In the aftermath of anti-Muslim violence in Gujarat, India, in 2002, NGOs and activists encouraged survivors to testify against Hindu perpetrators in court. This article shows how state officials and perpetrators used legal procedures to transform Muslim survivors into unreliable witnesses in the courtroom. These formal and informal techniques to destabilize Muslim witnesses are best understood not as byproducts of the law’s failure to address mass violence but as a legal performance of Hindu supremacy. Not only do such trials discredit witnesses of mass violence, but they also give a legal form to the subordinate status of religious minorities within a majoritarian political regime

On February 28, 2002, when Muslims were attacked across the state of Gujarat, Muslim shopkeepers, including Asad, one of two key eyewitnesses, watched Hindu mobs burn his shop, recognizing several Hindu neighbors in the crowd. Eight years later, the men who burned and looted Asad’s shop were acquitted of all charges by a lower court in Gujarat. The judge dismissed the case because the testimony of the witnesses was not credible. Asad did not identify the accused in the courtroom, and the second eyewitness, Gafar, identified the men who had burned and looted his shop but could not give the judge the “full name” of some of the accused and could not prove his presence at the scene of the violence.

Since 2005, both witnesses had been working with a local legal aid NGO that was supporting their quest for ‘‘justice.’’ But over the years, most prosecution witnesses and the main complainant in the case turned “hostile” (changed their statements during the trial and did not support the prosecution’s case). And the trial rested on the testimony of two NGO-supported eyewitnesses, one of whom still lived in the same neighborhood as the accused.

The day before their shops were burned, 59 Hindus died during an altercation between Muslim vendors and Hindu pilgrims. Allegedly, Muslims burned the S-6 coach of the train “Sabarmati Express” just outside the railway station of Godhra, a small town in Gujarat, India.1 The deceased included many Hindu nationalist activists of the organization Vishwa Hindu Parishad (World Hindu Council, VHP), and in protest the VHP declared an All India bandh (shutdown). In the following months, large mobs attacked Muslims with impunity. National and international human rights organizations blamed the state government, ruling politicians and the police for conniving with “rioters” to attack Muslims, which left over 1,000 dead, mostly Muslims, and displaced 150,000.2

Between 2010 and 2013, I attended trials in the lower courts of Ahmedabad, the capital city of Gujarat, in order to understand the work of activists and lawyers who were helping Muslim survivors to fight for legal justice. Witnesses in “rioting” cases were being summoned by the courts to give evidence before the judge, after nearly a decade of NGO activism and political contestation around legal interventions in the aftermath of violence. But inside the courts, lawyers for the Hindu accused shouted at the witnesses, the public prosecutor looked down and shuffled papers and the accused approached witnesses to compromise the cases.

Witnesses watched their case being deferred endlessly. The accused would not show up in court, and months turned into years, as the case gradually lost all relevance for the survivor. I observed this endless process of deferral congeal into a growing stack of acquittals, some of which are being reported in Indian newspapers even as I write this article.3 These acquittals in Ahmedabad conclude the larger arc of unaccountability that unfolded in Gujarat. Despite widespread national and international attention around the ‘‘riots’’ in Gujarat, and the subsequent rise of Narendra Modi, the Chief Minister of Gujarat in 2002, and the current Prime Minister of India, most perpetrators of the violence have gone unpunished. In 2012, the Stanford Human Rights Clinic estimated that the conviction rate in 2002-related riot cases was 1.2%.4

How did the institutions designed to help survivors end up reinforcing the power and legitimacy of Hindu nationalist rule in Gujarat? Or conversely, why and how did survivors of India’s ‘‘first televised riot’’ become ‘‘unreliable and inconsistent’’ witnesses? After the anti-Muslim riots, in the absence of truth commissions and tribunals, survivors and activists approached the Indian criminal legal system, hoping to use it as a space for Muslims to bear witness and reestablish their rights as equal citizens within a secular democratic regime. However, state officials and perpetrators mobilized the legal process, including the state’s control over the architecture of the trial, the conditions in the courtroom under which witnesses were expected to testify, and formal procedures relating to criminal law and ‘‘due process,’’ to reinforce the Hindu majoritarian nature of rule in Gujarat.

Mechanical approaches to seeking and applying “the rule of law,” however well-intentioned, can further marginalize victims, especially in contexts of state-sanctioned and majoritarian violence. The actors who organize the trial and the political contexts that elicit testimony can not only disable testimony, but can also use the legal process to send a message to the witness about the restricted nature of citizenship and rights afforded to certain groups, especially marginalized minorities, within a specific regime of power. Certain forms of state-sanctioned postcolonial violence—often labeled as communal violence and rioting—persist because the law is able to recast public acts of sectarian violence as non-criminal, even if the perpetrators are identified by witnesses, and well known to the wider public.

