7 The Limits of the Criminal Law

In the last decade, numerous studies have analysed the genocide courts and increasingly questioned the effectiveness of judicial responses to mass atrocities, and crimes against international humanitarian law. Ironically, their appearance coincides with the creation of the first permanent International Criminal Court at The Hague. As noted in an earlier chapter, for many advocates of the recent ad hoc UN tribunals for the Balkans and Rwanda, and the hybrid tribunals for Sierra Leone, Cambodia, East Timor, and Lebanon, as well as the ICC, the development of such transnational institutions is the fulfilment of Immanuel Kant's dream of ‘cosmopolitan justice’. Cosmopolitan justice would abolish sovereign immunity for crimes and would hold the political and military elite accountable for atrocities against their own citizens, and for aggression against their neighbours. At the end of the Second World War, the Nuremberg trials and the International Military Tribunals for the Far East were colossal steps in this direction, and successfully prosecuted German and Japanese leaders for the supreme crime of ‘making aggressive war’, and crimes against humanity (Ratner and Abrams 2001). The UN provided an ongoing institutional home for the preservation of the hope for transnational justice throughout the Cold War. The 1948 Convention on the Prevention and Punishment of the Crime of Genocide was the single most important achievement in this regard, followed by the Universal Declaration of Human Rights in the same year. For many, the creation of the ICC following the decline of communism gave the world a permanent Nuremberg-like solution to genocide, war crimes, and crimes against humanity (Schabas 2007). How confident should we be that a legal response to such conflicts is the optimal one? In this chapter I examine some important ‘report-cards’ on the recent UN and hybrid courts. To put them in perspective, I review a key historical trial—Auschwitz—that has recently attracted a great deal of attention among historians and legal scholars.

At the end of the Second World War, numerous former Nazis were extradited to Eastern Europe to face trials for wartime atrocities. Within Germany, tens of thousands of former Nazis were ‘lustrated’, i.e. stripped of political rights to neutralize their threat to the emerging democratic institutions. The National Socialist ideology was subsequently discredited. However, there were complaints throughout the 1950s that concentration camp guards remained free in West German society, and their crimes in Poland were being ignored. This story was researched independently by two historians, Rebecca Wittman (2005) and Devon O. Pendas (2006). The Auschwitz trial was the most dramatic, and well-publicized post-war Nazi trial to be convened in the Federal Republic of Germany. Unlike Nuremberg, which was created, prosecuted, and adjudicated by the victors, this was a trial held under German law, prosecuted and tried by Germans. Because the events of the war pre-dated the 1948 Genocide Convention, persons charged with the mass murders, tortures and other indignities that occurred in Nazi-occupied Poland—the ‘general gouvernement’—were tried under ordinary criminal law. Twenty-two persons were indicted before the Frankfurt criminal court. They represented a cross-section of the camp's administrative units, and ranged in rank from Major to Private, and included a single kapo (inmate-guard). The case arose from a local initiative in the Stuttgart prosecutors’ office based on allegations from a petty career criminal who identified a local citizen, Wilhelm Boger, as a notorious torturer at the Auschwitz concentration camp. The case may have been dropped were it not for the involvement of the International Auschwitz Committee (IAC), a network of survivors who supplied authorities with the names of suspected war criminals and witnesses who might offer testimony to prosecute them. Despite the prickly interaction between the IAC and German prosecutors, the case eventually made it to the office of Hessian Attorney General, Fritz Bauer. He secured jurisdiction from superior courts to bring the matter before the Frankfurt court, and gathered sufficient evidence to initiate a preliminary investigation under Judge Heinz Düx. Düx interviewed over 1 500 witnesses, and an indictment was prepared against the twenty-two suspects (two escaped trial for medial reasons). The charges included murder. The crimes consisted of selecting persons on the ramp at Auschwitz for immediate gassing or enslavement for forced labour designed to bring about death through exhaustion and maltreatment. Other indictments dealt with the killing of hospital inmates by lethal injections of phenol. There were also periodic purges of units of workers within the camps, and individual executions for attempting to escape and disobedience, and deaths resulting from ‘intensive interrogation’ and torture.

