The challenges of long-delayed prosecutions in fighting impunity in Bangladesh

Abstract This article focuses on the challenges of ‘long-delayed’ prosecutions, that is, criminal prosecutions that begin decades after the conflict, using the experience of the International Criminal Tribunal for Bangladesh (ICT-BD) as a case study. This issue is still an insufficiently discussed topic even though such prosecutions are likely to become more common in the future. The focus of this article is mainly on the legal and broader, transitional justice challenges of long-delayed prosecutions at the ICT-BD. The article examines how such prosecutions have had a contradictory, two-fold effect: on the one hand, they have partially broken the endemic culture of impunity that was allowed to prevail for decades in Bangladesh. On the other hand, however, they have been highly controversial and may have served to deepen alienation of the Islamist opposition in Bangladesh. The article concludes that the question of whether long-delayed prosecutions are desirable for a particular society remains highly context-dependent and, in some cases, mechanisms other than criminal trials may be better suited to dealing with the past.


Introduction
This article focuses on the challenges of long-delayed prosecutions, using the experience of the ICT-BD as a case study. 1 The issues of long delayed prosecutions and old evidence are still an 'insufficiently discussed topic' in the criminal justice literature. 2 The ICT-BD was established in 2010 to address crimes that were committed in 1971, almost 40 years earlier. The ICT-BD, therefore, belongs to a category of post-conflict criminal tribunals that are engaged in conducting 'long-delayed' prosecutions, that is, criminal prosecutions that begin decades after the conflict. 3 By December 2020, the ICT-BD had delivered trial judgments in 41 cases relating to crimes against humanity and war crimes committed during the 1971-conflict. Thirty-nine of these were appealed to the Appellate Division of Supreme Court of Bangladesh. The ICT-BD convicted 95 of the 105 accused who came before it. 4 It handed down 65 death sentences and 24 life sentences. Other sentences imposed included imprisonment for periods between 20 and 90 years. 5 This tribunal is a controversial domestic criminal court, with some commentators insisting that it was established 'to end the culture of impunity and to serve justice to the victims of 1971', 6 while others hold that it is politically motivated, and fraught with problems concerning due process and procedural fairness. 7 This article will largely sidestep those controversies. 8 It will focus mainly on the use of long-delayed prosecutions and old evidence in the fight against impunity in Bangladesh. It is considered that, while the circumstances of Bangladesh are highly specific, 9 it remains important to examine this case study for what it can tell us about the challenges of long-delayed prosecutions and old evidence more generally.
As greater focus is placed on fighting impunity around the world, and on legal and historical reckoning with past atrocities, long-delayed prosecutions may become more common in the future. 10 As one ICT-BD prosecutor put it, the establishment of the ICT-BD in Bangladesh has opened the door for the possibility of accountability in other South Asian countries and more broadly. One day, the political leaders of such countries may find the political will to try perpetrators of mass atrocities even after a long delay. 11 Indeed, it is possible that such prosecutions may become more common: as the pursuit of individual accountability for such crimes becomes a norm, rather than an exception, with societies increasingly willing and able to investigate atrocities perpetrated in their past. 12 While cases of long-delayed prosecutions share certain similarities, every situation is also contextdependent. For instance, with respect to prosecutions of Holocaust-era crimes, 13 those prosecutions, even if long delayed, occur in the context of the facts of the Holocaust having become facts of common knowledge -'indisputable historical fact[s]' 14where the roles of the different parties have previously been juridically established. The situation is different with respect to the 1971-conflict in Bangladesh where, even though thousands of trials were initiated in the immediate aftermath of that conflict under the 1972 Bangladesh Collaborators (Special 4 A number of accused persons passed away before their cases were completed. 5 T. Huq, 'Half of Bangladesh War Crimes Convicts Are Fugitive from Justice', Bdnews24.Com, 16 December 2020, available at www.bdnews24.com/bangladesh/2020/12/16/half-of-bangladesh-war-crimes-convicts-are-fugitive-from-justice. 6 S. Hosen, 'What Lessons May Be Learnt from the Operation of the ICT-BD in the Areas of International Criminal Law and Transitional Justice?' (2020) (PhD Thesis, Anglia Ruskin University, Cambridge, available at www.arro.anglia.ac.uk/id/eprint/ 706744/1/Hosen_2019.pdf), at 131. Tribunal) Order (the 1972 Collaborators Order), very few records of those trials have survived. 15 To a large extent, therefore, the facts of the 1971-conflict had to be juridically established anew by the ICT-BD in 2010. However, as will be discussed below, with respect to important aspects of that conflict, the tribunal was all too ready to treat them as facts of common knowledge. 16 Because of these controversies, and also other reasons such as language barriers, the work of the ICT-BD has remained relatively overlooked in the transitional justice and international criminal law literature. With some notable exceptions, such as a 2012 anthology on old evidence, 17 as Beringmeier has noted, the ICT-BD has been absent from most discussions on international criminal justice. 18 In this respect, a Bangladeshi NGO officer noted that this important issue 'had not been talked about enough'. 19 This lack of attention to the work of the ICT-BD reflects, in part, a broader disinterest in the 'circumstances surrounding the violent birth of Bangladesh in 1971'. 20 Debnath has made the point that Bangladesh, financially destitute and on the margins of geopolitics, appears to have 'aroused little interest in any discipline except perhaps development studies'. 21 And in the political arena, an ICT-BD prosecutor commented that, as a poor country, Bangladesh was marginal to the interests of the international community and some actors would have liked to 'erase' the 1971-conflict out of history if they could. 22 As a result, as the South Asia specialist, D'Costa, has held, the war of 1971 remains 'one of the most under-researched conflicts in the world, and the traumatic experiences of the civilians after the war remain virtually unknown despite growing interest in nationalism and ethnic violence'. 23 This article is organized in six sections. The next section will briefly set out the methodology of the research. It will then examine two frameworks for delayed prosecutions in the transitional justice literature. This is followed by a brief historical overview of the 1971-conflict and the 2010 trials in Bangladesh. The article will then briefly touch upon some of the controversies around the 2010 trials, bearing in mind, however, that this is not the main focus of the research. It will then examine some of the challenges of long-delayed prosecutions with specific reference to the experience and lessons of the ICT-BD. Finally, the article will offer some concluding remarks on the impact of the trials in Bangladesh.

