subject of a substantial new book that approaches one of Bangladesh’s most politically sensitive historical questions from an unusually prosaic direction.

Bangladesh Shaheed 1971: The First Comprehensive Compilation of Liberation War Martyrs from Secondary Sources, edited by Dr Tarek Fazal and Dr Md. Habibullah and published by Riddho Prokashon at the 2026 Ekushey Book Fair, runs to 1,029 pages. 

Its compilers examined 102 district-level books, covering all 64 districts, and attempted to determine how many wartime deaths could actually be identified in the literature already produced about 1971.

Their findings do not establish a new national death toll, and the editors explicitly say that they do not. But they expose the weakness of the documentary foundation beneath some of the numbers that have circulated for decades.

The books they examined yielded references to 46,135 martyrs. They found another 213,043 deaths presented through estimates or what they call “notional” calculations. But when they narrowed the test to people documented with names and addresses, the number fell to 30,755.

The significance of the exercise lies less in any of those numbers than in what produced them. More than five decades after the war, Bangladesh still lacks a comprehensive state-led effort to identify its wartime dead. That omission has become harder to separate from the politics surrounding 1971.

A national number without a national count

There is little dispute about the fundamental character of 1971. The Pakistani military launched Operation Searchlight on March 25, beginning a campaign of mass violence against Bengalis. Civilians were killed, villages destroyed and intellectuals targeted. Millions fled to India. The war ended in December with India’s direct intervention, Pakistan’s defeat and Bangladesh’s independence.

Bangladesh’s official estimate of three million became established immediately after independence. It was repeated by the new state and eventually embedded deeply in national memory. But the kind of nationwide documentation exercise that might have transformed an estimate into an independently verifiable record never took place.

Wars rarely produce perfect casualty figures. Records disappear, families are displaced, bodies remain unidentified and mass graves are difficult to investigate. Estimates are unavoidable, particularly immediately after a conflict. But estimates can subsequently be tested.

Bangladesh had decades in which to establish district-by-district registries, collect testimony, examine burial sites, compare population records, reconcile duplicate names and build a publicly accessible database of the dead. No government completed such an exercise.

That absence has periodically attracted researchers who challenged the established figures, most controversially author Sarmila Bose.

Her 2011 book, Dead Reckoning: Memories of the 1971 Bangladesh War, argued that scholarship on the conflict suffered from a basic methodological weakness. Sarmila Bose described 1971 as a conflict that had been extensively discussed but inadequately documented, and argued that exaggeration, selective memory and politically useful claims had entered the historical record.

Her work provoked intense criticism, including challenges to her methodology and interpretation. But one question raised by the controversy proved more difficult to dismiss: Why had Bangladesh itself never produced a systematic accounting of the people killed?

Bose argued that what she called “victor’s history” had hardened around the conflict. She also criticized Pakistani nationalist accounts, which developed their own explanations of 1971 around Indian conspiracies and minimized or disputed Pakistani military atrocities.

The competing narratives differed dramatically, but they shared a weakness: politics repeatedly got ahead of documentation. Bangladesh Shaheed 1971 enters this argument differently.

Why is the book different?

Its editors do not claim to have discovered the real number of people killed. They have not conducted excavations, demographic reconstruction or nationwide interviews. Their project is explicitly based on secondary sources.

Instead, they ask a narrower question: What do Bangladesh’s existing district histories actually document? That approach makes the book simultaneously less ambitious and potentially more useful.

The compilers surveyed 102 books from across the country. They found an extraordinary lack of uniformity. Researchers in one district might provide names and addresses; another might describe “hundreds” killed; another might offer an estimate without explaining how it was reached.

There was no common methodology because there had never been a coordinated national project. The editors therefore divided the information into categories. One contained people specifically identified by name and address. Another recorded numerical claims made by source authors. A third contained estimates or “notional” numbers.

If a source said between 10 and 15 people had been killed, for example, the compilers used 15 — the upper figure — in calculating the estimate-based total. They also confronted duplication. The same person could appear in several books, sometimes with variations in spelling or identifying details. 

Without careful reconciliation, combining local histories could produce the appearance of several victims where there was only one. More striking were some of the methods used to produce large casualty estimates.

The editors single out a 44-volume series surveying genocide sites, killing fields and mass graves. They say those volumes contained an especially high concentration of estimates and potentially misleading calculations.

In one example described by the editors, an author assumed that killings occurred every day over a period stretching from April 19 to November 29, 1971 and multiplied an assumed daily casualty rate across that period. That produces a number. It does not produce a count. The distinction is central to the new book.

