1. Introduction
During the winter of 2019, public protests erupted across India in response to the controversial Citizenship (Amendment) Bill, which was introduced by the Bharatiya Janata Party (hereafter, BJP) government and passed into law in December. The Citizenship (Amendment) Act (hereafter, CAA) provides a route to Indian citizenship for undocumented residents, based on their religion. The Act expressly excludes Muslims by specifying that the amendments apply to “any person belonging to Hindu, Sikh, Buddhist, Jain, Parsi or Christian community from Afghanistan, Bangladesh or Pakistan.” The protestors believed that the exclusion and the reasons behind it furnished proof of the anti-Muslim policy which threatened India’s secular Constitution.Footnote 1 In several states governed by the BJP, including Uttar Pradesh (hereafter, UP), the protests were met with harsh police action and counter-protests, the latter organized by pro-government and pro-CAA nationalist groups. Violent incidents and police excesses were recorded in more than 90% of the 24 UP districts where protests were held (Singaravelu, Reference Singaravelu2020). According to media and fact-finding reports, between 19 and 21 December 2019, across the state, the UP police, accompanied by Hindu nationalist groups, clashed violently with the protestors, resulting in the deaths of at least 23 people, of whom 22 died from gunshot injuries (Johari and Subramanian, Reference Johari and Subramanian2019; Suresh et al., Reference Suresh, Firdous, Khan, Kapoor, Prasad, Shaheen, Jamil and Ghufran2020, p. 13; Singaravelu, Reference Singaravelu2020; Zargar and Tanha, Reference Zargar and Tanha2022, p. 9). During this period, it was public knowledge that the UP police had complete impunity, while they picked up Muslim men from both affluent and poor Muslim neighbourhoods (Johari and Subramanian, Reference Johari and Subramanian2019; Kuthar and Gulfam, Reference Kuthar and Gulfam2019; Shroff, Reference Shroff2019). It was reported that Muslim homes, shops, and businesses were ransacked, and their places of worship vandalized (Dhara, Reference Dhara2019; Zargar and Tanha, Reference Zargar and Tanha2022, p. 6).
Such violence must be understood within the broader transformation of penal governance in UP under the Yogi Adityanath-led BJP government—a transformation marked by the increasing criminalization and targeting of Muslims through state-sanctioned punishment. This targeting of Muslims extends well beyond overt police brutality, such as in extrajudicial killings and staged “half-encounters” (Dixit, Reference Dixit2018), or the violent repression of anti-CAA protests. It also takes more insidious forms, blurring the lines between the legal, illegal, and extra-legal. Local authorities have turned to the strategic use of civil suits and municipal laws, like building codes, to justify the demolition of Muslim-owned homes and businesses, a practice identified by scholars as a form of domicidal violence (Amnesty International, 2024; Bhatia, Reference Bhatia2024; Jamil, Reference Jamil2024). At the same time, selective enforcement of laws—from the anti-conversion statutes to cow-protection measures—has produced the disproportionate criminalization of Muslims. Within this penal apparatus, the police have become the principal agents of selective punishment, deploying informal and extra-legal practices to realize political objectives. In a striking rebuke, the Supreme Court of India, in April 2025,Footnote 2 condemned the UP police for their role in what it called an “absolute breakdown of the rule of law,” highlighting the rise in the UP police’s tendency to escalate and frame civil matters as criminal cases (Ananthakrishnan, Reference Ananthakrishnan2025). It observed the UP police’s abuse of power by way of levying multiple first information reports, and unjust arrests.
Along with the police’s “generally violent approach to everyday policing” directed against Muslims and other minorities (Vasudev and Hansen, Reference Vasudev and Hansen2024, p. 10), in the six years following the post-CAA deaths and destruction, a pattern of state action has emerged, through which the police and local bureaucracy have undermined the principles of natural justice for the victims of violence. In their engagements with the law, the police have deployed techniques and processes that have incrementally caused an erasure of victimhood for Muslims and the transformation of (Muslim) victims into culprits. This paper, thus, seeks to understand the process of the de-recognition of Muslim victims through formal procedures by law enforcement agencies, and their identification as culprits and criminals. I trace the way this renders the crimes perpetrated against Muslims invisible and diminished in law. Such a denial of victimhood to Muslims impacted by, and killed during, the CAA-protest violence is rooted in the construction of Muslims historically to be the perpetrators, and contributes to the repositioning of the majority, and in this case, the majoritarian state, to be the sole and perpetual victim.
The starting point of this paper, thus, is that not everyone killed in incidents of anti-Muslim violence is recognized as being a victim of violence. The recognition of a person as being a victim is a promise by the state to redress the harm done to them. The accordance of the “victim” status is often based on the socio-political context, which then determines the entitlements that the victim has to any legal obligations under the state. While victims of all kinds of caste, gender, and political violence symbolize the presence of injustice in a society, victims of anti-minority violence, in the context of a Hindu nationalist state, require particular attention.
By examining the afterlife of the CAA-related violence, we may explore how ordinary, everyday police work legitimizes the de-victimization of Muslims killed in the violence. I analyse how legalities and legal tools are transformed and manipulated beyond recognition in their utilization by state agencies. In the following sections, the paper shows how, in the aftermath of violence, the filed cases and subsequent investigations leading to the assignment of blame are not about uncovering criminal culpability, but about the production and retrenchment of criminalized subjectivities. The paper explores the shift in the BJP state’s penal policies to target Muslims, through the deployment and interpretation of existing laws and legal structures.
2. Research method
This paper takes a qualitative approach to study 20 of the police violence cases that occurred during the December 2019 CAA protests in UP. Primary ethnographic fieldwork, with semi-structured interviews and participant observation, was conducted between 2021 and 2023 with 80 interviews across 7 districts in UP. The interviewees included retired High Court Justices, current district court judges, retired and current officers from the Indian Police Service, politicians and societal elites, local journalists in Delhi and UP, civil rights activists and NGO workers, academics, advocates, and survivors and family members of victims of violence.
The interviews were supplemented by documentary sources including first information reports (hereafter, FIRs) filed after the anti-CAA violence, letters written by the complainants in cases where FIRs were not filed, petitions filed in the Allahabad High Court, charge sheets and final reports of cases (where available), judicial orders and legal documents, fact-finding reports of the episodes of violence across UP, and the media reportage during the period. The interviewees have been anonymized herein; the deceased, however, have been named, since the cases were reported extensively in the press. The interviews were in English, Hindi, and Urdu, indicated as such when cited. The translations are mine.