Methodology
For this study, we sought the perspectives of Bangladeshi stakeholders who were familiar with the work of the ICT-BD. To this end, between July and December 2020, we conducted a small number of semi-structured, key informant interviews with respondents recruited from amongst Bangladeshi politicians, prosecution and defence lawyers, as well as officials from non-governmental organizations. 24 In order to do this, we initially relied on contacts within Bangladesh that one of the co-authors had established in the course of his doctoral research. 25 We also used 'snowballing' to recruit additional respondents. We were particularly interested in their perspectives around the challenges that delayed prosecutions posed to the workings of the ICT-BD and the role of this tribunal in fighting impunity in Bangladesh. 26 While some of these interviews were conducted in English, others were conducted in Bengali. This enabled us to reach stakeholders who may otherwise not have been able to participate because of language barriers.
Our focus on key stakeholder falls within the category of what Van der Merwe refers to as 'institutional impact' studies, that is, studies aimed at asking critical questions about the ability of justice mechanisms to contribute to institutional transformation. 27 While admittedly our pool of respondents is small, many of them, such as the Bangladeshi judges and prosecutors, would be considered 'difficult-to-reach population[s]'. 28 This is because of the restricted and insider nature of the information they may own or have access to, and their potential reluctance to take part in academic research.
As such, by making available for the first time a small but important number of perspectives from hard-to-reach key informants, this study seeks to contribute to the scholarship on the ICT-BD and the transitional justice processes in Bangladesh. We have tried to ensure balance in the viewpoints by targeting politicians from across the political spectrum, and also from both prosecution and defence lawyers. Nevertheless, this study is based on a small pool of perspectives and this limitation needs to be borne in mind when considering the findings. Moreover, while the stakeholders we interviewed were able to shed light on the more institutional and technical challenges of long-delayed prosecutions, the study had to rely on published opinion surveys to assess the broader, public impact of prosecutions in Bangladesh. 29 Another important limitation is that respondents in this study did not include victims/ survivors or victims' groups. While acknowledging the importance of this key demographic in discussions on fighting impunity in Bangladesh, given the very large number of victims/survivors emerging from the 1971-conflict, it was considered that a larger-scale study would be necessary to meaningfully capture their views. For these reasons, it is important to take into account the absence of the views of victims/survivors when considering the findings of this study, given that fighting impunity and delivering justice 'usually means different things to different people. It is a highly emotive debate among people speaking different conceptual languages'. 30 3. Long-delayed prosecutions in the literature A popular perspective in international criminal justice literature is that criminal accountability has to be 'swift'. 31 Indeed, that criminal trials should start promptly is something of 'an article of faith 24 In all, we were able to conduct eight interviews online. Although in the original plan, we had intended to travel to Dhaka, Bangladesh, to conduct further interviews in person (targeting particularly those stakeholders who were unwilling to speak using online means), unfortunately, the pandemic in 2020 disrupted these plans as international travel became impossible. 25 Hosen, supra note 6. 26 The interview protocol included questions such as: 'What were the main challenges that the ICT-BD faced on account of this almost 40-year delay? '  in post-conflict justice and international criminal law'. 32 From this perspective, in the aftermath of conflict or gross human rights violations, those suspected of atrocities should be charged, arrested, and tried expeditiously. Whiting notes that this view is considered 'so unexceptional that those who express it rarely examine it'. 33 This view is bolstered by the Rome Statute's complementarity regime, where 'undue delay' is seen as one of the key indicators of unwillingness or inability to prosecute. 34 Against this backdrop, however, there has been a steady stream of empirical research that has cautioned against 'a premature, reflexive response' to initiating immediate prosecutions in postconflict situations. 35 For instance, in their study, Fletcher et al. have argued that, rather than relying on the assumption that immediate intervention is necessary, the appropriate sequencing is to first gain a comprehensive understanding of the local context and then to ask what, whether, and when transitional justice interventions should be initiated. 36 The authors advocate a cautious, go-slow approach with respect to the deployment of criminal prosecutions and other retributive justice measures. 37 McAuliffe argues that, in some contexts, the ensuing delay may be protracted for many years without jeopardizing the success of the prosecutions and/or the transition. The author holds that criminal prosecutions undertaken many years after transition can be successful, can pursue many of the same penological justifications for criminal punishment as immediate trials, and 'can do so more often in better quality trials'. 38 While the above studies have therefore shed some light on the issue of timing of trials in transitional justice contexts, it is notable that the literature has mainly focused on 'short delays'on trials commencing within about ten years from the alleged human rights violations. This is so because, where trials have taken place after the transition, the majority generally have started within that timeframe. For instance, in their cross-national survey of transitional justice in 161 countries during 1970-2007, Olsen et al. found that trials occurred on average 4.2 years after the transition. 39 There have been fewer studies focusing specifically on the phenomenon of long-delayed prosecutionsthose beginning decades after the alleged violations. 40 This may be, in part, because of the relatively fewer cases of long-delayed prosecutions, 41 and in part, because the literature has generally not distinguished between short-and long-delayed prosecutions, treating them in the same fold. 42  Another study specifically exploring long-delayed prosecutions is Han's work on Argentina and South Korea, 'where criminal and civil justice are being pursued following several decades without legal accountability'. 44 In his study, the author argues that, notwithstanding the potential benefits emanating from bringing immediate legal prosecution and justice to past human rights abusers, in some cases, immediate or short-delayed trials may not be possible. In such cases, longdelayed justice (whether criminal or civil) may constitute a useful mechanism of transitional justice. Acknowledging that each transitional case is different and there is no one-size-fits-all solution, 45 the author discusses the strengths of long-delayed justice as a potentially useful approach, that is not meant to 'replace' other transitional justice mechanisms, but to be available as a useful mechanism in certain contexts. 46 In order to shed more light on the question of timing of trials in post-conflict contexts, the next section will discuss two frameworks for conceptualizing long-delayed trials, namely the 'Justice Delayed Is Justice Developed' and 'Better Late Than Never' frameworks.

Justice delayed is justice developed
In many transitional justice contexts, delayed trials are not only inescapable but can even be essential and beneficial to the pursuit of justice. 47 In this sense, justice delayed may be justice developed. In some situations, delayed trials are simply inevitable. In post-conflict contexts, governments are often in dire economic situations, lacking resources to spend on dealing with the past, especially for conducting expensive and time-consuming legal trials. 48 However, delayed prosecution may also be pursued as a deliberate policy choice, because it is desirable. Several scholars, including Snyder and Vinjamuri, 49 Paris, 50 and McAuliffe, 51 have developed normative theories demonstrating that, in some post-conflict contexts, delaying prosecutions may be desirable in order to allow for democracy to develop in a stable and sustainable fashion. 52 Han argues that delayed justice enables reform and the strengthening of civil society and governance before bringing legal justice. 53 Delayed prosecutions thus provide a society with time to improve its criminal justice system, restock its members, and institutionalize legal rights necessary in modern legal system and trials. 54 These processes should enable a certain level of stability to take hold which, as Whiting notes, 'is an essential precondition to successful prosecutions, which in turn promote long-term stability and peace'. 55 Delayed prosecutions may also allow the space and time for reconciliation and healing to take place. The passage of time may enable, as Chapman argues, painful memories to recede and a new generation, without direct experience of the hostilities and hence more open to new kinds of relationships, to emerge. 56 Immediately after the passions of war, it may be asking too much to expect  Han, supra note 44, at 654. 47 Whiting, supra note 9, at 326. 48 Han, supra note 44, at 677.  McAuliffe, supra note 31. 52 Han, supra note 44, at 694; McAuliffe, supra note 31, at 299. On the issue of overturning amnesties granted in this initial period see McAuliffe, ibid., at 308. 53 Han, ibid., at 675. 54 Ibid., at 676. 55 Whiting, supra note 9, at 329. 56 A. Chapman, 'Approaches to Studying Reconciliation', in Van der Merwe, Baxter and Chapman, supra note 27, at 151. criminal prosecutions of mass human rights violations. 57 However, delayed prosecutions can postpone what are potentially very divisive trials until society has 'sufficiently stabilized to absorb the divisional impacts of such trials'. 58

Better late than never
In addition to the above institutional perspectives, it is also possible to consider the question from the perspective of the victims/survivors. While they would generally prefer swift justice, there could be several reasons (political, economic, etc.) for which justice may have to be delayed. And, in situations where victims/survivors have had to endure several decades of impunity, they may consider it better late than never to have justice, even after a long delay.
This perspective is based on the premise that 'people do not forget'. 59 This may be why amnesties, governmental indifference and the subsequent emergence of other pressing social issues rarely dampen the need of victims' groups and civil society to campaign for truth and justice for human rights abuses committed in the past. 60 Indeed, even where, over time, internal and external political pressure to pursue prosecutions may dwindle, victims' groups and civil society have continued to demand accountability.
In Cambodia, for instance, victims' surveys showed that people supported the initiation of criminal prosecutions against the Khmer Rouge leadership in spite of the long delay. 61 And experience across several countries has shown that the desire for justice does not die over time. A timelag can be beneficial in allowing victims to 'process their experiences' before renewing the pursuit of justice. 62 According to McAuliffe, societies may be willing to postpone accountability for human rights abuses if it permits the cessation of conflict or authoritarianism, but they never forego it. 63 Indeed, Han notes that, in some cases where prosecutions for gross human rights violations have not occurred, while there may be an appearance of stability and unity on the surface after several decades, this does not mean that the issue is 'not boiling underneath'. 64 Writing in relation to both criminal and civil justice, the author posits that: Argentine and South Korean examples prove that one cannot escape from confronting the past: the choice is either you confront the past or have the past confront you. A theme of "better late than never" is equally applicable. 65 Studies in other countries where trials have not taken place several decades after the alleged atrocities suggest that the passage of time does not dampen the victims'/survivors' yearning for accountability. For instance, when the Indonesian government appeared unable or unwilling to pursue accountability for the 1965 crimes, survivors' associations established the International People's Tribunal 1965, an informal tribunal that sought to achieve symbolic accountability. McAuliffe, supra note 31, at 310. 63 Ibid., at 311. 64 Han, supra note 44, at 693. 65 Ibid. Indeed, the growing number of people's tribunals tend to reinforce the point that people do not easily forget such atrocities, even when governments may prefer to 'move on'.
In this respect, long-delayed prosecutions may be regarded as 'better late than never', because, although late, they may still contribute to the process of justice and fighting impunity. This needs to be qualified, however, in light of the counter-argument that such trials may serve to reopen old wounds. A study by Dancy and Wiebelhaus-Brahm found that the length of delay could matter. The authors set out to evaluate whether those transitions where trials began within five years of democratization were more or less likely to result in successful democratic consolidation. They found that quickly-initiated trials (within five years) tended to promote successful democratic transition. 67 Conversely, trials that began after that timeframe tended to be 'less consistent'. 68 This would thus suggest that context and the length of delay is important when considering the potential impact of long-delayed prosecutions.
The above discussion has examined some frameworks for thinking about long-delayed prosecutions. However, while such prosecutions could be useful in appropriate contexts, they must not be a political excuse for the endless avoidance of bringing legal justice. 69 Moreover, they may present significant challenges, an issue which will be considered later in the context of the ICT-BD. 70 The next part, however, will provide a snapshot of the events around the 1971-conflict and the 2010 trials in Bangladesh.