On the one hand, a strain of postcolonial legal fetishism (Comaroff and Comaroff 2008) and legal positivism imagines and promotes the judiciary and trials as apolitical and neutral spaces within fractious regimes like India. At the same time, the formal and informal structures within which trials take place, the conditions under which witnesses are called forth to testify, the delay and deferral that pervades the legal process, and the actual performance of the trial not only impede efforts by marginalized communities to expose the actors and institutions that performed the violence, but may even facilitate the power of the state to reinforce sectarian politics.5 In effect, the societal and political implications of such legal proceedings can be the exact opposite of what is normatively expected from transitional justice mechanisms or truth and reconciliation processes. Instead of “narrowing the range of permissible lies,”6 such trials are a key site for producing and recognizing the differential rights and status accorded to majority and minority communities.

Gujarat Pogrom: The Legal Aftermath

During fieldwork in the lower courts of Ahmedabad I met Bharat, an NGO paralegal, who was helping Muslim survivors to give evidence against Hindu accused. He introduced me to Asad during one of his routine visits to meet and update the survivors about the status of their legal case. When Asad (and many other Muslims like him) agreed to fight their case, Justice First (JF), a local legal-aid NGO, gave free legal services to them. It still took several years for the eyewitnesses to testify before a judge.7 When the trials were finally conducted in Gujarat, it was in the shadow of the electoral dominance of the Hindu rightwing Bharatiya Janta Party (BJP), which has won every single local election since 2002.

I accompanied Bharat and Yasim, paralegal and lawyer in charge of his case, to Asad’s shop, the same one that was looted and burned down in 2002. Asad was preparing to go to court and give his testimony the next day. The advocate-paralegal team had been in touch with him since 2006. Finally, the moment had come when as the lawyers put it, ‘‘the witness is on his own,’’ and they cannot speak for him or her in court, nor can they face the cross-examination on his behalf. They must watch silently as the defense confuses, intimidates and challenges their witness. I watched Yasim pull out Asad’s case papers and ask him to repeat his statement.

In between distributing bags of milk and small packets of confectionery to the customers, he said, “these boys (the accused) have grown up in front of me and often stroll along the street in front of my shop, I recognize them all.” The street in front of the shop was a busy road that connected two parts of the city; bustling with cars, scooters, cycles and pedestrians. I could barely hear him above the din of the street. Like many 2002-related incidents of violence against Muslims, the accused in Asad’s case were not faceless and nameless mobs without origin and history.

Violence arrived at the doorsteps of many Muslims like him with a face and a biography. “I even know their fathers,” he said. Impressed with Asad’s confidence, and having spent most of my time in courts watching judges defer the trial, acquiescing to lawyers who came to court simply to receive the next date of hearing, or submit a variety of applications to further draw out the legal process, I looked forward to the moment when the witness finally speaks. Ultimately, however, the judge found the Muslim witnesses’ testimony unreliable and dismissed the case.

Asad and Gafar’s case exemplified how legal accountability in Gujarat was transformed into a process that reinforced Hindu power. Despite numerous official commissions of inquiry, special investigations, and civil society and NGO interventions, the Indian state has a long history of condoning “communal violence.” Survivors and activists have routinely confronted the Indian criminal legal system, and attempted to use the courts to enforce accountability, which except in rare cases, end up exonerating the accused.8 Like in previous instances of state-sanctioned violence against minorities in India, the Gujarat police initially refused to register complaints, and later summarily closed more than half of the cases. In 2004, after the National Human Rights Commission approached the Supreme Court of India, it ordered the Gujarat police to reopen approximately 2,000 cases.9

Four years later, in 2008, the Supreme Court appointed a Special Investigation Team to re-investigate ten high-profile cases. These verdicts were hailed as a major victory for the “rule of law” in India and considered a major setback for the ruling Hindu nationalist party, BJP. Interventions by the Supreme Court of India and activists’ efforts did yield important convictions, including the imprisonment of a senior politician.10 But these convictions were possible due to a variety of exceptional circumstances, like the intervention of the Supreme Court of India, extensive media attention, the transfer of cases outside the purview of the Gujarat police, witness protection measures, and the extraordinary courage of survivors and activists.11 But most other cases were acquitted in the lower courts of Gujarat, unheralded by the media, activists and civil society.12

Judges, plaintiffs, accused, paralegals, and lawyers changed as I shuffled between courtrooms in Ahmedabad, but every trial was stacked against the witness: the police had not registered the complaint properly, or only after significant delay, the investigation did not include important witnesses, and the material evidence was not collected or missing. Lawyers and activists who have tracked the legal aftermath of previous “riot” cases (especially anti-Sikh violence in Delhi 1984), encountered a similar farcical process unfold in different parts of India, and were quick to point out that the “legal outcome (of Gujarat 2002) is largely a foregone conclusion” because “the present criminal justice system is woefully inadequate to deal with State sponsored genocide. The criminal justice system assumes the existence of an independent investigating and prosecuting agency, insulated from political interference, which is not the case in Gujarat … a case for the total breakdown of the constitutional machinery needs to be made out.”13

But the machinery did not quite break down; it was used to challenge, attack, and ultimately dismiss the survivor. The courtroom in Gujarat transformed into an official space to unmake the witness; sending a second, robust message about the subordinate status of religious minorities, especially Muslims, within a Hindu nationalist regime of rule.