3. Policing and protecting: Victims and perpetrators
This section engages with anthropological works on policing of minorities that explain the ways in which majoritarian politics intwine with the police practice through, not only the exercise of violence, but also several discretionary and documentary techniques that mark police complicity and expand state impunity. Scholars have sought to conceptualize majoritarian policing, or the policing of minorities within majoritarian contexts in various ways. Empirical studies have documented a pattern of prejudicial policing across several episodes of riots and pogroms, including the police’s overt and covert involvement in the violence (Rai, Reference Rai2008, p. 200; Verma, Reference Verma2021; Vasudev and Hansen, Reference Vasudev and Hansen2024). Such police complicity extends beyond immediate physical violence to, also, procedural practices during and after the incidents (Engineer, Reference Engineer1989). These include the police not registering complaints of Muslim complainants (Chenoy et al., Reference Chenoy, Shukla, Subramanian and Vanaik2002; Narula, Reference Narula2002), even when the names and addresses of perpetrators were clearly identified, and “lackadaisical crime reports” being written without conducting “serious investigation” (Srikrishna, Reference Srikrishna1998, p. 38). Such exercise of police discretion to circumvent established procedures, by not collecting substantial evidence, and avoiding the registration of complaints, was also noted in the 2020 Northeast Delhi violence that left 53 dead, disproportionately affecting Muslim communities (Delhi Minorities Commission, 2020, p. 19).
In the aftermath of violence, police documents, inquiry reports, media records, and official exonerations do not merely describe the failures and causal events, but play a performative role in shaping state narratives (Brass, Reference Brass2003) that justify the violence and delineate victims and criminals. Beatrice Jauregui’s concept of “discretionary and provisional authority” helps to theorize such police choices, not as anomalous failures, but as routinized forms of governance, where officers improvise within institutional templates, and those improvisations rapidly become everyday practices that produce differential legal outcomes (Jauregui, Reference Jauregui2016). During incidents of communal violence, such majoritarian discretionary policing works to produce an official narrative constitutive in police reports, legal documents, and judicial summaries, which also act as instruments that make certain stories administratively coherent while occluding others (Chatterjee, Reference Chatterjee2023a, p. 258).
In cases of violence between Muslims and Hindus, thus, official discretionary authority almost always denies Muslim victimhood, despite the deaths and destruction suffered. Moyukh Chatterjee demonstrates how legal structures and discretionary police action allowed the 2002 Gujarat violence against Muslims, which served the purpose of creating “wounded majorities and killable minorities” (Chatterjee, Reference Chatterjee2023b, p. 128). In his analysis of the police documentation and proceduralism in the aftermath of the 2002 Gujarat pogrom, Chatterjee found that official police documents comprising a legal archive of criminality were used in courts to absolve those very criminal acts (Chatterjee, Reference Chatterjee2017, p. 123). This led to framing the majorities as “wounded,” suggesting a pre-assigned status of victimhood, while the “killable” minorities insinuated that the punishment allocated to minorities was represented as legitimate. Scholars have observed similar techniques creating an “impunity effect” (Chatterjee, Reference Chatterjee2017) where the police create records (and mis-records) of crimes during episodes of communal violence, such as during the 1984 anti-Sikh pogrom (Das, Reference Das2007, p. 164; Grover, Reference Grover and Varadarajan2002, p. 364). Such policing, observed during large-scale violence, extends from discretionary practice and legal framing to administrative erasures and manipulations in the official documentary and procedural records.
One recurring documentary practice is the use of an “omnibus” format for FIRs in which several different charges that have occurred in one area are registered into a single document (Chatterjee, Reference Chatterjee2017, Reference Chatterjee2023a, p. 252; Grover, Reference Grover and Varadarajan2002, p. 363). For example, an omnibus FIR, as noted in Chatterjee’s legal ethnography, might record “rioting, arson, looting” across several localities under a single entry without naming individual victims, and neither citing witness statements, nor distinguishing temporally separate events. Such formal collapsing forecloses individualized investigation, and also creates an archival story to be read as coherent criminality, rather than as contested terrain. Veena Das argues that bureaucratic instruments, such as omnibus FIRs and selective entries, act as narrative technologies, which order incidents into state-legible storylines that determine which complaints are investigated, and which are omitted (Das, Reference Das2007, p. 165). Chatterjee highlights, as others have, that police practices that modify and alter legal documents lead to the exclusion and erasure of some incidents of violence, and thereby the evasion of criminal accountability (Chatterjee, Reference Chatterjee2023a, p. 255).
Here, it is important to highlight the colonial-postcolonial continuities in the policing of minorities, to understand how the ordinary, procedural forms of policing described above can render extraordinary violence to be administratively routine and everyday events. In analysing the historical continuities in majoritarian policing, several scholars have demonstrated that, in the case of minorities, Dalits, and Adivasis, there is a blurring of boundaries between “civilian and military policing” (Raman, Reference Raman, Heath and Lokaneeta2025, p. 47), as well as in the deployment of “ordinary and exceptional law” (Heath and Lokaneeta, Reference Heath and Lokaneeta2025, p. 27). This debate—on colonial continuities, martial-civilian blurring, violence against Adivasis and ordinary/exceptional law—demonstrates that police work can be simultaneously bureaucratic while also performative, or mundane while also spectacular (Gera, Reference Gera, Heath and Lokaneeta2025, p. 244; Heath, Reference Heath, Heath and Lokaneeta2025, p. 35). Thus, the historical continuities in policing have included procedural practices, such as filing FIRs, maintaining case diaries and charge sheets, which furnish ordinary face validity, even as they conceal extraordinary violence against minorities (Gera, Reference Gera, Heath and Lokaneeta2025; Heath, Reference Heath, Heath and Lokaneeta2025).
I read these studies as foregrounding techniques of majoritarian policing deployed during episodes of large-scale, spectacular communal violence. However, even in routine interactions, police engagement with Muslim citizens has been marked by erasures of accountability and inversions of victimhood that are shaped and influenced by majoritarian politics and securitization narratives (Ahmad, Reference Ahmad2023). To understand these forms of interactions, scholars have drawn attention to the role of the state as a protective force, where the deployment of violence against minorities is framed as essential to maintaining order and safeguarding the rights of the “lawful” citizen (Khanikar, Reference Khanikar and Khanikar2018, p. 96). Importantly, this security imperative relies on distinguishing the citizen from a perceived “criminal other,” thereby rationalizing the use of coercive practices such as police brutality and torture, under various pretexts (Khanikar, Reference Khanikar and Khanikar2018, p. 80).