A historical overview of the 1971-conflict and its aftermath
The historical background of the 1971-conflict in East Pakistan (now Bangladesh) has been welldocumented in the literature. 71 The conflict was a complex and extremely violent one, deeply rooted in historic antagonisms and steeped in the politics of the Cold War. 72 As Debnath points out, there were several currents of violence underway: a murderous state terror campaign against perceived supporters of Bengali independence (especially Hindus, who began fleeing the province in droves), a fledgling civil war between government troops and Bengali nationalists, and clashes between Bengalis and "non-Bengalis", apparently arising from pre-existing ethnic tensions. 73 Thus, the depiction of the 1971-conflict solely as a persecution by the Pakistani army against Bengalis fails to account for the atrocities committed by Bengalis against each other and against non-Bengalis. For instance, Bengali freedom fighters were responsible for targeting groups that were collaborating with the Pakistani army, or were merely suspected of doing somost notably the ethnic group of Biharis (who were refugees from India at the time of Partition, and who 67 Dancy and Wiebelhaus-Brahm, supra note 37, at 336. 68 Ibid. 69 Han, supra note 44, at 700. 70 Ibid., at 654. 71 See, for instance, Beringmeier, supra note 16, at 21; Debnath, supra note 21; Hosen, supra note 6; Hossain, supra note 15. opposed independence and wanted to remain within a united Pakistan). 74 However, while recognizing that different parties were responsible for committing atrocities, 75 the sheer brutality of the violence committed by the Pakistani armed forces, together with their local collaborators, against pro-liberation Bengalis stands out. 76 These atrocities included the mass killings of many hundreds of thousands of Bengalis in East Pakistan, accompanied by widespread use of torture and rapes, which led to the exodus of millions of refugees. 77 What stands out, according to Robertson, is the full-on barbarity of 'Operation Searchlight', in which: a monstrous regiment, with the latest military hardware, emerged behemoth-like from its barracks to kill the poor and burn down their houses and then to exterminate the intelligentsia. That was the beginning of the war. At the end, a few days before Pakistan's foreseeable surrender, came the most spiteful killingsof the professionals, teachers and community leaders who might have made a contribution to the nascent state of Bangladesh. There was genocide too, aimed at extinguishing or extirpating the large minority (ten million) Hindu population. 78 The actions of the Pakistani armed forces and their local collaborators went far beyond any conceivable defence of military necessity. 79 For instance, the Pakistani military undertook genocidal attacks on Hindus that forced millions of them to flee the country. 80 These groups were targeted partly because of their ethnicity/religion and partly 'for exercising a democratic choice to support the Awami League'. 81 There were four phases of violence. The first phase, beginning in March 1971, consisted in the above-mentioned 'Operation Searchlight', led by the Pakistani army and local collaborators, wherein several atrocities (such as deliberate targeting and killing of civilians) were committed. 82 In the second phase, the Pakistani army deliberately targeted individuals who supported (or where suspected of supporting) the Awami League Party, the pro-independence political party. Another feature of this second phase was the commission of widescale rape and sexual violence. 83 76 Ibid., at 399; D'Costa observes: 'Exactly how many people were killed in 1971 in the territory of Bangladesh is not known. On the higher side is the Bangladesh government's figure of three million, proclaimed quickly after the war and still accepted largely without question across the country. On the lower side is the Pakistani government's claim of 26,000 fatalities, a figure recognized by most as absurd. A number of scholars and analysts have estimated that one million is closer to the mark': see D'Costa, supra note 7, at 357. 77 Robertson, supra note 74, at 9. 78 M. Amir-Ul Islam, 'Towards the Prosecution of Core International Crimes before the International Crimes Tribunal', in Bergsmo and Wui Ling, supra note 2, at 217; Robertson, ibid., at 9. See also Travis, supra note 73, at 424, who notes that '[t]he West Pakistanis attacked Bengal, destroying about 24 city blocks in Dacca in the first two weeks of the attacks. The Pakistani air force bombed the main hospital in Dacca, killing most patients there. Federal machine guns mowed down groups of East Pakistanis. The army burned down and shelled countless homes and factories : : : At Dacca University, hundreds of students were lined up against walls and gunned down. British expats being evacuated to India reported that the army was killing every Bengali on sight. Pakistani firing squads targeted Bengali intellectuals and political leaders for death : : : According to several estimates, between tens and hundreds of thousands of Bengali women were raped. Some suffered nightly from multiple soldiers. Six hundred abducted women and girls were found in one Dacca camp abandoned by the Pakistanis. Federal forces and allied militia impregnated between 25,000 and 70,000 Bengali women against their will'. 79 Robertson, supra note 74, at 32. 80 Ibid. 81 Ibid. rounded up and arrested hundreds of Bangladeshi intellectuals, including doctors, engineers, professors and journalists, and tortured them to deathwith their bodies being found dismembered and mutilated. 84 In the fourth phasethe post-surrender periodthere were reports of revenge attacks by pro-liberation Bengalis against local collaborators and members of the paramilitary groups (Razakars), while mobs also targeted the Bihari community. 85 The 1971-conflict took place against the backdrop of the Cold War and, following the opening of some national archives, scholars have been able to shed more light on the political contexts and responses of foreign governments, including Canada, 86 the United Kingdom 87 and the United States. 88 Debnath notes how different actors, depending on their agenda, tended to frame the 1971-armed conflict either as a civil war in a sovereign state, which was therefore not an issue of international concern, or as a struggle for self-determination against a genocidal regime. 89 Despite knowledge of the atrocities in Bangladesh, for instance, the Canadian government chose to adopt a policy based upon public neutrality. According to Pilkington, this approach served to protect Canada's relationship with Pakistan, deemed desirable in terms of national interest, and maintained Canadian neutrality with regard to a foreign secessionist issue that might have stirred unwelcome comparisons with its own separatist debate over Quebec. 90 Similarly, the United Kingdom officially framed 1971-conflict as a civil war, enabling the British government 'to remain neutral and not interfere in Pakistan's internal affairs : : : '. 91 Moreover, amidst the mounting violence in East Pakistan, the United Nations Security Council largely remained silentparalysed by the tension between Soviet-backed India and a Pakistan firmly supported by the US administration. 92 The silence at the Security Council was finally broken on 3 December 1971, when Indian military jets began bombing both wings of Pakistan. 93 It was only after Pakistan surrendered to Indian forces on 16 December 1971, that the Security Council finally managed to issue a resolution calling for a durable ceasefire and the retreat of all armed forces to their own territories. 94 However, the international community's ostensible concern was short lived. Its lack of interest, support and, in some cases, direct opposition to calls for justice from Bangladesh, as discussed below, has been described as a 'failure of global justice'. 95 Reflecting on this failure, Bass refers to the case of Bangladesh: 84 See also Amir-Ul Islam, supra note 78, at 217; Robertson, supra note 74, at 9. See also Travis, supra note 73, at 424, who notes that '[t]he West Pakistanis attacked Bengal, destroying about 24 city blocks in Dacca in the first two weeks of the attacks. The Pakistani air force bombed the main hospital in Dacca, killing most patients there. Federal machine guns mowed down groups of East Pakistanis. The army burned down and shelled countless homes and factories : : : At Dacca University, hundreds of students were lined up against walls and gunned down. British expats being evacuated to India reported that the army was killing every Bengali on sight. Pakistani firing squads targeted Bengali intellectuals and political leaders for death : : : According to several estimates, between tens and hundreds of thousands of Bengali women were raped. Some suffered nightly from multiple soldiers. Six hundred abducted women and girls were found in one Dacca camp abandoned by the Pakistanis. Federal forces and allied militia impregnated between 25,000 and 70,000 Bengali women against their will'.  Pilkington, supra note 86, at 451. 91 Debnath, supra note 21, at 18. 92 Ibid., at 66. 93 Ibid., at 68. 94 United Nations Security Council, Resolution 307, UN Doc. S/RES/307 (1971). 95 Travis, supra note 73, at 442. as an illustrative case of the political processes by which the demands of international security can trump the prosecution of war criminals. In this important case, peace-making ultimately proved more important than accountability. At the same time, the long, uneasy aftermath of that eclipse of justice suggests that even when amnesty is necessary for the pursuit of peace, it can leave a toxic legacy for future politics. 96 After the ceasefire, the 'Father of the Nation' of Bangladesh, Sheikh Mujibur Rahman, returned to Dhaka from West Pakistan (where he had been detained), and immediately began to call for the prosecution of war criminals. 97 Without international support, however, Bangladesh had few options available for pursuing accountability. 98 At the end of the conflict, India had held approximately 92,000 Pakistani prisoners of war, 195 of whom had been suspected of international crimes such genocide, crimes against humanity and/or war crimes. 99 However, under pressure from Pakistan's Western allies and Islamic states (headed by the strong lobby of Saudi Arabia), and ultimately on the assurance of Zulfikar Ali Bhutto (the then Prime Minister of Pakistan) that he would ensure the trial of those 195 prisoners of war in Pakistan, the prisoners were returned to Pakistan. 100 In 1971, Bhutto set up the Hamoodur Rehman Commission of Inquiry to investigate the army's defeat, which issued a report blasting the Pakistani military's corruption and brutality. This Commission urged the government to create a high-powered court or commission of inquiry to: hold trials of those who indulged in these atrocities, brought a bad name to the Pakistan Army and alienated the sympathies of the local population by their acts of wanton cruelty and immorality against our own people. 101 However, Pakistan reneged on its promises to try the 195 suspects and no Pakistani trials ever happened. This was, after all, the 'era of impunity' 102 and as Islam has noted, a 'complex interplay of national and international politics' prevented the early pursuit of justice both in Pakistan and, as will be discussed, in post-independence Bangladesh. 103 In Pakistan, on the contrary, some generals suspected of the worst excesses went straight back into government and, when they died as 'innocent' persons, they were given 'a State funeral with full military honours'. 104 As to the Commission of Inquiry's report, this was so scathing that it was suppressed for decades, until it was enterprisingly published in India by India Today in 2000 and in Pakistan by Dawn in 2001. 105 While, therefore, Pakistan's assurances that it would prosecute the 195 Pakistani suspects rang hollow, the fledgling Bangladeshi government focused attention on prosecuting suspects present in its own territory. The country enacted new laws and effected constitutional amendments to pave the way for such prosecutions. On 24 January 1972, Bangladesh enacted the 1972 96 G. Bass, 'Bargaining Away Justice: India, Pakistan, and the International Politics of Impunity for the Bangladesh Genocide', (2016) 41 International Security 140, at 144. 97 Beringmeier, supra note 16, at 39. 98 Bass, supra note 96, at 180.  100 Beringmeier, supra note 16, at 47; ICT-BD Molla judgment, ibid., at 70. 101 Bass, supra note 96, at 156. 102 Robertson, supra note 74, at 11. See 104 Robertson, supra note 74, at 32. 105 Bass, supra note 96, at 156.
Collaborators Order for the purpose of bringing local collaborators of the Pakistani armed forces to trial. 106 Under this Order, 40,000 people were investigated, 20,000 were charged and taken into custody and less than a thousand people were convicted (with the right of appeal). 107 However, the strain on the makeshift judicial system was too great, and in 1973 amnesties were announced for all collaboration crimes except rape, murder, and arson. 108 Moreover, early in 1972, Sheikh Mujibur Rahman asked for an international tribunal to be set up to try war criminals. 109 While the idea received some support, there was not enough support to implement such a project. 110 As a consequence, and in the face of strong international opposition, 111 Mujibur Rahman decided that his only option was to establish a domestic criminal tribunal. As a result, on 20 July 1973, the International Crimes (Tribunals) Act, 1973 (1973 ICT Act), was enacted. The draft text of the 1973 ICT Act was developed in consultation with several international jurists. 112 The tribunal established by the 1973 ICT Act was given jurisdiction over crimes against humanity, crimes against peace, genocide, war crimes, violations of the Geneva Conventions 1949, and any other crimes under international law. Unlike the 1972 Collaborators Order, which focused on local collaborators, the 1973 ICT Act initially focused on members of armed, defence or auxiliary forces. 113 These enactments were accompanied by several changes to the constitution, aimed at precluding the possibility of declaring any law relating to the trial of war criminals as unconstitutional, and which also denied to the accused before the above tribunal certain constitutional guarantees available to accused persons in ordinary criminal trials, such as the right to move the Supreme Court. 114 Changes to the constitution also banned the religious party Jamaat-e-lslami, whose members had actively collaborated with the Pakistani armed forces during the 1971-conflict. 115 Finally, in 1973, the Bangladeshi Parliament proceeded to indemnify the actions of freedom fightersthat is, those who had fought for Bangladeshi independence in the 1971-conflict and its aftermath. 116 By virtue of the Bangladesh National Liberation Struggle (Indemnity) Order, 1973, therefore, any case connected to the actions of freedom fighters in the 1971-conflict was barred from prosecution. 117 Efforts to prosecute crimes of the 1971-conflict came to an abrupt end after the assassination of Mujibur Rahman, and the ouster of his Awami League government by military coup on 15 August 1975. 118 This signalled the start of what Hossain refers to as an 'endemic culture of impunity' and what a politician has referred to as a 'long winter sleep' in Bangladesh. 119 The military regime that 106 Beringmeier, supra note 16, at 40.  110 Beringmeier, supra note 16, at 39, 40. In this context, Triffterer states: 'I remember well that around 1971 and 1972, the immediate establishing of a universal international criminal court or an ad hoc regional tribunal in the context of Bangladesh was not seriously discussed': see O. Triffterer, 'Bangladesh's Attempts to Achieve Post-War (or Transitional?) Justice in Accordance with International Legal Standards', in Bergsmo and Wui Ling, supra note 2, at 262. 111 Robertson, supra note 74, at 9.  115 Islam, supra note 103, at 301. 116 Beringmeier, supra note 16, at 44. 117 Ibid.  120 It halted all trials, released the prisoners, and rehabilitated them by restoring citizenships and allowing people back into government positions. 121 Some of these suspects were able to re-enter politics and secure Ministerial positions. 122 The Repeal Ordinance also reversed the banning of Jamaat-e-lslami, allowing it to register and function as a political party. 123 The 1973 ICT Act, however, was never repealed. 124 As a result, according to Rafiqul Islam, because there was no legal reckoning for the crimes of 1971, a culture of impunity was allowed to develop. The military retained power until 1990, when the dictatorship headed by General Hussain Muhammad Ershad fell and gave way to multi-party elections and civilian rule. 125 Since 1990, Bangladesh has alternated between governments led by the Awami League, headed by Sheikh Mujib's daughter, Sheikh Hasina, and the Bangladesh National Party (BNP), headed by the widow of Bangladesh's first military ruler, General Ziaur Rahman. 126 Openly campaigning for justice during the military dictatorship was risky. A politician we interviewed observed that a popular slogan from the period was 'Forget, forget! Don't Divide the Nation', and that people who wanted justice were accused of inciting division. 127 In this period, it was grassroots movements that helped keep calls for justice alive. 128 For instance, in 1992, the activist Jahanara Imam, who had lost her son during the 1971-conflict, established the Forum For Secular Bangladesh and Trial of War Criminals of 1971 (Ekattorer Ghatak Dalal Nirmul Committee). 129 The Forum set up a People's Court to try Ghulam Azam, a suspected war criminal who had been re-appointed as the leader of the Jamaat-e-Islami party a year earlier, through a symbolic public trial in which people from different professional backgrounds participated and some 200,000 spectators attended the event. 130 Another organization, the National Coordinating Committee for Realisation of Bangladesh Liberation War Ideals and Trials of Bangladesh War Criminals of 1971, emerged from this movement and, in 1993, the Committee formed the National People's Enquiry Commission to investigate other alleged war criminals. 131 Another important development was the establishment of the Liberation War Museum in 1996, that began archiving documents and testimonies of the conflict. 132 Then, in December 2008, Bangladeshi voters elected the Awami League government with a landslide majority. 133 One of the electoral promises of the Awami League had been to bring accountability for the war crimes committed during the 1971-conflict. 134 The ICT-BD was 120 Beringmeier, supra note 16, at 44. 121 Chopra notes that, subsequently, religious-right leaders alleged to have participated in serious violence were allowed to re-enter and hold high public office in BNP-led coalition governments: see S.  125 Chopra, supra note 121, at 212. 126 Ibid.