For Fritz Bauer the purpose of the trial was ‘ultimately pedagogical’ (Pendas 2006: 52) in the sense that it would expose the deep reach of Nazism into the German psyche, and would contribute to the de-Nazification of German society. Bauer attempted to put the entire genocidal complex at Auschwitz on trial, and to demonstrate how the Holocaust was the outcome of widespread complicity of people from all walks of German life functioning under a complex criminal enterprise. Because the international Genocide Convention was ex post facto law, the defendants were tried under the 1871 German homicide law. That law limited liability for first-degree murder (‘Mord’) to those who were primary perpetrators, and who acted with base motives in taking the lives of others. Accomplices, while guilty, were considered to have significantly lower levels of culpability, particularly in terms of penalty. As a result, the routine activities of forcibly deporting millions of people from their homelands, imprisoning them in temporary ghettos, classifying them on the railway sidings for work or immediate death, and the subsequent act of gassing them, were viewed as regrettable, minor crimes (‘Totschlag’), akin to manslaughter. The law took a sterner view in the case of torture, and individual atrocities based on cruelty, sadism, or hatred. These crimes overshadowed the genocide itself. What was the result? The crimes of Auschwitz were equated with individually culpable acts of subjective barbarity. The state-initiated acts of mass murder disappeared as a focal point. The extermination system was ‘beyond justice’ (in the words of Wittmann), and escaped ‘the limits of law’ (in the words of Pendas). Six accused were convicted of first-degree murder (‘Mord’), but the majority of accused was convicted of being accessories to ‘Mord’ and received an average penalty of 6.3 years. Ironically, the ‘Totschlag’ convictions were implicated in the extermination of millions of victims, while the ‘Mord’ convictions were based on a few hundred deaths. Throughout the trials, the defendants acknowledged that crimes had occurred at Auschwitz, but denied their own guilt, since they had played ‘only’ a secondary role. In place of guilt, they often exhibited pride in the effectiveness with which they carried out their assignments.

According to Pendas (2006: 293), the trial ‘was unable to articulate adequately a historical account of the Holocaust that fully incorporated or even sufficiently acknowledged the extent to which it was a “total social event,” one in which every dimension of German society was implicated.’ Wittmann (2005: 271) concluded similarly: ‘The public gained a skewed understanding of Auschwitz. The sentences meted out to the defendants distorted the realities of the program of extermination [and] shifted the focus in the courtroom away from Nazi genocide towards individual acts of cruelty, suggesting that … the Nazi orders had been acceptable … ’ Hannah Arendt, whose riveting account of the Eichmann trial in Jerusalem appeared as the Auschwitz trial was winding down, also noted the failure of the proceedings to take hold in the minds of ordinary Germans: ‘Exposure for twenty months to the monstrous deeds and the grotesquely unrepentant, aggressive behavior of the defendants … had no impact on the climate of public opinion’ (cited in Wittmann 2005: 246). The trial failed to register the historical enormity of the events that created Auschwitz, including the explicit scheme of racial extermination, the plan to subjugate the entire continent of Europe, and the mobilization of the army, industry, and academy to bring this about. As far as the verdict at Frankfurt was concerned, most of the business at Auschwitz involved individual crimes that had little relevance for most of the Germans. Indeed, the majority of the crimes recognized by the Frankfurt trials, aside from genocide, were offences that would have been illegal under SS regulations.