Such securitization emerges from the prevalence of prejudicial attitudes within police ranks that treat Muslims as more “naturally prone” to committing crimes—a view concentrated in personnel from the so-called Hindi-heartland, including UP (Jain and Jha, Reference Jain and Jha2019, p. 152; Jha et al., Reference Jha, Prasad and Mittal2025, p. 55). These criminalized perceptions are compounded by the over-representation of Muslims among undertrials and detenues, as per official statistics (National Crime Records Bureau, 2022, pp. 67, 71), and by the pattern whereby Muslims are labelled as “anti-national” and are disproportionately charged under stringent terror laws such as the Unlawful Activities (Prevention) Act (Singh, Reference Singh, Ramraj, Hor, Roach and Williams2012, p. 427; Rao, Reference Rao2026). Studies on the long history of terror laws, which cut across different political regimes, establish that the state’s construction of internal and external threats produces Muslims as primary “suspect subjects” (Eckert, Reference Eckert2012, p. 336). Such criminalized identities have often served to legitimize the violent punishments, as justice, being meted out to them by the order-maintaining state, in the form of custodial violence or police encounter killings (Eckert, Reference Eckert2005).
Thus, to situate the criminalization of Muslims within the contemporary techniques of majoritarian policing in UP, it is necessary to note that, although there is an incremental history, the entry of the ethno-religious project of Hindutva into penal governance has, over the past decade, produced new forms of authoritarian law. Hansen, along with other scholars, argues that the rise of Hindu nationalism has helped to normalize state and public violence against Muslims as legitimate forms of policing, most evident in north Indian states (Basu, Reference Basu, Hansen and Roy2022; Hansen, Reference Hansen2021; Jaffrelot, Reference Jaffrelot and Schoch2021; Nielsen et al., Reference Nielsen, Selvaraj and Nilsen2023). A shift in majoritarian policing has been documented as an increase in “encounter” deaths and “deaths in custody” in UP (Dixit, Reference Dixit2018; Outlook Web Bureau, 2024. This, alongside incremental shifts in the forms of investigative malpractices in which the police lodge FIRs against the deceased, rather than the officers involved, further leads to criminalizing the victims and obfuscating police accountability (Verma et al., Reference Verma, Kumar and Appnender2021, p. 108). These practices of policing are also noted in judicial records.
Courts have questioned the increasing “hateful” intentionality behind the police’s punitive responses. In orders relating to police violence against a Muslim man in the 2020 Northeast Delhi pogrom, the Delhi High Court stated that: “The policemen…were motivated and driven by religious bigotry, and therefore [the violence] would amount to a ‘hate-crime’.”Footnote 3 The judgment further underscored the court’s concern that the investigation had been deliberately negligent and biased in favour of the accused officers, who, despite being entrusted with upholding the law, appeared to have acted from positions of prejudice and abuse of authority. It stated:
Without making too much of a harsh comment, this court is constrained to observe that the investigation in the present case has evidently been tardy, sketchy, and conveniently sparing of the persons who are suspected to be involved in brutally assaulting the petitioner’s son. What is worse is that the suspects were entrusted to act as custodians of the law, and were in a position of power and authority, but seemed to have been driven by bigoted mindsets.Footnote 4
Along these lines, majoritarian police violence against students during the anti-CAA protests has been named as forms of “hate-crimes” which highlight the intentionality and impunity through which police targeted Muslim protestors (Ali and Chakraborty, Reference Ali, Chakraborty, Heath and Lokaneeta2025, p. 143). In addition, to analyse the anti-CAA protests, it is important to place station-level FIR politics, ad hoc evidence-handling, and micro-level practices against the backdrop of macro-level institutional impunity and majoritarian ideology in UP’s policing ecology.
Therefore, I argue that the erasures of victimhood and “vanishings of crimes” (Bhat et al., Reference Bhat, Bajaj and Kumar2020) that mark majoritarian policing are not mere administrative oversights but the product of discursive, procedural, investigatory, and legal mechanisms that invert criminals and victims. I build on the literature that views proceduralism itself as a technology of erasure: formal investigations, omnibus FIRs, and terse charge sheets channel majoritarian politics into bounded bureaucratic processes, thereby neutralizing the violence as an administrative problem (Chatterjee, Reference Chatterjee2023b; Suresh, Reference Suresh2023). By looking at how paperwork, discretion, and legal framing interlock, I demonstrate how paper practices do more than mis-record—they help to determine who counts as a victim, and who can be killed with impunity. Further, I argue that mundane procedural acts, such as how an FIR is phrased, which law is applied, which witnesses are recorded, or how a “compromise” is reached (Baxi, Reference Baxi2013), function as technologies of statecraft that scale upward into durable majoritarian policing. The following sections develop this argument by analysing empirical case studies alongside doctrinal analysis of the court cases that were generated during the anti-CAA protests.
4. The “unmaking” of the victim
The CAA, passed in December 2019, was a precursor to the compilation of a National Register of Citizens (NRC) (Bhatia, Reference Bhatia2021). The country-wide NRC followed an exercise in Assam, which discovered 1.9 million undocumented people (700,000 Muslims and 1.2 million non-Muslims) (Jha, Reference Jha2019). Roy argues that the CAA is believed to have been enacted in order to serve also as a safety-net for those non-Muslims who were unable to prove their citizenship in the NRC exercise (Roy, Reference Roy2019). Simply put, the critique of the CAA was that by introducing religion in citizenship laws, it laid the basis for a legal process which could detain or deport those determined by the state to be illegal migrants: mostly, Muslims.
Protests against the passing of the CAA, which began in the national capital in mid-December 2019, spread throughout UP and were among hundreds that developed, rapidly, across the country. While the protestors comprised people of different faiths or those who did not profess faith-based politics, in some of the smaller towns in UP, they were led by Muslim men and women, although the sit-ins were mainly by Muslim women. While, on the one hand the protests were an exercise in democracy by rights-holding citizens, on the other hand they were severely repressed by violent police action in several states. In 8 UP districts, between 19 and 21 December 2019, at least 22 Muslim men were shot dead by the police: 7 men were shot dead in Firozabad, 5 in Meerut, 3 in Kanpur, 2 in Bijnor, 2 in Sambhal; and 1 each in Lucknow, Muzaffarnagar, and Rampur (BBC, 2019; Johari and Subramanian, Reference Johari and Subramanian2019; Shroff, Reference Shroff2019; The Polis Project, 2020; Yadav, Reference Yadav2023; Zargar and Tanha, Reference Zargar and Tanha2022, p. 9).