127
Interview with an Awami League politician (POL03) on 29 December 2020. 128 Ibid. 129 Beringmeier, supra note 16, at 51. 130 Ibid., at 50, 51. A number of Committee officials were subsequently sentenced to jail for sedition: see interview with NGO Officer (NGO02) on 30 October 2020. 131 Ibid., at 52. 132 M. Mohan, '"The Messaging Effect": Eliciting Credible Historical Evidence from Victims of Mass Crimes', in Bergsmo and Wui Ling, supra note 2, at 174. 133 Reiger, supra note 124. 134 In this respect, Paolantonio has suggested that even a failed attempt to achieve justice immediately after the transition deposits a memory of justice in people and acts as a catalyst for people's persistent normative aspirations for legal prosecution subsequently established in March 2010, and a second tribunal (ICT-2) was established in March 2012 in order to carry out trials more quickly (but has since closed down). 135 According to a BNP politician we interviewed, this tribunal was established as a domestic tribunal because 'the incidents happened in Bangladesh with the Bangladeshi people and the perpetrators are also Bangladeshi so it is appropriate to establish the tribunal domestically.' 136 According to the 1973 ICT Act, as subsequently amended, the tribunal has the 'power to try and punish any individual or group of individuals, or any member of any armed, defence or auxiliary forces, irrespective of their nationality, who commits or has committed, in the territory of Bangladesh' a wide range of conflict-related crimes. However, in practice the trials have only focused on alleged local collaborators of Bangladeshi (East Pakistani) origin, as many of the Pakistani forces suspects had fled to West Pakistan after the conflict. 137 According to an interviewee from the ICT-BD prosecution, the government's reinvocation of the 1973 ICT Act, as amended, for the ICT-BD prosecutions was highly symbolic. It provided a direct link to the earlier attempts to prosecute immediately after the 1971-conflict and, indeed, could be seen as a continuation of those attempts. It was, in the words of this interviewee, and attempt to complete 'unfinished business'. 138 Similarly, an Awami League politician noted that the establishment of the ICT-BD in 2010 was actually a 're-establishment' because the underlying law had already been in place since 1973. 139