For over a year and a half, I watched a trail of prosecution witnesses (Hindus and Muslims) step into the witness box and mechanically deny their statements against the accused.14 As the case dragged on for years, the accused would approach witnesses to forget about the case in exchange for money. With passing time, such behind the scenes, out of court ‘‘compromises’’ increased. In some cases, survivors continued to live and work in areas dominated by Hindus, and they did not think it prudent or necessary to sour relations or start an enmity with their neighbors.

Similarly, police-chosen on-site witnesses entered the dock only to deny their statements and even presence at the scene of the violence—no one wanted to get involved in Gujarat 2002 trials. Unfazed, the public prosecutor and the judge declared these witnesses “hostile” without further interrogating them or cross-questioning them, and the trial staggered on, weakened and diluted by the lack of corroboration, resting on the testimony of a few NGO-supported witnesses.

Judicial proceedings in the aftermath of 2002 were polarized along the same religious lines that allowed a one-sided attack against Muslims in the first place. The atmosphere inside the courtrooms captured the paradox of legal activism: the Hindu accused were neither remorseful nor afraid, they walked in and out of the courtroom with the swagger of those who did not really care for the rituals performed in front of them; while Muslim complainants were tense, fearful, and needed to be cajoled and escorted by the JF paralegals.

The demeanor of the accused inside and outside the court, the tone and tenor of Hindu defense lawyers who often shouted at witnesses is an important part of victims’ experience of what it means to fight for “justice” in the aftermath of Gujarat 2002. Ironically, while the accused roamed around unrestrained inside the premises of the court, Muslim witnesses and JF paralegals exercised caution and restraint in the courtroom. The public prosecutor chanted prepared lines and shuffled papers and kept the case moving till it reached its conclusion—acquittal. But things were also relaxed inside the premises of the court complex because everyone had to wait endlessly, like in any other trial in India. Half-bored, I saw the accused chatting with witnesses, who were often their neighbors and acquaintances. Since most cases ended in acquittals, the confidence of the accused was not misplaced.

By examining how state regimes deployed existing legal processes—used official documentation to challenge testimony and intimidated witnesses inside the courtroom—I illustrate how courts play an important role in legitimizing and normalizing discriminatory regimes.

Unmaking the Witness

Legal settings can often constrain and flatten the victim’s testimony, not only ‘‘distorting’’ the survivor’s experience, but can effectively censor key social and political aspects of both past and ongoing violence.15 The Hindu defense lawyer’s aggressive tone and demeanor in court transformed the courtroom into yet another space that reasserted Hindu supremacy in Gujarat. Activist efforts to help witnesses testify did not account for the opacity of the trial “where knowledge is masked or excluded for reasons of both efficiency and policy” and the ability of the state to turn “highly stylized and rule-governed proceedings”16 against the witness.

A week before the final verdict acquitting the accused that burned Asad’s grocery shop in 2002, I sat on the last bench of the courtroom listening to the defense’s concluding oral arguments. Bharat, the JF paralegal who had been helping Asad for five years in preparation for this one day, stood outside the courtroom, for fear of being recognized as an NGO worker giving legal aid to Muslim survivors. He chatted on his mobile phone like dozens of others milling about outside the room. The courtroom was basic: a large, poorly maintained room with five rows of wooden benches.

I sat on the last bench, farthest from the honorable judge, because I was pretending to wait for my turn in some other case. The judge sat alone on a long, raised desk, and on his left a clerk examined documents before placing them before the judge. Soft-spoken Asad looked out of place inside the wooden witness box and spoke in spurts. The only words booming in the courtroom were those of the defense lawyer, who was also an influential member of the Hindu nationalist group, VHP. He was cross-examining Asad, pacing the small space in front of the judge, gesticulating with his arms, almost scolding him. Visibly startled by the defense lawyer’s assertion that “perhaps rival Muslims had burned his shop in 2002” Asad mentioned the names of the accused present in the court, but did not identify them before the judge. Instead, he told the judge, he “was not in a state to identify the accused in the court.”