In the aftermath of the violence, rather than initiating an independent commission of inquiry, or a suo motu intervention by the judiciary, the police intensified its crackdown on alleged protesters (all Muslim) under the guise of search operations, and placed the blame for the police violence squarely on them. During interviews in Muzaffarnagar, local journalists recounted how the police were effectively granted unchecked authority to target Muslim neighbourhoods, then restricted for journalists. One journalist told me,
Instead of investigating the protest-violence and killings, the police rained down on Kacchi Sadak [an affluent Muslim-majority neighbourhood] and other Muslim mohallas [areas], undertaking violent search operations. For 24 hours, no journalist was permitted to take photos or visit these areas.Footnote 5
The wide combing operations by the police to punish protestors and non-protestors, to “teach them a lesson” and discourage further large-scale protests, were framed as “orders from the top.” As another local Muzaffarnagar journalist recounted to me the cover-up of the violence was pre-planned:
The police broke into Muslims’ houses and vandalized them. When we went there, the police did not let us enter the lanes, not even to photograph the houses. The next day, they barricaded the lanes in Kacchi Sadak. The ADM [additional district magistrate] gathered all the journalists and said ‘No one will take any photos or publish anything from here. We have orders from the top to teach them a lesson’.Footnote 6
In this context, “teaching a lesson,” a form of pedagogical violence that has been documented as a technique of disciplining and punishing marginal castes, genders, and communities, is understood to be majoritarian and when this is the “order from the top,” impunity is both conserved and expanded. Now I turn to the techniques of policing that unmake victims.
4.1. Discursive processes of blaming victims: The gendered politics of prophylactic punishment
Police and media narratives displaced all responsibility onto the protestors, recasting those who were shot or killed as having been culpable actors rather than victims. This discursive shift—framing protestors as “rioters” or “anti-national”—performed a reclassification that neutralized any possibility of institutional redress.Footnote 7 In the immediate aftermath of the December 2019 protests, Chief Minister (CM) Yogi Adityanath publicly branded protestors as “rioters” and referred to them as “anti-social and anti-national elements” who were misleading people and spreading violence (HT Correspondent, 2019b; TNN, 2019b). The protestors were labelled as miscreants, and the sit-ins as public nuisances, portrayed as being disturbances to social peace, “order” and harmony, language that prefigured and legitimized the forceful interventions by the police. UP police and district officials echoed this language; for instance, a Lucknow police tweet noted that “‘some anti-national elements’ tried to disturb the atmosphere; about 150 ‘nuisance makers’ were taken in custody” (UP POLICE, 2019).
This narrative technology (Das, Reference Das2007, p. 81) of the state was consequential since it re-coded the political protests as criminal disorder, allowing state actors to produce an administratively legitimate justification for coercion. Indeed, the rhetoric of criminality was mirrored by police conduct at protest sites. Across many districts of UP, local police forces intervened pre-emptively in protest activities, pressuring demonstrators to dismantle encampments days before any outbreak of violence. In mid-December 2019, UP authorities invoked Section 144 of the Code of Criminal Procedure, 1973, for a two-day ban, with no permissions for any rally “in all of UP” (Suresh et al., Reference Suresh, Firdous, Khan, Kapoor, Prasad, Shaheen, Jamil and Ghufran2020, p. 131; TIMESOFINDIA.COM, 2019). This prohibitory order criminalizes gatherings of more than four people when applied in the case of riots, protests, pandemics, or any perceived threat to public order.Footnote 8 A factfinding report documented that the state issued “cautionary” notices to more than 3,000 people in UP, warning them not to protest on 19 December 2019, and detained about 3,305 people that day, which rose to about 5,400 after 2 days (Zargar and Tanha, Reference Zargar and Tanha2022, p. 5). It states:
The state administration and all its bodies indulged in intimidation tactics by serving ‘caution’ notices to over 3000 people, warning them to abstain from protesting on the eve of the first all India protest across the state scheduled on 19.12.2019. Later, they were also formally arraigned in various FIR’s. Approximately 3305 people were detained from the protests that day. Two days later, this number rose to 5400 (Zargar and Tanha, Reference Zargar and Tanha2022, p. 5).
These interventions were not neutral exercises of public-order maintenance but part of a deliberate strategy to foreclose spaces of protest. Importantly, the police response was gendered in their deployment of legal and extra-legal pressure in ways that reproduced patriarchal and communal hierarchies. A pathologist and social activist, involved in one of the protests in Sambhal, recounted the targeted pressure by both the police and the Local Intelligence Unit (LIU).Footnote 9 The protest was notably led by Muslim women of all ages—“mothers, daughters, and grandmothers”—whose organized and visible presence in public space challenged both gendered and communal norms. The interviewee shared:
There was pressure from the police a few days before the violence only. The Additional SP [superintendent of the police] wanted me to stop the women’s camp… When the police sprayed water on us, even the children asked, ‘Why are you doing this to us? Isn’t it our right to protest?’… A few days before [the violence], my daughter had given a speech at the protest site. I later got an intimation from the LIU [local intelligence unit], saying ‘Don’t send your daughter and wife to the protest or there will be an FIR against them’.Footnote 10
Rather than confronting the women directly, the police weaponized patriarchal family structures. The male relatives were threatened with legal action should they not “control” and restrain their “wives and daughters.” The tactic not only invoked familial authority, but it also re-inscribed it as a tool of state discipline. The message was clear: a woman’s political agency became illegitimate, unless it was sanctioned, mediated, and curtailed by male kin. In effect, the state outsourced gendered policing to patriarchal social relations, converting private family sanctions into instruments of public order.
