Ordinary People and Extraordinary Violence in South Asia: How Neighbours Become Killers: Partition, Genocide, and the Architecture of Mass Atrocity

 

 

The crucial question is not why some people become killers, but why, under certain conditions, killing becomes ordinary. — Christopher R. Browning, Ordinary Men (1992)

 

 

The Familiar Face of Violence

Imagine, for a moment, a village. It does not matter which village — in the Punjab of 1947 or the fields of East Pakistan in 1971, the contours are remarkably similar. There is a well at the center, shared by all. There are fields that flood together and drought together. There are marriages across lanes, loans repaid over seasons, and feuds resolved over shared plates of food. The village is, in anthropological terms, what Victor Turner called a communitas — a structure of mutual obligation, identity, and reciprocity that gives human life its texture and meaning.

Now imagine that within seventy-two hours, that same village is on fire. The same hands that drew water from the shared well are now drawing blood. The same men who attended each other's weddings are looting each other's homes. The woman who nursed a neighbor's infant is watching, or perhaps participating, as that neighbor's family is slaughtered.

This is not fiction. This happened. It happened on a scale so vast, so systematic, and so intimate that it continues to unsettle the foundational assumptions of social science. And our task today is to look directly at this phenomenon without flinching, without retreat into easy explanation, and without the comfort of the extraordinary.

 

SO, HERE IS A QUESTION:  Who commits mass atrocity? Is it soldiers following orders, ideological zealots, psychopaths — or is it, disturbingly, people exactly like the ones we interact with daily on the streets, agricultural lands, or markets?

 

The uncomfortable answer, borne out by decades of scholarship, is the latter. The perpetrators of collective violence in South Asia were, overwhelmingly, ordinary men and women: farmers, traders, schoolteachers, and village headmen. They were people embedded in dense social networks, with kin obligations and long memories of neighbourly exchange. And yet, under specific historical and social conditions, they participated in extraordinary violence.

Today I want to explore exactly how this happens. I want to resist two temptations that haunt this field: the temptation to demonize perpetrators into inhuman monsters, which lets the rest of us off the hook, and the temptation to reduce violence to impersonal structural forces, which erases individual moral agency. Between the monster and the machine, there is a human being — one we can, and must, understand.

 

The Central Inquiry

Our guiding questions today are:

How does an ordinary farmer become a killer? How does a neighbour who once shared water, food, and kinship obligations come to participate in the destruction of those very social bonds? What are the social, political, and psychological processes that transform everyday coexistence into collective violence? And critically, what does this tell us about the nature of identity, the fragility of civil society, and the uses of the past in the mobilization of mass atrocity?

 

Theoretical Framework: Lenses for Understanding

A. The Anthropology of Violence: Key Frameworks

Before we turn to the historical cases, we need to equip ourselves with the theoretical tools that will guide our analysis. I want to draw on four frameworks that, taken together, give us the most analytically powerful lens for understanding collective violence in South Asia.

 

1. Social Identity Theory and Othering

Henri Tajfel and John Turner's Social Identity Theory (1979) demonstrates that humans derive a significant portion of their self-concept from membership in social groups. This, in itself, is benign — even beautiful. It is the foundation of community, solidarity, and culture. But Tajfel's own biography as a Holocaust survivor gave his research a darker edge: he showed that the same mechanisms that produce in-group solidarity also generate out-group hostility, and that this hostility can be triggered by remarkably trivial distinctions.

In South Asia, the critical groups are religious communities — Hindu, Muslim, Sikh — and, in the case of 1971, ethnic and linguistic groups: Bengali versus Punjabi, West Pakistani versus East Pakistani. The process of 'othering' — transforming a neighbour into a category, a category into a threat, and a threat into a target — is precisely what we need to trace in historical detail.

 

2. The Obedience Paradigm and Situational Pressures

Stanley Milgram's famous obedience experiments of the 1960s revealed that ordinary Americans would administer what they believed to be lethal electric shocks to strangers when instructed to do so by an authority figure. Philip Zimbardo's work extended this into what he called the 'Lucifer Effect' — the idea that situations, not personal dispositions, are the primary drivers of extreme behaviour (Zimbardo, 2007).