During Asad’s cross-questioning, I expected JF lawyers (who were Muslims) to stand, interject and register their protest against the aggressive accusations made by the defense lawyer, but they simply sat there, unmoved and stoic. Later they justified their reaction structurally. “You see, JF is only a party to the case as a counsel to the state public prosecutor, so we can’t play an active role during court proceedings,” Yasim, the JF lawyer in charge of the case told me, “Since all criminal offenses are formally crimes against the state, only the public prosecutor has the power to object or say anything substantial in court.”17

Instead, JF lawyers mostly submitted written applications for documents and reinvestigation, added witnesses with the permission of the public prosecutor, and sometimes made an oral argument submission before the judge. They claimed anything beyond such circumscribed procedural interventions was impractical and unreasonable.

Asad’s cross-examination ended abruptly. In a single wave beginning with the judge everyone exited the room, and I followed Bharat, Yasim and Asad down the stairs to the canteen to get a cup of tea.

Yasim: So what happened? Why didn’t you identify the accused in the court?
Asad: This was the first time I entered a courtroom … I was very scared to stand in the witness box. My heart was pounding when they read the names of the accused and asked me to identify them.
Bharat: It’s okay, don’t worry, many people feel the same, it’s the way he (defense lawyer) shouts at the witnesses … He has scared other witnesses before this …
Asad:Well, it’s been eight years since I have seen the accused and they kept shuffling them before me. I also did not recognize the names by which they called them out… I was afraid of sending someone innocent to jail… My body started to shiver … in the dock… and I was not sure how the names and the faces matched. What if I made a mistake, they could say I am lying and send me to jail. The defense lawyer was saying that I am lying… I have never entered a courtroom before…

Asad’s hands trembled as he sipped his cup of tea. The lawyer gulped down his tea and left in a huff. Bharat continued to comfort Asad, saying he understood his fear, confusion, and anxiety in the courtroom. Asad’s refusal or inability to recognize the accused, whom he had previously named in his police statement, was in stark contrast to the activist and lawyer’s confidence that effective testimony meant reproducing the basic “facts” of the case: the name of the accused and the place, date and time of the offense. The tone, tenor and atmosphere of the cross-examination reminded Asad of the undiminished power of the accused (“my heart was pounding when they read the names of the accused”) and the relationship the accused enjoyed with the state apparatus (“they could say I am lying and send me to jail”).

This suggests that structures and processes designed to safeguard the rights of the accused were transformed into a message to Muslim witnesses about the sectarian nature of state power in Gujarat. The incapacity or reluctance of the public prosecutor and the defense to confront the Hindu defense lawyer (a lawyer with Hindu nationalist affiliations who represented many riot-accused Hindus) in the courtroom, coupled with the power of the state to appoint prosecutors and conduct the trial, produced a legal space that mirrored the public anti-Muslim violence in 2002 and that effectively unmade the witness. On the other hand, activists’ confidence in the basic procedures of the trial revealed a bureaucratic-positivistic faith in the law that failed to address the law as a performance embedded in the context of Hindu supremacy in Gujarat.

What’s in a Name?

Asad’s ‘‘confusion’’ between names and faces is a good place for us to broach the specific practices that disabled witnessing in the Gujarat 2002 trials. It helps us understand the legal constraints and limits of witnessing state-abetted violence. Ostensibly minor details like the physical arrangement of the accused in the courtroom, the manner in which the court summoned them, and the accusatory tone of the defense lawyer in the courtroom reinforced the hierarchy between Hindus and Muslims in Gujarat. Since Asad was not the first witness ‘‘unable’’ to identify the accused inside the courtroom, JF lawyers justified his behavior in the courtroom by saying that they lose control of the witness once he enters the dock. They also suspected that the witness had compromised with the accused in exchange for safety, land or money, without informing JF, which opposed such out-of-court settlements on ethical grounds. No doubt multiple factors, both inside and outside the courtroom, influenced Asad’s ability to testify before the judge.

In order to square Asad’s previous statements that the accused “had grown up in front of him” with his later inability to recognize them in the court, I turned to his case. JF paralegals and lawyers articulated an ostensibly straightforward method to prepare witnesses to testify in court. They focused on the four main “facts:” date, time, name and the role of the accused – that mattered during the trial. But Asad’s testimony and cross-examination show that even such basic ‘‘facts’’ like the names of the accused are never unequivocal or entirely isolated from the social relations that embed the legal process.

The second eyewitness in the case, Gafar, was a Muslim mechanic whose shop was in the same line of shops destroyed as Asad’s. He was the second JF-supported eyewitness in the case. He deposed in court a month before Asad in August 2010. And like other eyewitnesses, his police statement mentions four aliases.

On the afternoon of February 28, a mob attacked Gafar’s garage. He fled towards the nearby mill (like Asad and other Muslims in this neighborhood). He named seven people in his report, and went on to identify four in the courtroom. Three of them had aliases – “A person named H, B, and a person whose name I don’t know but I can identify him when I see him.” One of the men whom he identified by an alias was then asked his name by the judge. But he gave a different name, not his alias, recorded in the police report. In response, Gafar said, “the accused says his name is D but we recognized this accused as H.”