This mode of control indexes a deeper state anxiety, where the protesting Muslim woman disrupts two dominant expectations. First, of feminine passivity and second, of minority marginality, thus, making women’s presence particularly threatening to majoritarian imaginaries. Women’s occupation of public space is therefore doubly subversive, and elicits a distinctive policing response. By weaponizing the threat of FIRs and legal persecution through patriarchal lines of authority, the police exercised a gendered form of state power, wherein Muslim men, under duress, were made custodians of women’s obedience. This indirect mode of control, exerting pressure on men to manage the political expressions of women, demonstrates the manner in which state institutions continue to collude with patriarchal structures to suppress dissent, especially when it comes from politically articulate and organized Muslim women.
In Lucknow, the police disallowed public gatherings by issuing orders under Section 144 of the Code of Criminal Procedure, 1973. The police also placed several activists, initially under house arrest, and then later arrested and detained. As documented in civil society reports, in the early hours of 19 December 2019, police surrounded activist Dr. Sandeep Pandey’s house and “put [him] under house arrest … preventing him from stepping out to take part in [the Dec 19] protest” (Suresh et al., Reference Suresh, Firdous, Khan, Kapoor, Prasad, Shaheen, Jamil and Ghufran2020, p. 86). Similarly, 73-year-old human rights lawyer Mohammad Shoaib was “arbitrarily placed under house arrest” on the night of 18–19 December (Ellis-Petersen and Rahman, Reference Ellis-Petersen and Rahman2020), and 76-year-old retired IPS officer S.R. Darapuri was first kept under watch at home, and later taken to Ghazipur police station and then “to the jail on December 21” (SabrangIndia, 2019). While the police “kept [Mohammad Shoaib’s] house under watch” throughout, the 73-year-old social activist was later accused of leading the very protest that he could not attend (Ellis-Petersen and Rahman, Reference Ellis-Petersen and Rahman2020). These tactics—widely reported by media and rights groups—effectively silenced key organizers before the planned demonstrations.
At the local level, the police’s unchallenged engagements with the law, enabled them to deploy forms of control to discipline citizens’ behaviour. These measures were framed as preventive law enforcement but functioned as bureaucratic instruments to suppress dissent. A retired IPS officer involved in social activism told me that anti-CAA protests were criminalized from the outset, in Lucknow as in other districts in UP. Describing the administrative techniques designed to “frustrate” citizens, they said that these included a delay or denial of permissions, onerous undertakings, and prohibitions on loudspeakers. They observed:
It is almost impossible to organize a protest here – they [the state] treat dissent like it’s a crime. You apply for permission [from the police or district administration], but you won’t get it, or they will delay it [until after the scheduled event]. Even if you want to hold a meeting or a conference in a hall – you are required to seek permission and give an undertaking that you will be responsible. This is a method to frustrate the dissidents. They have shut [banned] the loudspeakers here completely… But the Hindu side make their processions with speakers in public areas without permission. No rules for them…Footnote 11
Such pre-emptive bureaucratic gating of protests—control over permits, venues, and sound systems—operates selectively to prevent oppositional gatherings. The prohibition of loudspeakers at public gatherings, for instance, starkly contrasts with the police’s usual approach during religious festivals or celebratory events. During my fieldwork in Lucknow and other districts, I observed that religious processions, public celebrations, and festival-making were ubiquitous, and a largely unremarkable part of the everyday rhythm of life. Throughout the year, streets across several districts of UP are alive with colours, deities, music systems, and bonfires. Against this backdrop, the exceptional restrictions imposed during the CAA protests—distinct from those applied to religious or festive gatherings—revealed how protesters were viewed through a lens of criminalization.
These measures were bolstered by the various warnings, sureties, and house arrests imposed by the police. The retired police officer also highlighted the stringent “precautionary” measures used by the police to subdue protests, and clamp down on protestors. They said:
[During the December CAA protests] a few of us were asked to give a one lakh rupees [∼£1000] bond to provide surety of not being part of the protests. If I violate it, then I will be held responsible. This was done to forbid us from participating in protests… I also got a notice [from the police, saying] that ‘There is a CAA-NRC [related] meeting [in Lucknow] and it is understood that you are likely to take part, and if you violate section 144 then action will be taken against you’.Footnote 12
Legal instruments such as bonds, notices, and house arrests function as predictive punishments that would hold the victims responsible by imposing individualized costs on collective action. Criminalization via narrative-creation and administrative measures thus prepared the ground for subsequent legitimization of force during the anti-CAA protests. Once the protests had been discursively and administratively marked as illegal, the heavy-handed police action that followed was retrospectively framed as necessary for the “maintenance of law and order.” Consequently, those killed or injured during crackdowns were pre-discredited by virtue of having been present at the time and sites of the, allegedly, illegal gatherings.
4.2. The law’s posthumous criminalization
After the anti-CAA police killings, some of the deceased men were, posthumously, accused of rioting and destroying public property, and subsequently framed in police files, as having been criminals instead of victims. Where investigative processes did proceed, charge sheets commonly constructed symmetrical narratives—suggesting mutual violence among groups—thereby diluting state responsibility. For example, the Meerut charge sheet framed incidents as “gang wars” between the protestors, naming an alleged “gang leader” who had been identified on a CCTV recording. It reported that two, alleged, protestors, Zaheer Ahmed and Asif Khan, whose bodies were found in the Lisari gate area in Meerut, had died as a result of being killed in a crossfire during the violence. Subsequently, the police expanded the charge sheet to include Section 304 of the Indian Penal Code, invoking the provision for “culpable homicide not amounting to murder” against the protestors. In these official narratives, deaths were presented as consequences of intra-community violence, or as “accidental,” rather than as outcomes of state action.
This legal framing produced two effects: (1) it shifted public and juridical focus away from state violence and toward alleged criminality among protestors; and (2) it conditioned evidentiary expectations—courts and oversight bodies were then presented with a record that emphasized reciprocity of violence or presence of a “conspiracy”, rather than police culpability. Scholars have observed this as an “eyes wide shut” approach in the Supreme Court’s January 2026 Gulfisha Fatima & Others v State (NCT of Delhi) order that relies on the police chargesheet alleging a “conspiracy” by protestors, with weak evidentiary thresholds (Bhatia, Reference Bhatia2026). People who were victims of police violence were categorically classified as being participants in their own injuries and deaths.