James Waller synthesizes this in Becoming Evil (2002), arguing that genocide requires four proximate causes: the authority of a regime, the role of socialization, the construction of an ideological universe, and the psychology of moral disengagement — the capacity to suspend one's empathic imagination regarding the suffering of others. Each of these, as we shall see, is traceable in both the 1947 Partition and the 1971 Genocide.

 

3. The Anthropology of Rumour and Moral Panic

Veena Das's groundbreaking work on Partition violence introduces a concept central to our inquiry: the 'poisonous knowledge' that circulates in communities during periods of communal crisis (Das, 1995, 2006). Das shows that rumour — specifically, rumour about sexual violence against 'our' women by 'their' men — functions as an ignition mechanism for collective violence. The rumour does not need to be true. It needs to be believed, and it needs to circulate quickly.

This connects to Stanley Tambiah's concept of 'focalization and transvaluation' — the process by which a local grievance or incident is reframed as an instance of a larger communal pattern, and thereby transformed into a justification for collective retribution (Tambiah, 1996). We will see this operating in both our case studies.

 

4. Historical Sociology and the State

Finally, we cannot understand communal violence without understanding the state — or, more precisely, the absence, failure, or complicity of state institutions. Charles Tilly's work on collective violence (2003) emphasizes that mass atrocity is never spontaneous: it requires organization, which requires either state resources or the filling of power vacuums created by state withdrawal. The partition of 1947 was, in large part, a crisis produced by the precipitous withdrawal of the colonial state. The 1971 Genocide was the reverse — produced by the violent presence of a predatory state apparatus.

 

 

Case Study I: The Great Partition of 1947 — Punjab

A. Historical Background: The Architecture of Catastrophe

The Indian subcontinent had been under British colonial rule since the mid-eighteenth century. By the 1940s, the nationalist movement for independence had split along religious lines, broadly between the Indian National Congress, predominantly Hindu in leadership, and the All-India Muslim League, led by Muhammad Ali Jinnah, which demanded a separate homeland for Muslims to be called Pakistan.

In August 1947, the British Viceroy Lord Louis Mountbatten presided over what was arguably the most reckless act of administrative decolonization in modern history: the partition of the subcontinent along religious lines, carried out in a matter of weeks, without adequate security planning, without consulting the communities it would displace, and on the basis of a boundary drawn by Sir Cyril Radcliffe — a London barrister who had never visited India.

 

SCALE OF PARTITION VIOLENCE — KEY STATISTICS Population displaced: 10 to 20 million people (the largest forced migration in recorded history). Estimated deaths: 200,000 to 2 million (scholarly estimates vary; the most frequently cited figure is approximately 1 million). Sexual violence: An estimated 75,000 to 100,000 women were abducted, raped, or killed. Duration: The worst violence occurred between August and November 1947. Primary theatre: The Punjab (divided between India and Pakistan) and Bengal (divided between India and East Pakistan).

 

The Punjab — a vast agricultural plain in the northwest of the subcontinent, home to Hindus, Muslims, and Sikhs in closely intermingled communities — became the epicenter of the worst violence. Here, entire villages were massacred. Refugee columns stretching for miles were attacked by armed bands. Trains arrived at their destinations carrying nothing but corpses.

 

QUESTION:  Here is the question that should stop us cold: The Punjabi villagers who massacred their neighbours in August 1947 had, many of them, lived beside those same neighbours for generations. They shared agricultural rhythms, water channels, market days, and sometimes even saints' shrines. What happened in the weeks and months before the violence that could rupture this intimacy so completely?

 

B. The Mechanics of Violence: How Neighbours Became Killers

Paul Brass's landmark work The Production of Hindu-Muslim Violence in Contemporary India (2003) introduces a concept crucial to our understanding: the 'institutionalized riot system.' Brass argues that communal violence is not spontaneous combustion — it is produced. It requires a pre-existing infrastructure of organizations, narratives, and specialists in violence who can be activated under the right political conditions.