Unlike Asad, Gafar identified the accused in court, but he identified one of them by an alias, and had to add that “this accused has given his name (during identification) as D but we recognize him as H.” This difference between formal names and aliases also confused Asad, who refused to identify the accused at all.

This is the “we” used by other Muslim eyewitnesses in their statements to the police. Some of the accused are in fact well-known criminals in the area, known by their alias, not their formal or legal names. During his cross-examination Gafar said, “I do not know the proper name of H.” Certain strategic approaches to language in the courtroom allowed the defense lawyers and the judge to deny legal claims anchored in the everyday life of marginalized communities.

Half of the accused in this case have aliases and many of them are also accused in other 2002 riot cases in the same area. In fact some of them were well-known criminals and bootleggers, who have popular nicknames and aliases in their neighborhood. While JF lawyers attempted to fit witnesses into the standardized format of the trial, including developing a one-size-fits-all method to help witnesses, the question of aliases muddled the simple story they used to explain Asad’s failed testimony.

Yasim, the lawyer in charge of the case, simply dismissed Asad as a weak witness. Lawyers frequently described the cross-questioning as a test of strength between the witness and defense lawyers, where the defense lawyer tries to “break the witness.” Bharat, the paralegal, took a more empathetic view of the case: Asad was an old man in poor health terrified by the defense lawyer. But neither explanation could satisfactorily answer the question about the confusion between names and faces.

The police FIR (First Information Report) recorded nearly a month after Asad’s shop was burned down is also the first written statement. It is striking that the police recorded the names of the accused in the FIR, something that is missing in many reports. Asad gives the police eleven names that are part of a “communal mob that destroyed and burned shops belonging to the Muslim community.” Out of the eleven, four have aliases. But it is not only Asad; all the eyewitnesses give identical narratives and mention the same accused. The only thing that changes in the reports is the name of the plaintiff, the name of the shop, the date, and the estimate of damages to their shop. Let me quote a typical police report:

Due to the communal incident (quomi banavana) in Godhra on 27.02.2002, the VHP (World Hindu Council) announced a “Gujarat Bandh” (state-wide shutdown) on 28.02.2002, but I still opened my shop for business, but suddenly in the afternoon communal violence (quomi toofan) broke out in X, Y and Z areas of the city and then all the shopkeepers in my line of shops closed their shop and fled toward the mill. In front of our eyes, mobs of several thousand damaged and burned shops, vehicles, cabins, and houses. The communal mob (quomi tola) that destroyed and burned the shops belonging to the Muslim community, including my shop, comprised of the following…18 The above named persons in the communal mob were seen setting fire to the shops. Due to the destruction, my shop suffered losses of 100,000 rupees.”

All complaints are in the first person plural, and use “our” and “we” instead of the singular ‘‘I.’’ The complaint abounds in details that signal the collective and organized attack on shops “belonging to the Muslim community” and each report is thus a part of a larger story that the report, however imperfectly and strategically, nevertheless documents. Asad’s testimony is not an individual complaint because it mentions clearly an attack on all the Muslim shops in the area. It is also clear from the reports that the real but absent complainant, unacknowledged during the trial, is in fact the “Muslim community.”

“He’s a Hindu, he’s their Judge”

The experience of Muslim witnesses in the lower courts is a good place to understand how legal trials reinforced sectarian politics and Hindu supremacy in Gujarat. Different aspects of the trial, including the frequent absence of the accused and the modes of cross-examination, sent a message to survivors about the futility of seeking legal punishment for violence that enjoyed the sanction of the state and the wider society. After a decade of attending trials, during which time witnesses lost a day’s worth of wages every time they went to court, only to find that the judge was away, or the court was on recess, or the accused were missing, many witnesses settled the case outside the court.

These out of court settlements—called samjhauta in Hindi and samadhan in Gujarati—are a common feature of the Indian legal system and referred to as “compro” by Gujarati lawyers. Such out-of-court settlements during rape trials produce a “culture of compromise” that is crucial to maintaining public secrecy around sexual violence in India.19 In the aftermath of mass state-sanctioned violence, such out-of-court settlements can give survivors some measure of control over a legal process that is not in their hands. But such settlements can also reinforce impunity, if witnesses compromise under duress; and they are also a sign of the formal and informal, legal and illegal processes by which courts regulate the relationship between Hindus and Muslims in the aftermath of state-sanctioned sectarian conflict.