The ICT-BD in the context of international standards
As noted in the introduction, from its establishment the ICT-BD has been the subject of significant controversy and there is a growing body of research focusing on those issues. 140 The purpose of this section is not to duplicate that research but to highlight some of the main controversies, in view of their relevance to the perceived legitimacy of the ICT-BD and, consequently, to its ability to fight impunity.
The literature on the ICT-BD tends to be polarized between two camps: highly critical or generally supportive. Critics have pointed to both substantive and procedural shortcomings in the legal framework and practice of this tribunal. Chopra, for instance, notes: [d]espite the effort to update the Tribunal's foundational framework, many of its substantive as well as procedural provisions have caused concern among human rights groups. The ICT Act's inclusion of the death penalty amongst the possible punishments for crimes under its jurisdiction has drawn criticism. Human rights groups and legal monitors have also urged, inter alia, that offences under the Act be delineated more clearly, and that due process rights for the accused be enhanced. The ICT Act, as amended, did not grant suspects a right against self-incrimination or a right to legal counsel when being questioned by the police, nor did it give them adequate time to prepare a defence. Other problematic provisions include restrictions on interlocutory appeals : : : to the Supreme Court and restrictions on challenging the composition of the judicial bench. 141 There has also been rather a lot to criticize with respect to the practice of the ICT-BD. Interview with an ICT-BD prosecutor (PRO01) on 7 November 2020. 139 Interview with an Awami League politician (POL03) on 29 December 2020. 140 See, for instance, Beringmeier, supra note 16. 141 Chopra, supra note 121, at 212, 213. revealed that the presiding judge of the ICT-BD had been receiving advice from an expatriate Bangladeshi lawyer who was simultaneously also communicating with the prosecution about the same issues. 142 In those leaks, the presiding judge confided, inter alia, that he held out the prospect of a promotion if he would deliver fast verdicts. 143 And when in 2013, the Tribunal sentenced Abdul Quader Mollah to life imprisonment for crimes against humanity, the government amended the ICT-BD Act to allow prosecution appeals against sentence, and applied this power retrospectively. Following such an appeal, the Supreme Court reversed the life sentence and sentenced Mollah to death. 144 These several issues have led critics to question the legitimacy of the ICT-BD. Beringmeier opines, for instance, that 'the flawed application of the modes of liability and the elements of crimes raises severe concerns as to whether the Tribunal has really established individual accountability'. 145 On the other hand, supporters of the ICT-BD have tended to point out that, in spite of its shortcomings, the tribunal should be lauded for 'breaking the protective garb of impunity and delivering justice to many victims'. 146  Supporters of the ICT-BD have argued that, where the tribunal had to depart from international standards, this was usually necessary because the accused or their representatives had sought to 'game' the system. For instance, the tribunal had to impose retrospective restrictions on the number of witnesses that could be called, in response to the defence counsel's tactics to prolong the trials by, for instance, submitting applications for more than a thousand defence witnesses. According to the ICT-BD, this demonstrated 'an ulterior intention to somehow haul the trial and disposal of the case.' 148 Reflecting on the adherence or otherwise of domestic trials with international standards, Mégret and Samson have argued that domestic trials of mass atrocity crimes should be accorded 'a fairly high tolerance' for violations of international standards, in part because they are not human rights courts and their main objective is to strengthen the anti-impunity norm. 149 In this respect, the authors observe: [t]rials in exceptional historical circumstances and for exceptional crimes may involve a number of oddities that one would not want to associate with normal fair trials but that nonetheless hardly compromise justice entirely. 150 Mégret and Samson go on to note, however, that there is a threshold where a trial becomes one 'that is not a trial at all'in cases where there is no attempt at, or semblance of, justice. 151 142 Ibid., at 213. 143 Beringmeier, supra note 16, at 56. 144 Chopra, supra note 121, at 214. 145 Beringmeier, supra note 16, at 273. 146 Islam, supra note 103, at 317. 147 Ibid., at 306.  150 Ibid., at 584. 151 Ibid., at 585.
With respect to the ICT-BD, it is arguable that, despite its many controversies and shortcomings, that threshold has not been crossed. Even the most ardent critics of the ICT-BD in the scholarship would generally admit that the ICT-BD has attempted to adhere, at least, to some level of legal process and to provide some degree of justice. 152 Few would suggest that the ICT-BD trials are not a trial at allthat they are completely devoid, as Kirchheimer would put it, of any 'irreducible risk'. 153 As a matter of fact, the bulk of the literature in this area has tended to focus on providing key recommendations on how the legal frameworks and processes could be improved. 154 Therefore, while taking account of the critiques of the tribunal, we consider that there may still be value in studying the ICT-BD for what it can tell us about the challenges of conducting longdelayed prosecutions.

Some challenges of long-delayed prosecutions in Bangladesh
While, the ICT-BD trials of 2010 have been framed by Bangladeshi officials as a 'continuation' of the accountability efforts that took place immediately after the 1971-conflict, it is clear that the lapse of almost forty years has had an impact on the ability of such trials to fight impunity. This part explores some of the challenges arising from long-delayed prosecutions, starting with broader transitional justice challenges and then proceeding to examine some more technical, legal challenges mainly relating to old evidence.