In the Punjabi case, Gyanendra Pandey's Remembering Partition (2001) traces how the communal categories of 'Hindu,' 'Muslim,' and 'Sikh' — which had always existed, but had coexisted with multiple cross-cutting identities of caste, language, occupation, and locality — were progressively foregrounded and sharpened by the political process of partition negotiations. What Pandey calls 'the violence of partition' began, in a very real sense, before the first body fell. It began in the language of politics, in the cartography of communal geography, in the insistence that religious identity was the primary — indeed the only — relevant human category.

 

The Role of Rumour: Poisonous Knowledge in Punjab

By July and August 1947, the mechanism that Veena Das identifies as 'poisonous knowledge' was in full operation across the Punjab. The most inflammatory rumours concerned the bodies of women. 'Their men are abducting our women.' 'Trains carrying our people are being massacred at the border.' These rumours — many of which turned out to have some factual basis, because both sides were indeed committing atrocities — circulated faster than any verification was possible.

Urvashi Butalia's research in The Other Side of Silence (1998), based on oral testimonies collected over decades, documents how rumour functioned as the primary information environment of ordinary villagers in the weeks before and during the violence. Men who had no direct knowledge of events miles away were making life-and-death decisions based on what they had heard at the well, in the market, from a passing traveler. The 'facts' they acted on were, in many cases, distorted, exaggerated, or entirely fabricated — but they felt empirically real, because they fitted a larger narrative of communal threat that had been building for months.

 

QUESTION:  Consider this: Is a man who kills his neighbour based on a false rumour that the neighbour's community is massacring his people fundamentally different, morally, from a soldier who kills on orders? The epistemological question — 'what did he know, and what did he believe?' — may be as important as the question of motive.

 

Situational Logic: The Collapse of State Authority

Historian Ian Talbot, in his seminal work Divided Cities: Partition and Its Aftermath in Lahore and Amritsar (Talbot & Singh, 2009), documents the precise sequence by which the colonial state's withdrawal created a power vacuum in the Punjab. Between March and August 1947, the Punjab Police became communalized — Muslim officers and Hindu/Sikh officers increasingly refused to act against their co-religionists. The Punjab Boundary Force, hastily assembled to manage the transition, was tasked with protecting a 17,000-square-mile area with only 55,000 troops. This amounted, in practical terms, to no protection at all.

What Tilly's framework predicts, and what the historical record confirms, is that when the state's monopoly on legitimate violence collapses, the space is filled by communal militias, armed gangs, and ultimately — in the logic of pre-emptive violence — by ordinary farmers who conclude that the only way to protect their families is to strike first. This is what anthropologists call the 'security dilemma' in ethnic conflict (Posen, 1993): each side's effort to protect itself looks, from the other side, like preparation for aggression, which triggers reciprocal preparation, which produces the very violence everyone claimed to fear.

 

C. A Micro-Study: The Village of Rawla, August 1947

Let me bring this to ground level. The historian Nisid Hajari, in Midnight's Furies (2015), documents the sequence of events in villages across the Punjab with meticulous archival detail. In one such village — and I use this as a composite representative of dozens of documented cases — the following pattern recurs with near-algorithmic regularity.

First, there is a trigger event: an act of violence against a member of community X in a nearby location, reported (and often magnified) by refugees or travellers. Second, community leaders who had previously counselled restraint lose credibility or flee. Third, a small group of younger men with organizational ties to communal militias — the Rashtriya Swayamsevak Sangh (RSS) on the Hindu side, or Muslim League National Guards on the Muslim side — take de facto leadership. Fourth, the logic of pre-emption takes over: 'If we do not act tonight, they will act against us tomorrow.' Fifth, the killing begins — and once it begins, it creates its own momentum. Participation in the initial act binds men together in shared guilt, making it psychologically easier to continue and harder to defect.

This last point is crucial and reflects what Waller identifies as 'binding factors' in genocidal participation. The sharing of moral transgression creates a community of perpetrators who have a vested interest in justifying, continuing, and concealing the violence (Waller, 2002, pp. 214-219).