In 2002, three Hindu boys stabbed Amalibhai, a Muslim factory worker, when he went back from the relief camp to his house to find his bicycle. Eight years later, the accused and Amalibhai were present at a Metropolitan Magistrate Court in Ahmedabad. The accused sat in the front row and one of them kept going outside to check if their lawyer had come to court. As soon as the clerk announced the case, the judge deferred the hearing to accommodate a more urgent matter. Meanwhile, the defense lawyer and the JF lawyer whispered in a corner. The defense lawyer offered the JF lawyer a ‘‘compromise’’ in lieu of some money, but I learned later that the money was too little. Both the JF lawyer and Amalibhai didn’t seem to take the offer seriously, at least in front of me. But the defense lawyer kept trying to push the compromise. He approached the judge and whispered a few words as soon as the case was announced again and the clerk placed a heap of papers in front of him. On cue, the judge began a short speech on the value of compromise (samjhauta).

Judge: Where do you live?
Amalibhai: I live in neighborhood D.
Judge: Exactly. It is not as if you and the accused live in two different parts of the city. Since you live close to each other, it is better you reconcile with them … As a judge, it is my duty to conduct the trial and give a judgment. But then someone will appeal and it will go to the High Court, and then after another appeal it will go to the Supreme Court. All this will mean unnecessary running around for you, so isn’t it better that you simply live together with the accused?

After this brief speech, the judge conferred with both lawyers and the public prosecutor and adjourned the court. He gave Amalibhai a month to decide and make up his mind. Outside the court Amalibhai and I chatted over a cup of tea. He had noticed the tattered clothes of the accused and understood that they were poor. But what really irked him was the judge’s speech in court. Even though the judge’s statement was couched as a suggestion, it also signaled the law’s relationship to state-sanctioned Hindu violence against Muslims.

In such cases, punishment was unlikely and presented a narrative of how the legal process would only extend a wasteful, if not farcical, process. Between sips of hot sweet tea, Amalibhai said, “What do you expect from the judge? He’s a Hindu, he’s their judge.” In this way, regardless of the motivations of the judge, the legal process acts as a manifestation of a Hindu state, where Muslims must reconcile to sectarian violence and communal riots, which are often orchestrated and supported by state and non-state actors.

Courts of Dismissal

How can law elicit testimony only in order to dismiss it? Feminist legal scholars of sexual violence20 have argued that testifying to violence can often become a form of re-victimization rather than a therapeutic exercise. Building on this insight that testimony may not always be a form of truth telling, I show that the procedural aspects of the trial—documentation and delay—are often deployed to dismiss testimony.

Ultimately, the question before the judge is whether or not the accused in the case constituted an “unlawful assembly.” The police applied the unlawful assembly charge to most Gujarat 2002 ‘‘rioting’’ cases. Using the defense’s argument that the mere presence of the accused at the site of the offense is not enough, the judge noted that

“it has not been established that the accused had a common motive to destroy the property of the Muslim community, the connection of the accused to the offense has not been established, and the deposition of two witnesses who have identified a few of the accused are doubtful therefore this provision is not useful for prosecution’s case.”

The defense lawyer in his final written arguments quoted previous legal judgments to argue that it is not enough that the accused are merely present at the site of the offense but there should be clear evidence that the accused had gathered with a motive to commit the offense. On the other hand, the prosecution argued that the accused had “unlawfully assembled” with the common aim of destroying and looting the property of the Muslim community.

In October 2010, after years of adjournments, warrants, and court dates, the accused in the case were acquitted. The contrast between written documentation and oral testimony, and the disjuncture between the on and off the record aspects of the case, played a crucial role in undermining the eyewitnesses’ testimony. The judge first dismissed Gafar’s testimony:

After reading the entire deposition of this witness in which he has mentioned the accused and mentioned that three of them are present in court today, I cannot accept his words because during his cross-examination he has admitted that he does not know the full name of the three accused and has not given any physical description of the same. In the same manner this witness has claimed that he can identify one accused but he doesn’t remember his name and after that has said that he remembers his name and his name is H, but when the accused was asked his name he has said that his name is D. Thus I am sympathetic with the defense lawyer’s submission that the deposition of the witness does not prove that the accused are connected to the offense.

The judgment is ultimately a document that privileges ‘‘the document’’ and that comes at the end of a series of other documents—FIRs, police reports, lawyers’ written and oral arguments and written and oral evidence—that precede it. This is important because the judgment largely relies on the defense’s written and oral argument, and the judge only adds at the end of a section, “on this point I am persuaded by the learned defense lawyer’s argument.” In the thicket of narratives (defense and prosecution written arguments), his voice juts out for a moment to announce “I agree” or “I disagree.”

Gafar is unable to identify the accused by his “full name” (aakha naam) but names him by his alias, which circulated outside the courtroom and was even mentioned in police documents. During the trial, the judge never pauses to investigate the mass volte-face by material witnesses (panch witnesses) who were supposedly present at the scene of the offense while the police collected on-site evidence. Such details recur in many 2002 trials, but are simply overlooked in the judgment. The witnesses claim that the police forced them to sign a blank sheet of paper or the police did not in fact take them to the scene of the offense. These statements are on record but ignored by the judge, not worthy of further investigation or interpretation.