Some eyewitnesses to the police actions who had been hit by bullets survived, but were arrested months after the incident, under the category of having been “unknown protestors.” In Kanpur, a labourer was shot in the thigh as he was crossing the road; he survived, as did two others who were also shot. Three months after the incident, the police arrested him, and he was jailed for five months before being released on bail. Accompanied by a human rights activist, I visited him at his home, a small shanty in a Muslim ghetto in Kanpur, where he told me that he tried to “lie low” and “not seek justice” for being shot by the police. He said,
All three of us who were shot have been accused of being rioters in the case under ‘agyaat’ [unknown]. The police came after the week of the incident to take my statement – I didn’t say anything and didn’t accuse anyone because I didn’t want to get them [the police] angry. But three months after the incident they arrested me.Footnote 13
These patterns—delayed arrests, retrospective charges, and the application of unknown-protestor labels—illustrate how criminalization can be temporally displaced to fit institutional narratives.
Authorities issued hundreds of recovery notices for alleged destruction of public property, sometimes months or years after the events themselves. For example, the state issued some 274Footnote 14 recovery noticesFootnote 15 against alleged participants in the protests, often with little substantive evidence, or adherence to procedural safeguards (Jain, Reference Jain2022). By bypassing robust judicial oversight, these administrative recoveries effectively made the state prosecutor, judge, and executioner in one. The private property was attached, fines levied, and public notices pasted in ways that publicly branded accused persons before any criminal conviction. The recovery of damages in the form of fines was termed as seeking “revenge” from the protestors. Print media reported that the CM vowed on 20 December 2019 that the government would confiscate the property of identified “rioters,” stating “The rioters have been identified on CCTV footage. We will seize their property and seek revenge from them by making them pay” (“Ve sab videography mein aa chuke hein, CCTV mein aa chuke hein. In sab ki property ko zabt kar ke inse iska badla lenge,” translated from Hindi) (TNN, 2019b). On 27 December 2019, tweeting with the hashtag “#TheGreat_CmYogi,” the CM’s office posted on X:
“Every rioter is stunned. Every troublemaker is astonished. Seeing the strictness of the Yogi government, everyone has fallen silent. Do whatever you want now, compensation will be taken from the one who caused the damage, this is Yogi Ji’s declaration. Every violent protestor will now weep because there is the Yogi government in UP” (Yogi Adityanath Office, 2019).
A Lucknow administrative order on 22 December 2019 directed a four-member panel to identify “troublemakers” and “rioters” from CCTV footage of the protests, with orders to fine them or seize assets (Abhishek, Reference Abhishek2019). Subsequently, the UP legislature codified this punitive practice in March 2020 by passing the UP Recovery of Damages to Public and Private Property Ordinance (hereafter, the 2020 Ordinance)—a move that was widely read as retroactive legitimation of ad hoc recoveries, and as the formal institutionalization of what had previously been discretionary, administrative punishments. The 2020 Ordinance not only sanctioned retrospective recoveries, but also authorized the public naming and fining of those deemed responsible for “public violence” (Express News Service, 2022). In Gorakhpur, the police put up “wanted” posters with photographs of dozens of people, at traffic intersections, listing them as “rioters” (Abhishek, Reference Abhishek2019). In some cases, the police threat and public humiliation were not essential. As a voluntary act of “repentance,” Muslims in Bulandshahr submitted a payment for 6.27 lakh rupees to the district magistrate in an attempt to seek peace and compensate for the damages to public property (Telegraph Special Correspondent, 2019). Santosh Kumar Singh, the Senior Superintendent of the Police, told the press that he viewed this act as a “positive step” which “has avoided us the long process of sending notices to the rioters and getting them to reimburse for the damage” (Rai, Reference Rai2019).
Within this context, the damage to public property legislation deserves to be read in relation to, but also as a departure from, colonial collective fines that “commodified offences” (Sharma, Reference Sharma2016). Like colonial fines written into the Indian Penal Code, 1860, the 2020 Ordinance aggregates individual alleged wrongdoing into collective financial liability. Yet it departs from earlier practices in three crucial ways. First, it transforms discretionary, locally applied penalties into a statutory machine. Second, it transfers the evidentiary burdens onto the accused. And third, it vests quasi-judicial powers in Claims Tribunals, and limits avenues for appellate review. In these ways, it converts an episodic administrative tactic into an entrenched legal instrument, raising serious concerns about due process, access to justice, and fundamental rights.
The 2020 Ordinance has been criticized for its procedural unfairness and for institutionalizing “naming and shaming” dissenters, which serve as practices that further stigmatize targeted communities, and expose them to vigilante reprisals (Thakur, Reference Thakur2020). The statute explicitly directs attachment of respondent property and public dissemination of names, addresses, and photographs once a recovery order is passed:
As soon as the order of recovery for damage is passed the property of the respondent to be attached and authorities shall be directed to publish the name, address along with photograph with a warning for public at large, not to purchase property attached (Uttar Pradesh Legislature, 2020, p. 13).
This legal tool criminalizes all those participating in any form of protest (“bundhs, strikes, riots, public commotion, protests”) in the anticipation of violence, and exposes the personal details of protestors to vigilante groups and non-state organizations known for public disciplining. Subsequently, claims tribunals were set up in Meerut, Prayagraj, and Lucknow and used to impose financial penalties on several hundred people under this Act, long after the events (Press Trust of India, 2022). The finality clause inscribed in section 22 of the 2020 Ordinance, which treats tribunal orders as effectively non-appealable makes judicial oversight illusory, and entrenches an administratively enforceable form of punishment.
The technique of legal framing under the 2020 Ordinance accomplishes three interrelated political effects. First, it materializes political exclusion, where economic sanction replaces, or supplements, criminal prosecution. The Claims Tribunals allow the police to penalize those who were not shot, injured, or arrested at the time of the protests. The codification of recovery powers and tribunals thus expands the policing repertoire, from coercive action in the streets, to structural dispossession and public shaming in legal form. These are measures that have enduring socio-economic and civic effects on targeted communities. Second, it produces public stigma, via naming-and-shaming, that could invite social and vigilante enforcement of majoritarian norms. The long-term impact is disciplinary and deterrent. During my interviews, I learnt that the financial and reputational costs of being labelled a “destroyer of public property,” became a mechanism that deterred dissent, and reordered civic inclusion along majoritarian lines. And third, it forecloses legal remedies by reversing burdens and narrowing appeal routes, thereby insulating administrative punishments from effective judicial review. Viewed through the literature on legal violence and archival production considered above, these provisions exemplify how law can be used to routinize majoritarian policing by converting political intolerance into enforceable, bureaucratic practice through legal framing techniques. The recovery notices and claims tribunals matter, not merely as administrative instruments, but also as core technologies of contemporary majoritarian governance.