Broader challenges of long-delayed prosecutions
While Dancy and Wiebelhaus-Brahm have shown that trials which began within five years of the transition tended to promote successful democratic transition, 155 it is not clear whether longer delayed prosecutions would have the same benefits. This is because a long delay may provide the time for suspects to rehabilitate their names, sometimes repent and reintegrate in society. In Bangladesh for instance, a politician observed that 'the accused have been living peacefully for a long time and they became normal citizens of the country, contributing towards the country's economy'. 156 Similarly, a defence lawyer noted that, in the period after the 1971-conflict, the accused had established their position in the country and people had started accepting them, they had become well known political figures and they did good things for the people of Bangladesh. 157 Indeed, the passage of time may enable suspects to gain from the proceeds of their crimes, thus strengthening their social, political or economic standing in society. According to an Awami League politician interviewed by Hosen, this is what happened in Bangladesh: [t]he properties they looted during the atrocities in 1971 were illegally earned property. A few people left the country, a few people were rehabilitated into politics and over time, their [status] rose in the Bangladeshi society and they became economically strong, these issues created some problems due to time lapses.  154 See, for instance, Beringmeier, supra note 16. 155 Dancy and Wiebelhaus-Brahm, supra note 37, at 336. 156 Hosen, supra note 6, at 144. 157 Ibid., at 147. 158 Ibid., at 143. some of the killers were given diplomatic assignments abroad and others were encouraged to form a political party and to become members of the parliament, under the patronage of this regime. 159 Over time, these individuals were able to use their power and influence to attempt to gain public acceptance. 160 There is therefore an argument that initiating long-delayed prosecutions after several decades would simply reignite politically-constructed schisms. In addition, over time, both internal and external political will to pursue prosecutions may dwindle. 161 Thus, after the passage of decades, it is possible to argue that transitional justice mechanisms other than criminal trials may be better suited to dealing with the past. There may come a point, depending on the context of a society, where 'better late than never' becomes 'better never than late' as concerns criminal prosecutions. A society may consider that it would be preferable to avoid, or postpone indefinitely, trials that reopened the past 'until the country was sufficiently reconciled to withstand the tensions that would likely result'. 162 Indeed, a fact often overlooked by those who insist on swift criminal prosecutions is that, as Mendeloff observes, '[m]any post-conflict states have carried out conscious policies of forgetting or suppressing the past and have not risked relapse into civil war'. 163 Criminal prosecutions may not always be necessary or desirable, therefore, particularly after a long delay. 164 In some countries where policies of forgetting or suppression have been pursued, a common justification has been that engaging the past would only provoke further conflict. 165 However, the countervailing view is that, unless and until criminal prosecutions of mass atrocities occur, hatreds and resentments will continue bubbling below and, in some cases, above the surface. A failure to mount trials would 'allow old wounds to fester'. 166 And, as Mulaj has noted, the ongoing 'suspicion and uneasiness diminishes the likelihood that the country moves forward to establish a shared historical chronicle of the past and develop a healthy dialogue between and within ethnic communities'. 167 In light of this, as a Bangladeshi NGO officer put it, the justice and reckoning brought by longdelayed prosecutions may be 'late, but not too late'. 168 Elaborating on this point, Amir-Ul Islam has argued that, in Bangladesh, the tragedies which occurred in the form of coups, counter-coups, killings, and assassinations, the destabilizing of the constitutional regime, the subverting of the electoral process, and the rise of extra-constitutional regimes, 'can be traced back to the failure and omission to hold trials and punish the perpetrators of the original crimes committed in Bangladesh in 1971, giving rise to an impunity culture'. 169 159 Amir-Ul Islam, supra note 78, at 229. In the decades when impunity prevailed in Bangladesh, a judge interviewed by Hosen noted that 'the victims suffered mentally as they were denied justice for a long time'. 170 They suffered not just because of the crimes committed against them, or the lack of recognition of their suffering, but also because they watched as 'people like Mir Qasim Ali and Salahuddin Quader Chowdhury became the richest people in the country' at the expense of the victims of the 1971-conflict. 171 On this point, an Awami League politician drew parallels with the Nuremberg and post-WWII trials, arguing that, if it would not have been possible to hold criminal trials immediately after WWII, Nazi leaders would have had the time to rehabilitate themselves and re-engage in political and social activities in their country. Then, if circumstances changed and it suddenly became possible to hold trials forty years later, the Nazi suspects would have argued that such long-delayed prosecutions against them had to be politically motivated. According to this interviewee, that is what happened in Bangladesh. 172 From this perspective, therefore, criminal prosecutions, even when they are long delayed, are necessary to achieve justice. Those prosecutions aim to document the untold suffering 173 and injustice of those who, as Amir-Ul Islam observes: have endured and suffered great injustice, who often have a powerful sense that what they experienced must not be forgotten, but must be cultivated both as a monument to those who did not survive and as a warning to future generations, so that a nation can be free from these crimes and atrocities; however much a government tries to bury these crimes by default, the crimes continue to haunt the nation from the debris of the history in countless ways. 174 Having considered some of the broader issues relating to long-delayed prosecutions, the next section examines some of the technical, legal issues with such prosecutions.

Legal challenges of long-delayed prosecutions
A first set of challenges that arise from long-delayed prosecutions concern the principle of legality, which requires that the conduct in question must have been criminalized by a law that was applicable to the individual at the time of the offence. 175 While this principle applies to all criminal prosecutions, it is brought into sharp relief in the case of long-delayed prosecutions, when there may be a significant evolution in the definitions of crimes between the date of the offence and that of its prosecution. The ECCC which, like the ICT-BD, was established several decades after the atrocities of the Khmer Rouge regime, had to grapple with this issue. For instance, in Duch, the Trial Chamber held that, rather than determining the state of the law at the time of prosecution, the Chamber had to: determine whether the offences and modes of participation charged in the Amended Closing Order were recognized under Cambodian or international law between 17 April 1975 and 6 January 1979. 176 170 Hosen, supra note 6, at 147. 171 Ibid. A matter that concerned both the ECCC and the ICT-BD was whether, in the 1970s, the definition of crimes against humanity required a nexus with an international armed conflict. The Nuremberg judgment had made clear that such crimes could only be committed at times of an international armed conflict. 177 However, in its analysis, the ECCC found that several subsequent instruments did not require such a nexus. 178 In the case of the ICT-BD, Robertson considers that it did not give this issue sufficient attention: [w]hat the Tribunal had to establish was that by 1971 the crime had shed the Nuremberg requirement of a connection with a war between states : : : This exercise might be satisfactorily done but this tribunal did not do it : : : . 179 Long-delayed prosecutions may also raise the prospect of time bars on prosecutions, even though the weight of international academic opinion considers that core international crimes are imprescriptible under general international law. 180 Already in 1967, the United Nations General Assembly expressed the view that it was necessary and timely to affirm in international law 'the principle that there is no period of limitation for war crimes and crimes against humanity : : : '. 181 And it is fair to say that by 1971, international crimes had become imprescriptible and the mere passage of time, even of several decades, could not serve as a bar to their prosecution.
Long-delayed prosecutions also bring with them various opportunities and challenges with respect to old evidence. Etcheson has argued that the passage of time has a contradictory, two-fold effect on evidence. On the one hand, it degrades the evidence. However, on the other hand, it allows for new evidence to come to light, for evidence to be found and documented. 182 The passage of time may also help surviving victims and witnesses to overcome their fear of denouncing the deeds of the erstwhile powerful, and may thus help bring to light more witnesses and more reliable evidence. 183 Klonowiecka-Milart lists a number of potential benefits of prosecuting crimes after a long delay, including that: [t]he passage of time makes available the established historical record. Some elements are at least ascertainedmaybe partiallyin the public conscience: the parties who stood on each of the good and bad sides, who started, or encouraged the armed violence, when did it start, and when did it end. 184 The passage of time, however, could also make the collection and evaluation of evidence significantly more difficult. In this respect, Combs notes that prosecutions that are delayed for longer periods of time are likely to feature more fact-finding impediments than those that are delayed for shorter periods of time. 185 As a result, conducting criminal prosecutions several decades after the alleged acts were committed may present significant dilemmas for lawyers. 186 Such long-delayed prosecutions may well necessitate more flexible and lenient rules of evidence, as otherwise the impediments caused by the passage of time may simply stop the prosecutions dead in their tracks. 187 In relation to the issue of old evidence, a useful framework that may be applied is that developed by Andrew Cayley, the former International Co-Prosecutor of the ECCC. Cayley developed this framework on the basis of his own extensive professional experience at the ECCC, frequently having to grapple with the challenges of old evidence in the course of his work. According to this framework, challenges relating to old evidence could be grouped in the following four, broad areas: (i) crime scenes, (ii) witnesses, (iii) documents, and (iv) expert evidence. 188

Crime scenes
Crime scenes include the places used to plan criminal acts or to conduct criminal activity, such as execution sites. The identification of such sites is important because, even though they may have become contaminated on account of the passage of time, the mere identification of the existence of these crime sites allows investigators to establish patterns of abuses so that the mass and systematic nature of these abuses becomes readily apparent. 189 However, with the passage of several decades, the locations themselves may have gotten contaminated to such an extent that, from a forensic perspective, they no longer contain any useful evidence. A mass grave that is examined a decade after the atrocity, for instance, would typically reveal less probative evidence than a mass grave that is examined three months after the atrocity. 190 Cayley refers, in this context, to a Khmer Rouge prison in Phnom Srok District which, after the fall of the regime, was rehabilitated and used as the district administration office. He notes that nowadays '[n]o traces remain of its previous use as a prison '. 191 This point is directly relevant to the situation in Bangladesh, where an ICT-BD prosecutor noted that several crime scenes, such as classrooms in a primary school which had been used as torture chambers, had not been preserved. 192 And a defence lawyer explained that identifying crime scenes for the prosecution was difficult because over time the places had changed significantly. 193 Cayley makes the point that evidence of many crime scenes in Cambodia today have survived only as memories in the minds of surviving witnesses, except where photographs and/or other documentary evidence was also available. While the investigation of crime scenes several decades after the events, therefore, presents evidentiary challenges, these are not insurmountable. What the above discussion shows, however, is the importance of documenting crime sites 'as soon as possible after the commission of the crimes, as well as at regular intervals thereafter'. 194 186