Finally, it is not simply the contradiction between written documents and oral evidence that weakens the survivor’s testimony. The defense presents the belated documentation of the violence (a common feature of most 2002-related complaints) as a reason to doubt the credibility of the witness. The judge notes that the witness has filed a complaint with the police only “24–25 days after the event and there is no explanation for this delay.” The extraordinary threat to Muslim life and property during this time does not suffice as an “explanation” for the delay.


Faced with a hostile state and public, NGOs and activists in Gujarat encouraged survivors to approach the courts for ‘‘justice.’’ However, inside the courtroom in Gujarat, legal procedures and reasoning hollowed out substantial questions about the context of state-sanctioned violence against minorities during the trial, producing an outer shell that appeared as the ‘‘rule of law’’ but acted as a mirror that showed the Muslim witness the depth of Hindu nationalist power in Gujarat. If it is not uncommon for legal trials in the aftermath of mass violence in India21 to fail to provide accountability and justice to victims and survivors, then what exactly is at stake in studying trials that end up reinforcing violent regimes instead of undermining them? Instead of diagnosing these processes as the law’s ‘‘failure,’’ I argue that such trials are better understood as modes of disseminating and recasting Hindu supremacy as a legal form of political rule.

Given the televised, public, and state-sanctioned nature of the violence against Muslims, the Hindu nationalist government in Gujarat was under pressure to reestablish its legitimacy within established imaginaries surrounding justice and the rule of law within a democratic state regime like India. By winning local elections, the Hindu nationalist government established its popular sovereignty among the people in Gujarat, but it still lacked legitimacy within national and international publics.

Subsequently, the legal process in Gujarat, except in rare cases, became an institutionalized space for dismissing victims’ testimony and reinforcing the subordinate status of Muslims within a Hindu nationalist regime; ironically, the state was able to reinforce the hierarchy between Hindus and Muslims, only this time in the name of the rule of law and due process. With the help of the police and judicial officials, the state turned the law against the witness. These trials exemplify the paradox of using the law to help witnesses of political violence in contexts of state-sanctioned and publicly approved violence.

Legal processes ostensibly designed to punish and account for public violence are also, depending on the architecture of the trials and the actors that interpret testimony, an effective tool for majoritarian states to whitewash atrocities and further consolidate its legitimacy. The activist and civil society recourse to the law for relief and justice in Gujarat is part of a broader global trend to transform political questions into legal ones.22 However, such transformations can have unanticipated effects. Activists, NGOs, and the media displaced their hope for a space untainted by Hindu nationalist politics in India onto the legal domain. But the legal-technical process of documenting and classifying harm, producing evidence, and staging the trial became a mode for state actors to give a legal form to the subaltern status of Muslims within Gujarat.

Excerpted from the Author’s Original Research Titled “Against the Witness: Hindu Nationalism and the Law in India”. Read the full paper here.


1. Two politically competing government inquiries came to contradictory conclusions on the cause of the fire that killed the Hindu activists. In 2005, an Indian railways investigation found that the train burning was due to an “accidental fire.” In 2008, a commission of inquiry appointed by the Gujarat government concluded that the train burning was a “planned conspiracy.” In 2011, a trial court in Gujarat convicted 31 people and acquitted 63 others, including the prime conspirator, but maintained that the incident was a “pre-planned conspiracy.”

2. I have used the following human rights organization reports on the 2002 violence: Concerned Citizens Tribunal (CCT), Crime Against Humanity. Volume 1 and 2 (Mumbai: Citizens for Justice and Peace, 2002); Human Rights Watch (HRW), ‘‘We have no orders to save you’’: State Complicity and Participation in Communal Violence in Gujarat (New York: HRW, 2002); International Initiative for Justice (IIJ), Threatened Existence: A Feminist Analysis of the Genocide in Gujarat (Mumbai: IIJ, 2003). reports/iijg/2003. Accessed on March 15, 2015.

3. Satish Jha, The Indian Express, February 14, 2014 (online). On February 14, 2015, a local court in north Gujarat acquitted all the 70 accused in a 2002 related case involving the murder of 14 Muslims. Most witnesses, including the main accused turned hostile, and did not name the accused during the trial. Nine persons died before the case ended. Accessed on March 15, 2015.

4. International Human Rights and Conflict Resolution Clinic at Stanford Law School, When Justice Becomes the Victim: The Quest for Justice After the 2002 Violence in Gujarat (Stanford, CA: Human Rights Clinic, 2014). Accessed on March 15, 2015.