4.3. The omnibus FIR: Blank papers and omissions
In the majority of the anti-CAA cases considered, the grieving families were unable to register FIRs, or were obliged to accept entries that systematically misrepresented facts. In several UP districts, their complaints, which were assigning responsibility for the deaths on police excesses, were blatantly turned down. Instead, the police framed omnibus FIRs in several districts, which consolidated charges of multiple deaths recorded in the city, along with charges of rioting, burning, and public destruction. As per the FIRs filed in Meerut, Muzaffarnagar, Firozabad, Bijnor, Sambhal, and Kanpur, the police produced a charge, for all these crimes combined against “four to five thousand unknown protestors.” The narrative of the incident, as stated in the FIRs, was marked by modifications and omissions. By altering locations and timings, the police registered themselves as being the complainants, and tailored the complaints to conceal any police participation in the incidents.
This formal consolidation altered spatial and temporal markers, effaced medical detail, and pre-shaped the investigatory record, so that subsequent inquiries proceeded from a preordained story of victim culpability. The Meerut case is illustrative: the deaths of two men, Asif Khan, a tyre-repairman, and Zaheer Ahmad, a cattle-fodder trader, were subsumed into a single FIR 1079/2019, the narrative of which attributed their deaths to “the city-wide disorder caused by unruly protestors.”
Firozabad district reveals a pattern of striking medico-legal/administrative dissonance. The gunshot entry and exit wounds, indicated in the medical records, were not reflected in several FIRs, which recorded stampedes or omitted bullet injuries entirely. Although seven men were killed in the city, allegedly from gunshot injuries (Mohammad Haroon, Muqeem, Mohammad Shafique, Nabi Jaan, Rashid, Mohammad Abrar, Arman), four FIRs do not mention bullet injuries; while Mohammad Shafique’s FIR states he died of stampede-related injuries although his medical X-ray report states a “gunshot wound to head.”
Almost all families interviewed reported police coercion—they being required to sign blank forms—and lawyers reported various fabricated entries. In an interview with a relative of one of the deceased men, they said, “After my [relative’s] death, I went to the police station to register a complaint, but the police made me put my thumb-impression on a blank sheet of paper.”Footnote 16 Similarly, while Mohammad Haroon’s medical report states “gunshot injury,” the FIR does not mention bullet injuries, but states that he was killed in a stampede. Mohammad Abrar’s and Muqeem’s FIRs do not mention gunshot wounds nor the locations of their injuries, although their medical reports show “gunshot entry and exit wounds.” In my interview with a lawyer in Firozabad, they told me that the family members of all seven men were made, by the police, to sign on blank sheets of paper and inaccurate FIRs.
Procedurally, these practices do not merely misdescribe events, they reconfigure the evidentiary universe, on which investigators and courts subsequently relied. These procedural techniques are best read as being part of a broader set of documentary technologies that produce official facts as established in the literature. Thus, the non-filing of FIRs and the misfiling or falsifying of FIRs had been done in such a way that they guided the future investigation processes. Since omnibus and misfiled FIRs pre-structure investigative priorities in order to determine who is to be treated as a suspect and which leads are to be pursued, and which forensic tests are to be commissioned, they therefore operate as gatekeepers of legal truth. This is crucial since what enters the register, largely determines what can be litigated. Courts are therefore constrained by these administratively produced records. Judicial intervention often depends on protracted litigation, medico-legal contradiction, or civil-society exposure to reopen a file. The result is an asymmetric burden on marginalized families who must undertake costly, risky, and emotionally taxing efforts to contest pre-shaped administrative narratives.
4.4. Closure reports: The terror of “faisla”
In the majority of the cases of anti-CAA killings, the UP police conducted investigations in such a way that has led to filing final reports which state “a lack of evidence,” for the killings, thereby closing each case. For example, in the case of the killing of Noor Mohammad in Muzaffarnagar, despite a recovered bullet and a forensic report, the police concluded that they could not link a weapon to an accused and, consequently, submitted a Final Report (hereafter, FR) recommending closure of case.Footnote 17 Interviews with family members and journalists in Muzaffarnagar led to the suggestion that investigative trajectories were often predetermined at the actual time of the incident. Witnesses were “rounded up,” forensic findings were downplayed, and investigative energies channelled toward narratives that exculpated state actors. A close relative of Noor Mohammad, whom I interviewed, reported: “The police did the investigation as they wished, and we were forced to accept that they can’t find the killers.”Footnote 18
The direction that the investigation took was essentially based on the way in which the FIR 0710/2019 was drafted and registered on 22 December 2019 at the Civil Lines police station in Muzaffarnagar. The FIR signed off by Noor Mohammad’s brother held that “…My brother Noor Mohammad alias Nooru … was going on foot from Meenakshi Chowk to Mahavir Chowk at around 3.30 pm, when someone from the crowd of rioters shot my brother, due to which he died.” The Final Report submitted by the police to the Chief Judicial Magistrate on 17 June 2020 reiterates the same language about the “riots against CAA” where Noor Mohammad was shot by “by the rioters.” It reads:
On 20/12/2020 during the riots against CAA, deceased Noor Mohammad died in Medical College, Meerut due to bullets fired by rioters. The bullet recovered from the deceased’s body during postmortem was found to be not of a prohibited bore. Sir, in the above case a lot of efforts were made to obtain information about the accused but no information about the accused could be obtained in any way, and neither is there any hope of getting any information in future; in such a situation there is no justification to continue the investigation. The investigation is concluded through final report no. 38/20 dated 16.06.2020 while continuing the search and investigation. Kindly accept the final report.
A local Muzaffarnagar journalist reported the ways in which the case was pre-decided, and the way in which the investigations followed suit.
A SIT [Special Investigation Team] was formed, and a magistral inquiry took place [in Muzaffarnagar]. The forensic report of the bullet [that killed Noor Mohammad] stated that it was shot by a licensed revolver. But the line of inquiry in the magistral report was along the fact that someone among the protestors had shot him. The witnesses were rounded up and evidence was framed around this. The SP city and an IPS officer were on the SIT of a 302-page murder report which amounted to nothing – this is a failure of the police.Footnote 19
Similarly, in Firozabad, the police filed closure reports in six cases stating, “We don’t know which rioter’s bullet killed which civilian. And there is no possibility of knowing it in the future either. There is no point in continuing the investigation” (Yadav, Reference Yadav2023). However, the families, along with lawyers and human rights groups, challenged the closure reports, which led to the district courts in Firozabad rejecting five closure reports, and asking the police to re-investigate the sixth death, that of Mohammad Shafique, a 40-year-old scrap dealer.