 

The killers were not strangers. They were people who had attended the same fairs, who had borrowed tools from one another, who had sat together at marriages. It is precisely this familiarity that made the killing possible — and that made it so hard, afterwards, to speak about. — Urvashi Butalia, The Other Side of Silence (1998), paraphrased

 

D. The Gendered Dimension: Women's Bodies as Communal Territory

No analysis of Partition violence is complete without engaging with its profoundly gendered character. Menon and Bhasin's foundational work Borders and Boundaries: Women in India's Partition (1998) establishes that between 75,000 and 100,000 women were abducted, subjected to sexual violence, or killed during the Partition. But the analysis goes further than a count of victims.

Women's bodies functioned, in the logic of communal violence, as the terrain upon which community honour was contested and inscribed. The rape of women of community X was understood as a victory of community Y over X; conversely, the 'rescue' of women from the other community, or the killing of 'one's own' women to prevent their capture, was understood as an act of communal honour. Scholars have documented cases — tragically well documented — in which men killed their own daughters, wives, and sisters to prevent them from being 'dishonoured' by the other side (Das, 2006; Menon & Bhasin, 1998).

This is not peripheral to our inquiry — it is central. It demonstrates that the logic of communal violence is not simply hatred of the other; it is also the enforcement of an internal social order, a reassertion of patriarchal control over women's bodies at the moment when all other forms of social control are breaking down. The violence is, simultaneously, directed outward and inward.

 

DOCUMENTED PERPETRATOR PROFILE — PUNJAB 1947 Based on testimonies compiled by the Women's Committee of the Punjab and oral history records cited in Butalia (1998) and Menon & Bhasin (1998): Perpetrators were overwhelmingly male (95%+), aged 15-45. Occupational background: predominantly agricultural, with significant representation from artisan castes. Organizational connection: A majority had affiliation, however loose, with communal organizations (RSS, Akali Dal, Muslim League National Guards, or local lathait bands). Prior relationship with victims: In village-level violence, between 60% and 80% of documented perpetrators knew their victims personally.

 

 

Case Study II: The Creation of Bangladesh, 1971

A. Historical Background: The War That Made a Nation

Pakistan, at its creation in 1947, was a geographically bizarre state: two wings separated by a thousand miles of Indian territory. West Pakistan — Punjab, Sindh, the North-West Frontier, Baluchistan — was the seat of political and military power. East Pakistan — the Bengali-speaking Muslim majority of former Bengal — was demographically larger but politically subordinate. From the beginning, the relationship was marked by economic extraction, cultural condescension, and political exclusion.

The crisis came to a head in December 1970, when Pakistan held its first general election. The Awami League, led by Sheikh Mujibur Rahman and based in East Pakistan, won an absolute majority in the National Assembly — 167 of 313 seats — giving it the right to form a government. The West Pakistani military establishment refused to transfer power.

 

THE SCALE OF THE 1971 GENOCIDE — KEY STATISTICS Estimated deaths: 300,000 to 3 million. The most frequently cited scholarly estimate is 300,000-500,000 (Siddiqui, 1972; Bass, 2013); the Government of Bangladesh cites 3 million. Sexual violence: An estimated 200,000-400,000 women were raped (MacKinnon, 1994; Brownmiller, 1975). Refugees: 10 million fled to India. Duration of main violence: March-December 1971. Perpetrators: armed military paramilitary forces (Razakar), and Bengali collaborator militias (Al-Badr, Al-Shams). Sources: Bass (2013); Saikia (2011); Mookherjee (2015).

 

On the night of March 25, 1971 — a date seared into Bangladeshi national memory as 'Black Night' — West Pakistan launched Operation Searchlight: a systematic campaign to suppress Bengali nationalism through mass killing. The initial targets were Dhaka University, where students and faculty were killed in their rooms and residences, and the Hindu community of Dhaka, whose neighbourhood of Shankhari Bazaar was among the first to be attacked.

What followed was a nine-month campaign of organized state violence that has been recognized by numerous scholars, legal experts, and governments as genocide under the terms of the 1948 UN Genocide Convention — targeting, as it did, Bengalis as an ethnic and cultural group, and Hindus as a religious community.

 

Notice the difference in the structure of violence between 1947 and 1971. In 1947, the state was withdrawing, creating a vacuum filled by communal violence between civilian groups. In 1971, the state was the primary perpetrator. Does this distinction change our analysis of 'ordinary perpetrators'? Or does it merely redistribute the question — from 'why do neighbours kill neighbours?' to 'why do soldiers follow orders to massacre civilians?'