One of the recurrent themes that runs through genocide trials, then and now, is that they are thought to be socially important by recording history through the testimony of witnesses and the examination of documents. The exposure of the lives of the accused, and of their victims, for the purposes of establishing guilt is equated with the historical task of establishing truth. The lesson of the Auschwitz trials is otherwise. The equation of guilt and truth is ill founded since such proceedings are adversarial and are arbitrarily limited by rules of evidence, procedure, and the slant of the laws. In this case, both Wittmann and Pendas laid the failure of the trial as history at the door of German law, particularly the procedural distinction drawn between perpetrators and accomplices and, in addition, the requirement of a base motive at the core of first-degree or capital murder charge. If the law had been written differently, the observers would have taken away an alternative lesson. There are two issues here. The first is that the legal lens will always shape and colour how events are portrayed, but so will the perspective of the historian. The second point relates to the limitations attributed to the specific German law. I offer an alternative reading to that of Wittmann and Pendas. The problem was not a deficit or oversight in the German penal code. A Durkheimian reading would suggest that the law follows changes in the collective consciousness of society. In The Division of Labour in Society, Durkheim (1893) argued that the forms of law (retributive, restitutive, commercial, and constitutional) reflected changes in historical patterns of affiliation, changing density, the rise of commerce, urbanization, etc. In this view, German law may well have reflected the hierarchical nature of German society by demarcating the role of perpetrators and accomplices so sharply. I would suggest that the subjective reactions of the accused during trial—smug indifference—reflected precisely the sort of mentality that Norbert Elias described in The Germans (1996). Elias notes that German civil society was marked by the substitution of strong leadership for individual autonomy in political and military matters. With the Prussian ascendency, national development occurred hand in hand with a deep aversion to democracy and individual political responsibility. The Auschwitz accused could not be shamed as a result of simply carrying out official orders, since they were not complicit in their design, although they followed them with panache, in accordance with duty. From this perspective, the 1871 law of homicide conceptualized accomplices as followers, in contrast to the Anglo-Saxon law, which viewed accomplices as associates, with near commensurate levels of responsibility.

The trial failed on two counts. It failed to capture the organizational nature of the Holocaust and the high level of complicity that it required, and it failed to awaken any sense of shame in the perpetrators. As I have noted earlier, we find the same absence of subjective guilt among the Rwandan génocidaires. The convention to prevent and suppress the crime of genocide created new international law without the limitations that Wittmann and Pendas attributed to domestic criminal law. How successfully has the 1948 Convention brought mass atrocities within the reach of the law in recent times?

The Nuremberg trials did not prosecute the crime of genocide. However, the General Assembly of the UN in December 1946 held that ‘genocide is a crime under international law which the civilized world condemns’, suggesting that it was already customary law. As noted earlier, the term which formed the core of the 1948 UN Convention was only coined by Raphael Lemkin in 1944. Certainly, people had been horrified by massacres of civilian populations in political conflict from the time of Melos (487 BC), Carthage (146 BC), and the Thirty Years War (1648). The 1948 UN Convention created the first positive law that described the elements of the offence, as well as the special mental element, or dolus specialis, required to establish guilt. This special element appears to presuppose a high level of agency among the perpetrators of mass murder. Not only do they have to kill, they have to be motivated to exterminate certain categories of people, and to do so deliberately because they belong to such categories. However, one of the recurrent observations in the genocide literature is the neutral, vacuous, or evanescent mentality of the legions of persons recruited for acts of mass murder. In the Rwandan case, the amnesty laws justified mass murder as part of the Hutu revolution, and hundreds of thousands of génocidaires were typically reluctant to accept any sense of guilt or remorse. Their killings were righteous, like those of patriots defending The Good. The legal process in the current genocide courts is premised on the idea that offenders are morally sensitive, and that the legal process will result in their acceptance of a guilty verdict, that they will stand condemned, and that their conviction will expedite acceptance of responsibility. But the evidence from these courts suggests that those assumptions which are associated with individual forms of offence and prosecution (mere murder) are misplaced in trials for mass murder.