Witnesses
A long delay will also impact on the evidence provided by fact witnesses in several ways. The availability and quality of witness testimony naturally erodes over time, due to a variety of factors: [w]itnesses will die. Witnesses will forget, or their memories may become frail and unreliable with advancing age. Witnesses may also lose interest, and no longer be willing to recall traumatic events of long ago. 195 According to one ICT-BD prosecutor, many key prosecution witnesses had died by the time the trials in Bangladesh began. He argued that, in some cases, if you go through the judgments, you will find that the prosecution is not 100 per cent successful in establishing all of the charges. Some of the charges we have failed as well. And the reason for failing is the lack of papers and lack of witnesses. 196 Similarly, a defence lawyer made the point that many witnesses had died due to delay and others could not remember the events accurately during the trial because their memory faded over time. 197 Moreover, long after a crime has taken place, some potential victim-witnesses would be difficult to locate. 198 As a result, those victims' accounts may remain untold and their perpetrators unidentified, leading to accountability gaps in the retelling of the historical events. 199 In addition, the impact of trauma on memory may become more pronounced over time. This is a challenge the ICT-BD has faced, and one prosecutor noted that 'if you go through the judgments of rape and physical torture, you will see that some of our witnesses are traumatised witnesses. Because they have gone through very horrible experiences in their lives'. 200 Several scholars have investigated the impact of trauma on the reliability of witness accounts. In their review of the literature on the subject, Topiwala and Fazel found that there was 'an apparently contradictory body of evidence supporting both an impairment and enhancement of memory in normal humans subjected to stress'. 201 The passage of time may therefore impact on the quality and reliability of witness evidence, including as regards something as central as establishing the identity of the accused. 202 At the ICT-BD, for instance, issues around identity evidence arose in the trial against Abdul Quader Molla, where the Defence raised doubts over whether the accused was indeed the notorious 'Butcher of Mirpur'. 203 Moreover, political and cultural contexts prevailing at the time of the investigations and/or testimony, as well as the operation of collective memory, may intervene to shape individual recollections during and after episodes of mass violence, and their impact may be exacerbated by the passage of time. As has appeared in a number of cases in various tribunals, this may result in witnesses testifying as if they had actually seen certain events when, in fact, they had only heard about them. 204 As Cohen argues: [t]he interpretation that such a witness is lying may in fact reflect a misunderstanding of the way in which knowledge and memory are shaped and expressed in some societies. Testimony International tribunals, such as the Special Court for Sierra Leone (SCSL) and the International Criminal Tribunal for Rwanda (ICTR), routinely had to contend with these issues. 206 With the passage of time, various collective narratives that sought to make sense of the violence for the community that experienced it would emerge. And over time, members of that community may come to believe and internalize such narratives or may not appreciate the difference between having seen the events themselves and having heard about them. 207 This issue arose also at the ICT-BD trials, where a Jamaat-e-Islami politician claimed that: because of time lapse, a wrong perception was developed by the [Awami League] party and witnesses are being manipulated by the perception that developed over time and could not provide an accurate description of the events. 208 This claim was echoed by a defence lawyer who argued that, although the delay may assist with the emergence of new evidence, it may also impact on the changing perceptions of the witnesses. 209 These challenges, however, may equally affect short, as well as long, delayed prosecutions. Given that witnesses remain a key source of evidence for mass atrocity trials, the key point here is that the accounts of potential victims/witnesses should be collected and preserved as soon as possible, and the process should continue throughout the investigation. As Cayley notes, even if a witness is unavailable at the time of trial to authenticate the statement, and/or to be cross-examined regarding their statement, the court may still consider their recorded testimony as probative and useful to ascertaining the truth. 210

Documents
There are many kinds of potentially relevant documentary evidence in mass atrocity trials, including original documents from organizations or individuals involved in carrying out atrocities, contemporaneous photographs and films, as well as contemporaneous domestic and international news accounts. 211 With the passage of time, however, such documentary evidence may degrade and be lost forever. 212 The passage of time may also enable perpetrators or their associates to destroy potentially incriminating evidence. Cayley notes that individuals engaged in mass atrocity crimes often go to great lengths to destroy evidence of their culpability. They can, for instance, burn or otherwise destroy documents and other traces of their acts. 213 This point is pertinent to the situation in Bangladesh, where it has often been claimed that extensive evidence was destroyed after the conflict, particularly in the period during the military dictatorship, and is therefore no longer available. 214 The documents claimed to have been lost or destroyed in Bangladesh include reports produced by the Special Branches of the Pakistan police, case registers of trials under the 1972 Collaborators Order, official letters concerning the transfer of lands and ownership issued to 205 members of the auxiliary forces, and other documentary evidence such as contemporaneous news reports. 215 One interviewee, for instance, held that the archive of the Bangladeshi Interior Ministry was almost emptied in the period starting from 1975. 216 And with respect to newspaper records, an ICT-BD prosecutor observed that: when we visited [public] libraries for copies of old newspapers, we found that the whole newspaper is there, but the particular news regarding the auxiliary forces were cut [out], with the cut of a blade, so that no one can read it or take a record of it. 217 Similarly, a politician interviewed by Hosen claimed that when war crime suspects or their associates assumed power in the government in or after 1975, over the course of time, they destroyed a lot of significant evidence and documents relating to gross violations of human rights. This interviewee mentioned the case of Salahuddin Quader Chowdhury and stated that 'while he was in power, he destroyed a lot of documents relating to the cases which were filed against him in 1972'. 218 It is important to note that these allegations have neither been proven nor disproved independently. Nevertheless, for instance, of the thousands of trials that were conducted under the 1972 Collaborators Order, very few records have survived. 219 As a result, it is not clear how many suspects were charged, convicted and punished under the Order and for what crimes. This, in turn, has also given rise to questions around double jeopardy. 220 Moreover, even where documents have managed to survive several decades after the event, challenges may arise concerning the chain of custody of such documents. However, Cayley notes that international courts have had to accept the inescapable reality that, the longer the delay, the more risk there was that the chain of custody over old evidence would become contaminated. As a result, there have been liberal rulings on admission and probity of such old documents, often relying on other evidence to corroborate them. 221