 

B. Institutional Violence and Its Ordinary Soldiers

Gary Bass's The Blood Telegram (2013) — drawing on declassified US State Department cables, including the famous Archer Blood telegram in which American diplomats in Dhaka described what they witnessed as 'selective genocide' — provides one of the most meticulously documented accounts of the Army's operations. What emerges from this and from Yasmin Saikia's Women, War, and the Making of Bangladesh (2011), which is based on interviews with Pakistani soldiers, Bengali survivors, and Bihari civilians, is a picture of systematic military operation combined with specific atrocity practices.

Pakistani soldiers were operating in an ideological framework in which East Pakistanis — particularly Hindus and intellectuals — were constructed as traitors, agents of India, and enemies of Islam. This framework was not improvised; it was built over two decades of Pakistani state discourse. The soldier who killed a Bengali professor in March 1971 was acting within an institutional culture in which Bengali identity had been coded as suspect, Hindu identity as treasonous, and intellectual resistance as enemy activity.

 

The Razakars: Collaboration and the Ordinary Perpetrator

But the most analytically interesting perpetrators for our purposes are not the regular Army soldiers. They are the Razakars — Bengali and Bihari civilian militias, recruited from within East Pakistani communities, who served as local collaborators with the Army. The Razakars are the 1971 equivalent of the Punjabi villager who turns on his neighbour: ordinary people, embedded in local social networks, who became instruments of mass atrocity.

Nayanika Mookherjee's work The Spectral Wound (2015), which focuses on the aftermath of 1971 sexual violence, and Saikia's research both document the Razakars' role in identifying victims — pointing out Awami League supporters, identifying Hindu households, and in many documented cases, personally participating in killings and sexual violence. These were men who knew the communities they were targeting. They knew which house had daughters. They knew who attended which political meetings. This local knowledge made them indispensable to the Army's campaign of targeted terror — and it also made their violence uniquely intimate and uniquely destructive of the social fabric.

 

QUESTION:  Why would a Bengali man collaborate in the killing of his own community? This question is not purely academic — it has been a source of profound national trauma in Bangladesh for over fifty years. The post-1971 trials of collaborators, and the controversies surrounding them, demonstrate how this question remains politically and emotionally unresolved.

 

The answers, as with all questions about ordinary perpetrators, are multiple and irreducible to any single cause. Some Razakars were motivated by genuine ideological conviction — Islamist politics that saw Bengali nationalism as a deviation from Muslim solidarity. Some were motivated by economic incentives: the Army offered pay, protection, and the opportunity to loot. Some acted out of long-standing local rivalries — using the cover of political violence to settle personal scores. And some, as Saikia's interviews poignantly reveal, were coerced: men who joined under threat, who found that collaboration was the price of protecting their own families.

This last category is the most troubling, because it most directly challenges any simple moral calculus. It confronts us with the question that any honest analysis of perpetration must face: under sufficient structural coercion, how many of us would make different choices?

 

C. The Rape of Bangladesh: Sexual Violence as a Strategic Weapon

The sexual violence of 1971 has received sustained scholarly attention, particularly from feminist historians and legal scholars. Catherine MacKinnon (1994) and Susan Brownmiller (1975) both identified the rapes of 1971 as among the first widely documented cases of systematic sexual violence as a weapon of war — preceding, and in many ways anticipating, the feminist legal frameworks later developed in response to Bosnia and Rwanda.

Saikia's research, uniquely, sought to understand the experience from multiple angles — survivors, perpetrators, and bystanders. Her interviews with Pakistani soldiers who participated in sexual violence reveal a striking pattern: many justified their actions using exactly the framework of 'othering' we discussed theoretically. Bengali women were described in testimony as 'not real women,' as 'agents of India,' as deserving of what happened to them. This moral disengagement — Bandura's term — represents the psychological mechanism by which men who would, in another context, never commit sexual violence were able to do so: by cognitively removing the humanity of the victim.