We have been focusing on comparatively ancient history: the Auschwitz trials from West Germany in the Cold War period. To what extent have the weaknesses of earlier genocide trials persisted in the more recent trials? The first ad hoc UN tribunal to investigate genocide, war crimes, and crimes against humanity was created during the civil war in the former Yugoslavia in 1993 before the hostilities had ended. If law had a deterrent effect, presumably that would be evident from experience in the former Yugoslavia, since the combatants had knowledge of liability while conflict was ongoing. A second tribunal was created following the end of the Rwanda genocide in 1994. The former was confronted largely with allegations of war crimes, and crimes against humanity (‘ethnic cleansing’). The latter was overwhelmingly confronted with genocide, complicity in genocide, and incitement to commit genocide. The two ad hoc tribunals were created to deal with breaches of international humanitarian law separately for each conflict, one located in The Hague, the second in Arusha, Tanzania. A single prosecutor initially directed the two courts. When this proved ineffective, a separate prosecutor was designated for each court, although the appeal process was amalgamated under a single bench to ensure some continuity in the new jurisprudence that arose from trials in each region. The courts were initially under-resourced, and their progress was halting. However, as they attracted more reliable funding, their caseloads expanded, and they reported progress. Nonetheless, there were recurrent allegations of professional incompetence of tribunal staff, financial mismanagement, political interference, and disconnection of the legal process from the post-conflict societies. When subsequent calls were made for ad hoc tribunals for atrocities elsewhere in the world, the UN adopted a ‘hybrid’ approach in which the legal process would be based on international laws developed in the initial ad hoc tribunals, combined with input from domestic law and local judges, and funded in part by the post-conflict states themselves, and by voluntary state donations. The aim of the hybrid courts was to better integrate the stakeholders in justice: the victims, the national governments, and the United Nations. Unlike the Frankfurt court, which conducted its prosecution under national criminal procedures, and national criminal law, the ad hoc courts attempted to create something out of nothing. They were to act in specific regions with a global mandate created by the Genocide Convention, and consistent with the jurisprudence created famously at Nuremberg. They were in a certain sense operating in the dark since Nuremberg never prosecuted genocide. There was no settled jurisprudence for genocide. There were no clearly determined operational rules for criminal procedure comparable to those found in the nations that had created them. This opened the chasm between Anglo-Saxon adversarial procedure and continental inquisitorial procedure. The selection of judges and prosecutors was eminently political since the parties at the UN who created the courts were averse to creating judicial processes to which they themselves or their countrymen might be subjected, while at the same time assuming responsibility for the costs. And there was no institution corresponding to the police. Apprehension of international suspects was fraught with challenges to sovereignty that could result in state-to-state conflict. Those who designed such courts presumed that there existed sufficient transnational communication, recognition, and cooperation that a new layer of courts could simply be added to the existing national, provincial, and municipal levels of society. The national systems evolved over centuries. Expectations that a higher level of justice could be achieved in a short period of time as a result of globalization were clearly over-optimistic.

Have the new ad hoc tribunals succeeded where the Auschwitz trials failed? After all, the new courts have had the advantage of legal doctrines that explicitly recognized liability, which the 1881 German homicide law lacked. Genocide does not require a base motive, but a special motive. Conspiracy to commit genocide, incitement to genocide, and complicity are all major crimes like genocide per se, not minor offences. They also enjoyed the moral and financial support of the Security Council of the UN. In 2004 the UN Assistant Secretary-General for Legal Affairs shocked his colleagues when he publicly expressed doubts about the ICTR and ICTY. Ralph Zacklin, who helped create the tribunals, wrote in the most disparaging terms about them (2004: 545): ‘the ad hoc tribunals have been too costly, too inefficient and too ineffective. As mechanisms for dealing with justice in post-conflict societies, they exemplify an approach that is no longer politically or financially viable.’ Wittmann and Pendas said the fault lay with the 1871 German law. By contrast, Zacklin claimed it was the institutional success of the tribunals that had appropriated a world to themselves disconnected from both the realities of finance, and responsibility to the victims in whose name the proceedings were convened. The accomplishments, in terms of convictions, were modest, their progress was glacial, and their contribution to the restoration of social peace was questionable.

What has changed since 2004? Adam Smith (2009) examined the performance of the tribunals. His report reinforces Zacklin's earlier misgivings.

The courts have been monumentally expensive. The yearly ICTY budget from 1993 to 2007 expanded a thousand-fold, from $276,000 to $276 million; at various times in its tenure, the ICTY alone has accounted for 10 percent of the UN's entire annual operating budget. In all, from 1993 through 2007 the ICTY cost $1.2 billion, and is on pace to cost as much as $2 billion by the time it completes its mandate … Judicial productivity, however, has seemingly not matched the expense. The average cost per conviction at the tribunal has been estimated at nearly $30 million, more than fourteen times the average cost per capital conviction in the United States … The Rwandan tribunal is somewhat less costly, though it is projected to also have spent more than $1.4 billion by the time it finishes operations … In 2004, the Rwanda and Yugoslavia tribunals together constituted 15 percent of the total UN budget (Smith 2009: 182–3).