5. It is important to note that the Indian criminal justice system, like other common-law systems, operates upon a presumption of innocence for criminal defendants, and the burden of proof (“beyond reasonable doubt”) on the prosecution is particularly high in order to safeguard the rights of the accused. For a discussion of the challenges posed by state-sanctioned violence to the Indian criminal justice system and its relationship with international law, see Usha Ramanathan, “India and the ICC,” Journal of International Criminal Justice 3(3) (2005), 627–34.

6. Wilson, The Politics of Truth and Reconciliation in South Africa, p. 225.

7. Judicial pendency and the huge backlog of cases is a systemic and longstanding feature of the Indian legal system. Several official reports have highlighted this issue and it is a common feature in discussions around judicial reform in India. The 77th and 79th Law Commission of India Reports in 1978 and 1979 on “Delay and Arrears in Law Courts” specifically addressed this issue in the lower and upper courts. Law Commission of India. Accessed online. For a discussion of structural factors that produce delay in Indian courts, see Robert Moog, “Delays in the Indian Courts: Why the Judges Don’t Take Control,” The Justice System Journal (1992), 19–36. Moog explains how various state and nonstate actors including judges, lawyers, and litigants, use delay and deferral as a strategy in the Indian legal system.

8. Zoya Hasan, ‘‘Mass Violence and the Wheels of Indian (In)Justice,’’ in Violence and Democracy in India, Amrita Basu and Srirupa Roy (eds), (London: Seagull Books, 2006).

9. The Supreme Court of India reopened nearly 2,000 cases, out of a total 4,252 cases, which were closed by the police. It also transferred a few cases out of Gujarat on the grounds that a fair trial was not possible inside the state. It later set up a Special Investigation Team that concluded there was “no prosecutable evidence” against Mr. Narendra Modi and other state officials. However, an Amicus curiae report suggested that Mr. Modi could be prosecuted on grounds of “promoting enmity among different groups on grounds of religion.” http:// Accessed on March 15, 2015.

10. Gardiner Harris and Hari Kumar, The New York Times, India Ink New York Times (online). Accessed on March 15, 2015.

11. Stanford Human Rights Clinic, 2014, p. ii.

12. For an overview of issues relating to the lower courts in India, See Robert Moog, ‘‘The Significance of Lower Courts in the Judicial Process,’’ in Veena Das (ed.), The Oxford Companion to Sociology and Social Anthropology, Vol. II (New Delhi: Oxford University Press, 2003), pp. 1389–1412.

13. Vrinda Grover, ‘‘The Elusive Quest for Justice,’’ in Gujarat: The Making of a Tragedy, Siddhartha Varadarajan (ed.) (New Delhi: Penguin Publications, 2002), pp. 355–89, 384.

14. The problem of witnesses turning hostile is a structural feature of the Indian criminal justice system, and is certainly not unique to cases related to state-sanctioned violence or Gujarat 2002. See Pratiksha Baxi, Public Secrets of Law: Rape Trials in India (New Delhi: Oxford University Press, 2014) and Daniela Berti, “Hostile Witnesses, Judicial Interactions and Out-of-Court Narratives in a North Indian District Court,” Contributions to Indian Sociology 44(3) (2010), 235–63. Based on field work in different courts in India, both studies analyze the phenomenon of witnesses changing and denying their statements during the trial as effects and signs of the power of out-of-court structures and actors to fundamentally shape legal outcomes.

15. Kirsten Hastrup, “Violence, Suffering and Human Rights Anthropological Reflections,” Anthropological Theory 3(3) (2003), 309-23.

16. Robert P. Burns, A Theory of the Trial (Princeton, NJ: Princeton University Press. 1999), pp. 9–10.

17. Most JF lawyers were Muslims, who only worked part-time for the NGO. Even though in legal meetings, the lawyers discussed legal matters (like legal judgments pertaining to the case or the evidence ‘‘on record’’), various political and social factors impinged on their handling of 2002-related cases, which they were unwilling to discuss for fear of speaking like ‘‘activists.’’ For instance, the political dominance of the Hindu nationalists in Gujarat and practical considerations of sustaining their private legal practice in Hindu-majority lower courts played a key, but unacknowledged, role in their legal interventions.

18. The report mentions the names of the accused, including four who have aliases.

19. Baxi, Public Secrets of Law: Rape Trials in India.

20. Kristin Bumiller, In an Abusive State: How Neoliberalism Appropriated the Feminist Movement against Sexual Violence (Durham, NC: Duke University Press, 2009).

21. Surabhi Chopra and Prita Jha, On their Watch: Mass Violence and State Apathy in India (New Delhi: Three Essays, 2014).

22. Antina Von Schnitzler, “Performing Dignity: Human Rights, Citizenship, and the TechnoPolitics of Law in South Africa,” American Ethnologist 41(2) (2014), 336–50.