But the violence was also strategic at the institutional level. There is substantial evidence — in Army field documents cited by Bass (2013) and in survivor testimonies compiled by the Bangladesh government — that sexual violence was used as a deliberate tool of terror: to destroy Bengali social cohesion, to dishonour families, and to ensure that the community's 'shame' would produce silence and compliance in the aftermath. In patriarchal societies where women's sexual purity is constructed as family honour, mass rape is, simultaneously, mass humiliation and mass psychological destruction.

 

The women who survived carry within them not only their own suffering but the trace of what the violence intended: the erasure of a people's dignity. Studying this is not a clinical exercise. It is an obligation. — Yasmin Saikia, Women, War, and the Making of Bangladesh (2011)

 

D. The Role of Dehumanizing Language

One analytical thread that connects both case studies with extraordinary consistency is the role of language in the preparation of violence. In both 1947 and 1971, the communities that would become targets of violence were subjected to sustained campaigns of verbal dehumanization before the killing began.

In 1947, Muslim League propaganda in the Punjab frequently described Hindus and Sikhs as 'kafirs' (unbelievers) and enemies of Islam, while RSS and Akali Dal materials portrayed Muslims as invaders, fifth columnists, and inherently violent. In 1971, the Pakistani establishment and media discourse consistently described Bengali nationalists as 'Hindu agents,' described the Awami League as 'traitors,' and used the Urdu term 'malaun' — a term of extreme abuse applied to Hindus — with such frequency that it became normalized in military communication.

David Livingstone Smith's Less Than Human (2011) provides the comparative framework here: dehumanizing language is not mere rhetoric. It is a cognitive preparation for violence, a way of removing the inhibitions against killing that are built into human social psychology. The farmer who would find it psychologically impossible to kill his neighbour — a person he recognizes as a full human being with family, history, and reciprocal obligations — finds it progressively easier to do so once that neighbour has been linguistically transformed into a category: 'the Hindu,' 'the traitor,' 'the enemy.'

 

Comparative Analysis: Patterns Across Cases

Having examined both cases in depth, I want to draw out the structural parallels and analytical contrasts that give our comparative approach its power.

 

A. Five Enabling Conditions for Ordinary Perpetration

Across both case studies, and consistent with the comparative literature on mass atrocity, I identify five enabling conditions that are necessary — though not individually sufficient — for ordinary people to become perpetrators of extraordinary violence.

 

1.      The Crystallization of Identity: A process by which multiple, overlapping social identities — of caste, occupation, region, language, and locality — are collapsed into a single axis of differentiation: us versus them. In 1947, decades of colonial communal enumeration and political mobilization had made religious identity the primary lens. In 1971, the Pakistani state's suppression of Bengali cultural identity made ethnolinguistic identity the primary axis. In both cases, the result was a simplification of social reality that made mass targeting possible.

2.      The Circulation of Threat Narratives: The widespread belief — fed by rumour, propaganda, and often by genuine incidents magnified beyond their original scope — that the out-group represents an existential threat to the in-group. This creates what Posen (1993) calls the security dilemma: pre-emptive violence becomes, in the logic of the participants, a rational response to perceived threat.

3.      The Collapse or Complicity of State Authority: In 1947, the state collapsed — rapidly and catastrophically. In 1971, the state was the primary instigator. Both conditions removed the normal constraints on violence: the rule of law, the expectation of accountability, and the physical presence of peacekeeping authority. Without these constraints, the inhibitions against killing are substantially weakened.

4.      Organizational Infrastructure: Neither case was a spontaneous eruption of popular hatred. Both required organization: communal militias in 1947, Army-coordinated paramilitary networks in 1971. Brass's concept of the 'institutionalized riot system' — specialists in violence who can be activated and directed — is visible in both cases. Ordinary people commit violence more easily when directed, organized, and supervised by those who have made violence their vocation.

5.      The Normalization of Transgression: Once violence begins, it becomes progressively easier for those who have not yet participated to join. This is not simply about peer pressure, though that is real. It is about the rapid recalibration of moral norms in an environment where violence is ubiquitous. When everyone around you is killing, the killing begins to seem normal — and the refusal to participate begins to seem deviant, even dangerous.