The ICTR budget for 2010–11 was $245 246 500. The ICTY budget for 2010–11 was $301 895 900. Both failed to meet their original 2010 ‘completion strategies’. The hybrid courts in Sierra Leone, Cambodia, East Timor, and Lebanon were created in part as a result of cost escalation at the ad hoc courts. How have they fared in comparison? Again, we turn to Smith. In the case of the Special Court for Sierra Leone, he reports as follows: ‘The original budget was very ambitious and called for only $54 million over three years; since the special court's opening in 2002, that amount has more than tripled’ and the three-year mandate has morphed into eight (Smith 2009:183). That was for ten cases, but the most important case, the indictment of Charles Taylor, was moved for security reasons to The Hague, duplicating much of the costs already invested in Freetown.

The Khmer Rouge atrocities were mentioned earlier. The costs of their genocide trials at the Extraordinary Chambers in the Courts of Cambodia were reported by Rebecca Gidley (2010: 14). ‘In terms of its finances, the budget for the court from 2005 until 2010 is US$142.6 million.’ Gidley comments that this is a lot less than either the ICTR or the ICTY, but she fails to mention that there were only five accused in the docket. Also, the UN News Service (UN News 2010) at the time pointed out that ‘for 2011, the total budget of $46.8 million is unfunded’. This suggests that the projected costs were nearly $189 000 000 if one includes 2011, or about $38 000 000 per case (RNW 2010). The hybrid courts do not appear to be less costly than the ad hoc courts on a per capita basis. But they raised more than financial questions.

As of mid-2012, there had been only one conviction at the Extraordinary Chambers in the Courts of Cambodia (ECCC), that of Kaing Guek Eav (aka ‘Duch’) who was commandant of the Tuol Sleng prison during the Khmer Rouge rule (1975–79). At least 14 000 people were tortured, and sent to their deaths there. Virtually no one survived. Throughout the trial, Duch cooperated with prosecutors. He said he was deeply remorseful for all the suffering he had created, and apologized to the dead, to their families and to all Cambodians (Mydans 2009). However, on the last day of his nine-month-long trial in November 2009, he petitioned the court to ignore all the evidence of murder and torture carried out by him or under his supervision, and asked that he be released without any further penalty. The public was shaken. On appeal of his thirty-five-year sentence on 30 March 2011, he reiterated his request to the Supreme Court Chamber, and asked them to acquit him and set him free. The court noted that this was equivalent to rejecting his culpability for his crimes. His behaviour raised grave questions about the sincerity of his remorse, a point related to the peculiar subjectivity of guilt raised earlier in this chapter. One is reminded of Hans Frank's very limited acceptance of responsibility at Nuremberg for Nazi crimes in Poland, which was mentioned earlier. The Rwandan genocide? Same story. Cruvellier (2010: 14) writes of Froduald Karamira, one of the ringleaders of the Rwandan genocide, who was convicted of genocide before the criminal courts in Rwanda: ‘Froduald Karamira never showed the slightest hint of remorse. He always maintained an air of defiance before his judges in Rwanda.’ He was publicly executed in Kigali in 1998.

There are two sets of reasons that the UN ad hoc and hybrid courts have been so costly, made such slow progress, and have so little to show for the investment. The first has to do with paradoxes of genocide outlined earlier. The vast majority of the perpetrators are not psychopaths with obvious criminal tendencies, but ordinary people who do not exhibit the usual symptoms of shame and remorse when confronted with evidence of their crimes. In addition, the crimes with which they are charged have become conventionalized. i.e. treated as above the law, or prosecuted only in the case of those on the political margins of power. The task of the courts is also different in terms of the scale of atrocities. Unless the courts focus exclusively on the most senior leaders in politics, and the military, the docket will have more cases than it can comfortably handle. The second set of problems concerns the very weak capacity of the courts to induce guilty pleas. In the common law countries, the vast majority of criminal matters are disposed of through guilty pleas, primarily negotiated guilty pleas (Brannigan and Levy 1983). These are the exception at the ad hoc tribunals. If crime were tried in the domestic courts without plea negotiations, this would slow their progress considerably. But more to the point, many of the high-level white-collar crime charges can only be established by testimony from co-accused working in the same operation against the more senior actors, a process that may lead to sentencing differentials associated with pleading, and appearing as a prosecution witness against future co-accused (Hagan, Nagel and Albonetti 1980). The ad hoc tribunals have been reluctant to do this because of the perception that the gravest international crimes would be discounted by plea-bargaining. We see evidence for this in the case of Jean Kambanda.