What 30,755 names do and do not tell us

The headline figures from the compilation are certain to attract political attention: 46,135 martyrs mentioned in the surveyed books; 213,043 deaths generated from estimate-based claims; and 30,755 people whom the editors say they could document by name and address.

None is presented as the final death toll of the Liberation War. That caveat is essential.

A person absent from the 102 books examined by the researchers did not therefore survive the war. A family may never have reported a death. A village massacre may have left few witnesses. Records may have disappeared. Some districts have been researched far more extensively than others.

The book is a compilation of available secondary material, not a census of 1971. Its 30,755 named victims therefore cannot responsibly be converted into an alternative national casualty figure.

But neither can the enormous difference between documented names and frequently repeated estimates simply be ignored. The book reveals how much remains unknown.

That is particularly important because Bangladesh has often treated the debate as though only two positions were available: accept three million as an established historical fact or deny the atrocities of 1971.

There is a substantial analytical space between those positions.

It is possible to accept the overwhelming evidence that Pakistani forces and their local collaborators committed mass atrocities while also asking how many people were killed. Indeed, documenting victims individually can strengthen the historical record rather than diminish it.

The editors introduce another complication by examining the meaning of “shaheed,” or martyr. They distinguish among people killed fighting for Bangladesh’s independence, those killed while supporting a united Pakistan, and those who belonged to neither category but died because of the war.

That classification challenges the tendency to turn every wartime death into a single political category. A civilian murdered in a village, a freedom fighter killed in combat, a person killed while supporting Pakistan and someone who died amid wartime violence were all victims of the same historical conflict. They were not necessarily participants in the same political project.

This matters because “martyr” in Bangladesh is not merely descriptive. It carries moral, political and sometimes administrative significance. The state’s own records demonstrate the difficulty.

The government maintains a gazetted list of martyred freedom fighters whose families qualify for state recognition and benefits. According to a January 26, 2026 report by The Daily Star, citing the Muktijoddha Kalyan Trust, the number of gazetted martyred freedom fighters stood at 6,757.

That figure cannot be compared directly with the three-million estimate. It covers a specific category — recognized combatants who died — rather than all civilian and military deaths. But it illustrates the broader problem.

A problem of its own 

Bangladesh has managed to identify particular categories of war dead for administrative purposes while never constructing a comprehensive national register of everyone killed during the conflict. The result is an unusual historical imbalance. The country has an exceptionally precise official national figure but no correspondingly precise list behind it.

That should be a question for historians before it becomes an argument for politicians. There is also a danger in moving too quickly in the opposite direction.

Bose estimated that 50,000 to 100,000 people may have died in the conflict. Her estimate has itself been disputed. Replacing three million with another number unsupported by a comprehensive enumeration would merely reproduce the problem in a different form.

The strongest conclusion available from Bangladesh Shaheed 1971 is therefore not that Bangladesh’s official figure has been disproved. It is that it has never been properly tested. So, the dispute is ultimately about more than numbers.

Since independence, competing political forces have repeatedly used 1971 to establish legitimacy. The Awami League possessed the strongest historical claim to leadership of the independence movement and made that inheritance increasingly central to its politics under Hasina.

The more tightly the party attached itself to the Liberation War, however, the easier it became for opponents to blur the distinction between criticism of the Awami League and criticism of the historical narrative it promoted.

That was damaging in both directions.

Critics of Hasina could be portrayed as hostile to 1971. At the same time, resentment toward her government could migrate toward the historical symbols the government claimed to defend. The aftermath of the August 2024 revolution demonstrated the cost.

Bangladesh now has an opportunity to separate those questions: the historical significance of the Liberation War from the political interests of any government that invokes it.

Doing so requires greater scrutiny, not less. The casualty question provides an obvious starting point. A serious national project would not begin by deciding whether three million is correct or incorrect. It would begin with the dead.

It could collect names from families, local archives, newspapers, freedom-fighter records, refugee documentation and existing district histories. It could digitize the information, identify duplicates and publish its methodology. Mass graves and killing sites could be catalogued. Researchers could distinguish confirmed identities from probable victims and demographic estimates.

Such a project might never produce an exact number. But it would produce something Bangladesh does not currently possess: a transparent evidentiary basis for discussing the scale of its defining conflict.

—

The writer is a London-based researcher in history, political analyst and activist. He can be reached at mnhasan.blog@gmail.com.