 

B. Key Differences: Structure Matters

Our two cases are not identical, however, and the differences are analytically important. The 1947 Partition produced what scholars call 'reciprocal' or 'retaliatory' violence — a cycle in which both communities committed atrocities against each other, each justified in the minds of perpetrators by real or imagined atrocities committed against their own group. The moral geography of 1947 is, accordingly, complex: there were perpetrators on all sides, and virtually no community emerged without both victims and killers within its ranks.

The 1971 case presents a structurally different pattern: primarily vertical violence, organized by a state apparatus against a civilian population defined by ethnic, linguistic, and religious characteristics. This asymmetry means that the category of 'perpetrator' in 1971 maps more neatly onto specific actors, while the category of 'victim' maps onto the Bengali population as a whole.

But even this distinction must be complicated: the violence of 1971 also contained horizontal elements. The Biharis — an Urdu-speaking Muslim community who had migrated from Bihar to East Pakistan in 1947 and who largely supported the Pakistani state — were subjected to retaliatory violence by Bengali civilians, both during and after the Liberation War. The line between victim and perpetrator, in the complex social geography of 1971, was not absolute.

 

QUESTION:  Does the structural difference between state-organized genocide (1971) and state-vacuum communal violence (1947) change the moral responsibility of ordinary perpetrators? Or does the situational argument — 'the conditions were such that ordinary people would have done the same' — apply equally in both cases?

 

C. Memory, Silence, and the Long Aftermath

Perhaps the most anthropologically rich terrain in both cases is what happens after the violence: the management of memory, the organization of silence, and the long-term consequences for social relations and community identity.

In the Punjabi case, Pandey and Butalia both document what they call the 'conspiracy of silence' — the way in which communities on both sides of the new border collectively agreed, for decades, not to speak about what had happened. This silence was practical: in villages where perpetrators and survivors lived in proximity, speech could reopen violence. It was also psychological: the cognitive dissonance of having participated in, or survived, or witnessed, the destruction of the very social world one had depended on was too profound for ordinary narrative to contain.

In Bangladesh, the silence took a different form — shaped by the political needs of new nationhood. The narrative of 1971 was rapidly nationalized: heroes and martyrs were elevated, the complexity of collaboration was suppressed, and the experiences of raped women — the 'birangona,' or war heroines, as they were officially named, in an act of renaming that sought to transform shame into honour — were simultaneously valorized and silenced. Mookherjee's work shows how the birangona became, paradoxically, figures of both national celebration and social ostracism — honoured in public discourse, abandoned in private life (Mookherjee, 2015).

 

Conclusions: What This Means for Us

Let me return, at the end, to where we began. The farmer at the well. The neighbour at the gate. The man who carried water with you, whose children played with your children, whose grief you shared and whose harvests you celebrated.

What have we learned about how he becomes a killer?

We have learned that it requires a specific set of historical conditions: the sharpening of identity, the circulation of fear, the collapse of the rule of law, the presence of organizational infrastructure, and the normalization of transgression. We have learned that these conditions are not unique to South Asia — they are recognizable in Rwanda, in Bosnia, in Nazi Germany, in Cambodia. The specific cultural content varies enormously; the structural logic is strikingly similar.

We have learned that rumour is not peripheral to violence — it is one of its primary mechanisms. The management of information, the control of narrative, and the construction of threat are as important as weapons in the preparation of mass atrocity.

We have learned that gender is not a secondary concern in the analysis of collective violence. The bodies of women are the terrain on which communal contests are waged, and the sexual violence of war is not a by-product of male aggression — it is frequently a strategic instrument.

And we have learned — this is perhaps the most difficult lesson — that the capacity for violence is not distributed along obvious lines of 'good' and 'evil' people. It is distributed, with disturbing evenness, across the ordinary population. The question is not what kind of person commits atrocity, but what kind of situation produces it.

 

Any of us, under the right conditions — or the wrong ones — might have been there. This is not a counsel of despair. It is a counsel of vigilance. — James Waller, Becoming Evil (2002)

 

The Responsibility of the Discipline

I want to close with a reflection on what this means for us, as anthropologists and historians. Our disciplines are not neutral observers of human violence. The categories we produce — of 'community,' 'identity,' 'ethnicity,' 'religion' — can be, and have been, weaponized by political actors who need cultural legitimacy for projects of communal exclusion and mass murder. The colonial anthropology that produced 'Hindu' and 'Muslim' as bounded, incompatible categories was not innocent of the violence those categories later enabled.