Kambanda was made prime minister of Rwanda after the coup d’état on 6–7 April that occurred with the murder of the president and the prime minister. The military junta installed a new government within days, and appointed Kambanda as prime minister. He was apprehended in Kenya in 1997, and returned to Arusha to face genocide, conspiracy to commit genocide, incitement to commit genocide, complicity in genocide, and crimes against humanity. He cooperated extensively with the prosecutor and provided days of interviews outlining how the junta had organized the mass murders, who did what, when, and where, and voluntarily submitted a plea of guilty to all charges. At the time, he declined court-appointed representation, and was kept isolated in custody from other offenders. He did not enter into a quid pro quo agreement to testify against other suspects, and was sentenced to life imprisonment without any possibility of release. At that point, he attempted to recant his confession on the grounds that his open cooperation ought to have resulted in some diminution of penalty, given his remorse, his confession, and his cooperation with the office of the prosecutor. The court found the crime so grave that these mitigating factors were unworthy of any weight, a view advanced by Madam Prosecutor, Carla Del Ponte. This sent a signal to every other accused who may have considered cooperating with the tribunal to bring the more recalcitrant offenders to justice through their evidence against them. Of the thirty-two completed cases at the ICTR, seven were convicted as a result of guilty pleas. At the ICTY, twenty of the 161 cases have been similarly resolved. These trends are the opposite of what is found in the common law national courts. In my view, the tribunals have already discounted the gravity of ‘the crime of crimes’ by eliminating the death penalty. The progress of the tribunals would have been greatly expedited by negotiating guilty pleas. The cost might be a reduction in sentence by several years—but a much greater capacity to register convictions, and to promote a culture of remorse. In fact, it might be argued that the inability of the courts to expedite guilty pleas contributes to the maintenance of the culture of impunity where no one is expected to admit guilt.

As I have stressed, the hybrid courts as well as the ad hoc tribunals have been very costly, and experienced appalling delays. On the other hand, the Special Panels for Serious Crimes (SPSC) in East Timor was starved of funding, receiving about $6 million annually from the UN in 2001, rising to $7–8 million by 2005. It conducted fifty-five trials involving eighty-seven accused. Eighty-four persons were eventually convicted for serious crimes arising from the mayhem of Indonesian-sponsored militias operating in East Timor in 1999 that were designed to enforce Indonesian dominance and to suppress the island's aspirations to political independence (Cohen 2006a). Some 1 400 murders were recorded during the September 1999 referendum. When the UN shut down the courts, almost 600 other murders were under investigation. Conviction of over eighty accused in four years was a higher figure than any other tribunal at the time. Despite this, the SPSC was shut down by Security Council Resolution 1543. The vast majority of the offenders were illiterate farmers who had been enrolled in pro-Jakarta militias. There were grave questions about the competence of the tribunal. All this was documented by David Cohen (2006b). For example, during the first fourteen cases, no court reporter, stenographer, or audio recording was utilized. As a result, all the appeals were conducted without benefit of trial transcripts, except for notes recorded by one of the judges. There was no experienced defence bar; it was essentially unfunded. Not surprisingly, the defence failed to call any witnesses at all in the first fourteen trials. Cohen (2006b: 93ff) also points out that a UN programme designed to create judicial infrastructure ended badly. Some twenty-five legal workers were schooled for several years in the basic principles of law. When they were evaluated in 2005, every single candidate was judged to be legally incompetent, including four native persons who had served as judges in the earlier trials, and whose judgments had contributed to the convictions. In spite of the relatively large number of convictions, the UN withdrew funding and the chief judge sealed all the case files, preventing future scrutiny.