Expert Evidence
Long-delayed prosecutions may relate to crimes that occurred in very different political, social and historical contexts, and expert witnesses may help shed more light on those contexts. Their analytical evidence can be particularly valuable in old cases, when there are only fragmentary bits of information available regarding certain events, but when there are enough fragments with which skilled analysts can reconstruct aspects of the criminal events which may help judges understand the relevant contexts. 222 Experts may also assist the court with forensically establishing aspects of the official narratives of the conflict, such as official death tolls. This is particularly important given that, with time, official memory may become deeply ingrained in a society, leading it to acquire 'sacred' status. 223 As a result, members of that society, including its judges, might find it difficult to extricate themselves from the official narratives. In this respect, experts may assist by providing independent estimates of the number of victims, on the basis of scientific analysis. This technique was used 215 Interview with an ICT-BD prosecution lawyer (PRO01) on 7 November 2020. 216 Interview with an Awami League politician (POL03) on 29 December 2020. 217 Interview with an ICT-BD prosecution lawyer (PRO01) on 7 November 2020. 218 Hosen, supra note 6, at 141. 219 Hossain, supra note 15, at 77. 220 Islam, supra note 103; Robertson, supra note 74. On the issue of double jeopardy, however, an ICT-BD prosecutor has argued that the 1972 Collaborators Order on which thousands had been prosecuted and the 1973 ICT Act, as amended, on which the ICT-BD was established, gave rise to different sets of legal offenses: Interview with an ICT-BD prosecution lawyer (PRO01) on 7 November 2020. 221 Cayley, supra note 188, at 117. 222 Ibid., at 119. 223 Klonowiecka-Milart, supra note 184, at 42. at the ECCC, where a team of independent expert demographers was engaged to carry out a focused study of the question. The demographers concluded that the death toll under the Khmer Rouge regime was between 1.7 million and 2.2 million, with between 800,000 and 1.3 million violent deaths. 224 As Cayley notes, that 'is still a substantial range of uncertainty, but it is much better than the virtually open-ended estimates that had been previously circulating among the public'. 225 Such scientific studies by independent experts, in addition to assisting the court in making more accurate findings on death tolls, will also enable it to generate more responsible historical narratives of the conflict. 226 Seeking to develop responsible historical narratives is particularly pertinent to the context of Bangladesh where successive military regimes have tried, over the course of decades, to rewrite significant portions of the history of the 1971-conflict, including in classroom textbooks. 227 According to an NGO officer, after the murder of Sheikh Mujibur Rahman in 1975: we saw a number of martial law dictators come into place and what they have done really during the course of decades is really tried to erase a significant portion of our history and that includes the history of our 1971 conflict. 228 This has led to many areas of contestation around this conflict, including with respect to the death toll. In academic sources, different numbers ranging from 500,000 to 1.7 million to 3 million deaths circulate. However, very often the sources of these differing estimates are not clear. 229 Judges at the ICT-BD have regularly made reference to the higher 'official' death toll of some three million people killed in the 1971-conflict. 230 In so doing, the ICT-BD judges have remained faithful to Bangladesh's official narrative, 231 and have severely closed down any attempts to question that number. 232 It is easy to see why the judges would have adopted this stance. The questioning of death tolls is a technique frequently used by historical revisionists, who tend to claim that the total number of victims was drastically lower than the official number. 233 However, the fact remains that the 3-million estimate was not established by a scientific study of independent experts. One factor for which the ICT-BD did not engage independent experts could have been time: such studies take time to complete and commissioning them may have hindered the expeditious pace of the trials. Given how long Bangladesh had waited for accountability, ICT-BD judges may have been keen to prioritize expeditiousness and '[i]n practice, the Tribunal has proven itself extremely quick in conducting the proceedings as well in the delivery of the judgments'. 234 However, quite apart from questions around due process that this approach raises, 235 it does not reflect a commitment to seeking truth and writing history responsibly. 236 The experience of the ICT-BD, therefore, draws attention to the importance of expert witnesses in assisting courts to study and verify official death tolls impartially, in order, on the one hand, to provide an estimate based on a scientific study and, on the other hand, to shrink the space for denial, 237 or, as Ignatieff put it, reduce the number of lies that can 'be circulated unchallenged in public discourse'. 238 7. Concluding remarks Many international commentators have called the ICT-BD trials 'a step backward' and have criticized their lack of adherence to international standards. 239 Within Bangladesh, however, these trials have been 'extremely popular'. 240 According to one opinion poll, the majority of Bangladeshi voters supported the ICT-BD trials and wanted them to proceed. 241 And, as one ICT-BD prosecutor observed, the victims of the 1971-conflict and their relatives themselves have been strongly supportive of the process. 242 Indeed, according to a politician from the BNP, victims, civil society and young people always wanted such trials to take place in Bangladesh. 243 In part, this may be because, as a judge of the ICT-BD has put it, although these trials have been very much delayed, 'the ICT-BD has taken a great step to end the culture of impunity'. 244 In his view, it was only by ending the culture of impunity resulting from the 1971-conflict that a culture of peace could truly take hold in Bangladesh. 245 It should be noted that, because of jurisdictional limitations, the ICT-BD has at best only partially addressed the impunity gap arising from the 1971-conflict. It has heard cases concerning 105 accused. 246 It has not been able to try any of the 195 Pakistani suspected war criminals. 247 Nor has it been able to try any Bangladeshi freedom fighters. Nevertheless, the support that the ICT-BD has received from the majority of Bangladeshi society seems to reflect the view that 'some justice would be better than none'. 248 In countries such as Bangladesh or Cambodia, where perpetrators have, for decades, enjoyed relative impunity for mass human rights violations, there is a view that: social, economic, and political development requires the rejection of impunity : : : through criminal trials. Impunity : : : breeds fear, distrust, and an inability to imagine a future together. 249 237 From this perspective, criminal prosecutions that lead to the arrest, detention, and eventual trial of criminal suspects who had previously enjoyed impunity, even if they are long-delayed, would be seen as 'a step forward-and perhaps should be embraced as the last chance to make such a step'. 250 But even if we accept this perspective, it is important for Bangladeshi policymakers to take heed of the countervailing view. This view, as expressed by an ICT-BD defence lawyer, posits that if the ICT-BD had been established soon after the 1971-conflict, there might not have been outrage or dispute. However, 'after 45 years, it is really strange to bring accusation[s]'. 251 This is because such trials may be perceived as politically motivated and may serve to stir up 'divisions in the Bangladeshi society by accusing people after more than three decades since the crimes were committed'. 252 Rather than inculcating a culture of peace, therefore, such long-delayed prosecutions may simply perpetuate a cycle of vengeance, something that, as Bass has observed, 'has become all too familiar in Bangladesh's politics today'. 253 These trials, happening as they are, after a long delay, may lead to deepening alienation of the Islamist opposition and hardening some of the worst factions that have divided Bangladeshi society. 254 As some commentators have cautioned, this in turn may lead to 'retaliatory proceedings when power changes hands'. 255 In order for the work of the ICT-BD to be perceived as legitimate, stronger efforts need to be made to ensure that the trials are, and are perceived to be, as fair, impartial and independent as possible. Several recommendations have been made for strengthening due process safeguards for the accused, as well as witness support and protection at the ICT-BD, and those will not be repeated here. 256 Moreover, the tribunal should be given jurisdiction over all possible sides to the 1971-conflict. At present, the ICT-BD has only been able to focus on the activities of local collaborators, as pro-independence fighters have been indemnified. The ICT-BD's single focus on local collaborators has given rise to the perception of a partisan tribunal focusing only on the losing side. It is thus open to the critique of victor's justice: those on the right side of history would be forgiven their war crimes, whilst those who fought for a united Pakistan would always be treated as traitors. 257 It is recommended that the ICT-BD should be given jurisdiction to settle questions of criminal responsibility relating to all sides of the conflict. As Robertson put it, 'there will be no peace without justice, but only justice that is two-sided, equitable and can be seen to have been done'. 258 One ICT-BD prosecutor commented that the ICT-BD should be able to consider alleged crimes of freedom fighters as well, where there is sufficient evidence to sustain an investigation. This was because, in his view, justice was a 360 degrees process and 'you have to try all the parties'. 259 But in the same vein, the group of 195 Pakistani suspects have also, so far, been able to evade justice. The international community also needs to shoulder its responsibility, therefore, to challenge Pakistan's continued amnesia about the role and brutal record of its army during the 1971-conlfict. 260 Indeed, in light of the considerable challenges arising from conducting long-delayed prosecutions, and the significant resources necessary for such prosecutions, rather than sending signals of indifference to the international community, it is crucial for Bangladesh to 250 be ready to have a genuine exchange and collaboration with its international partners. 261 On its part, the international community needs to work to counter the perception, expressed by an ICT-BD prosecutor, that 'Bangladesh was never important to the international policy makers'. 262 Having already once 'turned its collective head' on justice in Bangladesh immediately after the 1971-conflict, 263 the international community should now be more prepared to resolve key sticking points and provide much-needed resources and support for the ongoing work of the ICT-BD. Such support could include providing resources for a victim/witness protection programme (which in its current state has been described as 'very poor'), 264 supporting documentation activities, 265 assisting the tribunal in gaining access to documents, including archival materials through international co-operation, 266 and supporting training and outreach. Closer engagement between Bangladesh and the international community aimed at addressing the problem of impunity in this country is a duty owed to the victims and families of the 1971-conflict.
In the final analysis, in light of the endemic and destabilizing culture of impunity that was allowed to prevail for decades in Bangladesh, it is perhaps not surprising that, in their electoral manifesto for the 2008 election, the Awami League prioritized criminal trials over other transitional justice measures. Initiating prosecutions almost 40 years after the events was always going to be challenging. However, as this article has discussed, while such challenges may be significant, they are not necessarily insurmountable. Indeed, as the above opinion surveys of Bangladeshi society have shown, the ICT-BD trials have proved popular domestically and have been seen, at least, as partially successful in breaking the culture of impunity.
However, the ICT-BD trials have also been highly controversial and, to some extent, may have served to deepen alienation of the Islamist opposition in Bangladesh. A key question that Bangladeshi policymakers could consider at this stage, therefore, is whether the time has come to adopt other forms of transitional justice mechanisms to supplant or supplement the ICT-BD. 267 Indeed, as one NGO officer put it, 'it isn't just about the trials : : : There are issues that : : : still need to be addressed which haven't been addressed'. 268 Bangladeshi society is divided not only due to political and social allegiances and alliances, but also along class and generational lines. According to D'Costa, these divisions: cannot be mitigated only by a legal process and there remains a need for a broader reconciliatory approach towards healing the nation. Bangladesh remains a case of international interest. How and whether the country is able to achieve this ambitious goal will offer import lessons to transitional justice scholars and practitioners alike. 269 Such a broader reconciliatory approach could potentially include the establishment of a Truth and Reconciliation Commission (TRC) for Bangladesh. Such a TRC, which could potentially include 261 D'Costa, supra note 7, at 364. 262 Interview with an ICT-BD prosecution lawyer (PRO01) on 7 November 2020. 263 The international community's indifference, lack of support and, in some cases, direct opposition to criminal prosecutions immediately after the 1971-conflict has been described as a 'failure of global justice': see Travis, supra note 73, at 442. 264 Interview with an Awami League politician (POL03) on 29 December 2020. 265 Etcheson, supra note 43, at 66. 266 Klonowiecka-Milart, supra note 184, at 56. 267 Criminal trials have been favoured by the Bangladeshi government. However, some other forms of transitional justice have also been implemented, such as memorialization. For instance, the 26th of March is remembered as 'Independence Day', the 14th of December as 'Martyred Intellectuals Day' and the 16th of December as 'Victory Day'. There are also several monuments memorializing the events of the 1971-conflict across Bangladesh. Moreover, the Liberation War Museum also continues to play an important role in collecting, researching and disseminating information on the1971-conflict. 268 Interview with an NGO Officer (NGO02) on 30 October 2020. 269 D'Costa, supra note 7, at 366.