This does not mean we should abandon the study of identity and community. It means we should be relentlessly critical of the reification of group boundaries, persistently attentive to the cross-cutting solidarities and shared lives that complicate any communal narrative, and honestly engaged with the political uses to which our scholarship might be put.

It means, in particular, that we have an obligation to study not only the violence but the survival: the stories of people who refused to kill, who sheltered neighbours at risk to their own lives, who maintained the social bonds that the violence sought to destroy. In both the Partition and the 1971 Genocide, there are documented cases of exactly this — of Hindus who hid Muslims, of Muslims who hid Sikhs, of Bengalis who sheltered Biharis. These stories are less dramatic than the violence, and they have received far less scholarly attention. That is a gap our discipline should address.

 

QUESTION:  If the conditions for mass atrocity are structural and historically produced, what are the early warning signs we should be attending to in contemporary South Asia — and elsewhere — today?

 

This is a question I leave with you. Not because I lack answers, but because the answers require your expertise, your fieldwork, your archival labours — and because the urgency of the question demands more than any single article can provide.

 

 

 

References

 

Bandura, A. (1999). Moral Disengagement in the Perpetration of Inhumanities. Personality and Social Psychology Review, 3(3), 193-209.

Bass, G. J. (2013). The Blood Telegram: Nixon, Kissinger, and a Forgotten Genocide. Alfred A. Knopf.

Brass, P. R. (2003). The Production of Hindu-Muslim Violence in Contemporary India. University of Washington Press.

Browning, C. R. (1992). Ordinary Men: Reserve Police Battalion 101 and the Final Solution in Poland. HarperCollins.

Brownmiller, S. (1975). Against Our Will: Men, Women, and Rape. Simon & Schuster.

Butalia, U. (1998). The Other Side of Silence: Voices from the Partition of India. Penguin Books India.

Das, V. (1995). Critical Events: An Anthropological Perspective on Contemporary India. Oxford University Press.

Das, V. (2006). Life and Words: Violence and the Descent into the Ordinary. University of California Press.

Hajari, N. (2015). Midnight's Furies: The Deadly Legacy of India's Partition. Houghton Mifflin Harcourt.

Khan, Y. (2007). The Great Partition: The Making of India and Pakistan. Yale University Press.

MacKinnon, C. (1994). Rape, Genocide, and Women's Human Rights. Harvard Women's Law Journal, 17, 5-16.

Menon, R., & Bhasin, K. (1998). Borders and Boundaries: Women in India's Partition. Rutgers University Press.

Mookherjee, N. (2015). The Spectral Wound: Sexual Violence, Public Memories, and the Bangladesh War of 1971. Duke University Press.

Pandey, G. (2001). Remembering Partition: Violence, Nationalism and History in India. Cambridge University Press.

Posen, B. (1993). The Security Dilemma and Ethnic Conflict. Survival, 35(1), 27-47.

Saikia, Y. (2011). Women, War, and the Making of Bangladesh: Remembering 1971. Duke University Press.

Smith, D. L. (2011). Less Than Human: Why We Demean, Enslave, and Exterminate Others. St. Martin's Press.

Tajfel, H., & Turner, J. C. (1979). An Integrative Theory of Intergroup Conflict. In W. G. Austin & S. Worchel (Eds.), The Social Psychology of Intergroup Relations (pp. 33-47). Brooks/Cole.

Talbot, I., & Singh, G. (2009). The Partition of British India. Cambridge University Press.

Tambiah, S. J. (1996). Leveling Crowds: Ethnonationalist Conflicts and Collective Violence in South Asia. University of California Press.

Tilly, C. (2003). The Politics of Collective Violence. Cambridge University Press.

Turner, V. (1969). The Ritual Process: Structure and Anti-Structure. Aldine.

Waller, J. (2002). Becoming Evil: How Ordinary People Commit Genocide and Mass Killing. Oxford University Press.

Zimbardo, P. (2007). The Lucifer Effect: Understanding How Good People Turn Evil. Random House.