However, even when courts intervene, the re-investigations are vulnerable to administrative pressure, witness intimidation, and procedural shunting that can ultimately return cases to administrative closure. In an interview with civil rights activists in Firozabad, I learnt that the police were pressuring Shafique’s and others’ families to accept the final reports. “The police are pressuring us to do “faisla” (settlement) in exchange for money. They want us to agree with their false narrative that protestors did the shooting.”Footnote 20 Challenging the final reports, the protest petition was deposited by their advocate stating that Shafique’s FIR was filed arbitrarily, without mentioning his gunshot injuries. The petition argued that the Rasoolpur police station SHO VD Pandey shouted, “Goli maaro saalon ko (Shoot those traitors)”Footnote 21 following which, the violence erupted, and subsequently Shafique was hit by a bullet. The 2021 court’s dismissal of the FR led to reopening the case. However, after re-conducting it, the police shut the investigation and, once again, filed the FR. In a telephone interview with one of the lawyers, they told me that the police had threatened and pressurized the plaintiffs to accept the FR this time.
The police said that if he testified, he would have to face a lawsuit in a different matter. So, the family has backed out of the case out of fear and frustration. So, on 3 Jan 2024, they gave an application to the court to shut the case. On the same day itself, the police ended the investigation and submitted the FR.Footnote 22
In this way, in some of the cases of gunshot violence during the anti-CAA protests, even with some court intervention, the police managed to misdirect the re-investigation process, and coerce the victims’ families to accept the final reports that the victims had died at the hands of “unknowns.” A retired district judge encapsulated the problem; in an interview, he noted that a critical tension emerged between judicial knowledge and the formal boundaries of adjudication. The judge acknowledged that there was a pervasive issue where police investigative processes are often flawed, not incidentally, but structurally, in ways that systematically undermine cases of anti-Muslim violence. Reflecting on this, the judge remarked:
What we do is paper or file justice. The judiciary can only adjudicate based on what has been found in the case: the evidence, witnesses and how the prosecution plays out the case in court… Even I read the newspapers; everyone knows that the lynching has happened and who has done the lynching. But if the investigation is poor or made to be poor by the police, then the entire case can fall through in court.Footnote 23
This reflection reveals a disjuncture between lived, or public, knowledge and juridical proof. What the judge refers to as “paper or file justice” is, in fact, the product of a deeper structural manipulation, that might, more accurately, be termed “paper or file injustice.” Far from being a passive limitation of the legal system, this form of injustice is an active technique of police power. The judge’s statement reveals the dependence of adjudication on investigatory processes: when investigations are structurally compromised, courts are reduced to formal enforcers of pre-shaped administrative narratives. By controlling the content of case files, through determining what evidence is collected, how witnesses are recorded (or ignored), and how charges are framed, police shape the very terms on which justice is legally possible. In this light, “paper injustice” becomes a bureaucratic strategy of impunity, one that exploits the judiciary’s reliance on procedural evidence, while masking systemic bias and intentional negligence. It constitutes a subtle, yet powerful, method of insulating perpetrators from accountability, especially in communal violence cases. Thus, while the courts may formally hold discretionary power, their function is deeply circumscribed by the documentary scaffolding provided by police institutions. The paper file, rather than the event, becomes the site of truth. In this way, police investigations can be conducted in such a way as to subvert the evidence and witnesses and thereby weaken cases of anti-Muslim violence.
Taken together, the discursive, legal, procedural, and investigatory strategies analysed above constitute an interlocking system through which the victimhood of those who were killed, shot, and injured during the anti-CAA protests, was actively unmade. Discursive reclassification and blaming the victims legitimizes pre-emptive policing while gendered policing uses patriarchal structures to suppress dissent. Administrative and legal criminalization, and misfiled FIRs, produce archival stories that incarcerated meaning, and enacted pedagogical violence. The manipulated investigations converted public knowledge into juridical silence and “paper injustices.”
5. Conclusion
In conclusion, this article has sought to show how the killings of Muslim men by the police often do not lead to the provision of victim status for the persons shot dead. This is because the processes involved in the construction of victims of communal violence also impose a simultaneous need for the identification of the perpetrators of that violence and a subsequent call for justice, neither of which aligns with the Hindutva state. By deconstructing the state’s techniques and practices of pre- and post-violence management, which allow for erasures of victimhood, this article seeks to reconstruct UP as a case study for a Hindutva state (Hansen and Roy, Reference Hansen and Roy2022), which defines majoritarian policing as embedded within everyday democratic and bureaucratic systems. This is because the practices involved in unmaking victims are not an exception, but a general rule in cases of anti-Muslim violence, as can be seen subsequent to the post-Northeast Delhi violence, 2020.
This article has sought to demonstrate how following the violence, the apparatuses of investigation and everyday bureaucratic processes, deployed by state agencies in the management of cases of violence, contribute to the “unmaking” of the victims. Such unmaking takes place through layers of discursive, legal, investigatory, and procedural techniques which serve to erase victimhood and convert victims into criminals. The utilization of these modalities of the law and engagements with the legal technicalities are significant because they cause the unmaking of victims, and making of perpetrators even before the cases enter the criminal justice system. This analysis of the anti-CAA protests ties the micro-level practices of policing to the macro-level institutional majoritarianism, by building on literatures on policing, legal violence, and archives. It shows how everyday paperwork, paper “injustices,” and decision-making authority are not incidental to state violence, but are its central technologies. The erasures of victimhood are created through an exercise of systems and means of state power, which lie outside of democratic practice, but within state practice. Such majoritarian policing in a Hindutva state may thus be extended to different categories of people at different points in time.
Acknowledgements
The author gratefully acknowledges the anonymous reviewers for their insightful comments and constructive suggestions, which have strengthened this article. Heartfelt thanks to Pratiksha Baxi and Julia Eckert for this inspiring special issue and for the immense care they invested in guiding it to completion. I also thank the Henry Luce Foundation for its generous support during the fieldwork undertaken in Delhi and Uttar Pradesh.
Open access