The SPSC was plagued with under-funding, lack of access to competent translators, inability to apprehend senior Indonesian military and political suspects, and a hostile Portuguese-speaking government that was more interested in peace with Indonesia through diplomatic reconciliation than through criminal trials. From my perspective, what made these trials important was that the vast majority of those who were convicted pleaded guilty. Linton and Reiger (2002: 2) note that ‘one of the most interesting and unique features of these cases is the fact that almost every accused person admits to being involved in some aspect of the crime with which he is charged, for example, that he killed’. This may have been because there was no access to experienced defence counsel, or because of ‘the desire of many of them to acknowledge their involvement in crimes and be reconciled with their communities’. The evidence against them was presented at trial nonetheless. The courts resisted acknowledgement of the fact that the accused wanted to plead guilty. Like the Auschwitz accused, they had little remorse. Because they were peasants, they felt little autonomy in their actions, but unlike the Auschwitz accused, they wanted to repair the breach of the peace by acknowledging their guilt. ‘In Timorese culture, the expected practice for the accused was to confess his crimes … In order to promote the culture of the not guilty plea required by the Western court systems, the UN experts had to train the Timorese to lie’ (UN diplomat Shashi Tharoor, quoted in Moghalu 2008: 14).

If common law experience is any guide, guilty pleas can expedite the legal process. However, it may be that a diplomatic process and/or a national court process may be a better outcome than what is offered at the international level. Adam Smith makes this case in his analysis of the conflict in the Balkans. These are two lessons based on his research in this region. Specifically, the first is that the cessation of hostilities following the Dayton Peace Accords led many to call for the cancellation of the Hague prosecutions, and the release of persons apprehended by the ICTY. Since the Serbs had been persuaded to put down their arms, and President Milošević had helped broker the deal, further prosecution appeared to be perverse, since the initial call for the tribunal was designed to halt aggression. Smith argues that the objectives of justice, and those of peace, sometimes operate at cross-purposes. Moghalu stresses the same point in his analysis of the competing agendas of security and diplomacy. Diplomats can often negotiate the end to conflicts without bringing anyone to trial; frequently, it is easier to de-escalate the tension between combatants when aggressors effectively receive immunity, or pardons, as a guarantee of the establishment of security. One reason that Serbs developed such an aversion to the ICTY was that many thought it was made redundant by the Dayton Accords. The same issue has led former Bosnian Serb President Karadžić to refuse to plea, claiming that his participation in the Dayton Peace Accords was negotiated to leave him free of subsequent prosecution (MacDonald 2009).

Smith's second point is that nation states should prosecute breaches of international humanitarian law internally. He points out that before Milošević was surrendered to the ICTY, he ‘was arrested, indicted, and set for trial in Serbia’ (2009: 326). Croat prosecutors were already prosecuting Croats for crimes against Serbian populations and Bosnians were likewise prosecuting their own citizens for war crimes against Serbs and Croats. Smith rejects the idea that domestic prosecution for these sort of inter-ethnic atrocities is impossible.

The questions about trials and tribunals raised by Wittmann, Pendas, Smith, Zacklin, and other critics raise issues that probe whether the courts contribute to an accurate history of atrocity, and whether they are efficient and effective in restoring peace. Nancy Combs (2010) raises a more fundamental question: are the tribunals effective as courts? Her book, Fact-Finding Without Facts raises the uncertain evidentiary foundations of international criminal convictions. Her critique is based on a detailed investigation of how the chambers actually acquire information from witnesses. She focuses on the ICTR, and all the hybrid courts except Lebanon. Although her sampling frame is exploratory, her position is convincingly argued.

If one reviews the viva voce evidence at the ICTR, the EECC, the Special Courts for Sierra Leone, and the SPSC in East Timor, one finds that the witnesses are by and large illiterate peasants who are unable to provide basic reliable information required by the prosecution to establish the actus reus of the crimes. These deficiencies include the following basic points: inability of witnesses to report the dates (specific months or weeks) when the crimes were observed; inability to describe the duration of the events associated with massacres, and/or the distances travelled by survivors during the events. In addition, the witnesses repeatedly fail to provide estimates of the number of victims or perpetrators involved, and the witnesses are often unable to understand two-dimensional representations of crime scenes, such as maps or drawings. They cannot recall important details such as the make of a vehicle used by paramilitaries, or whether they were using some type of gun or ‘bows and arrows’ (Combs 2010: 39).

In addition, the witnesses are unable to appreciate such adversarial procedures as cross-examination, and appear insulted when questioned about their evidence. They struggle to answer simple questions such as ‘what happened next?’ or ‘what else did you see?’ Even more disturbing, in important cases as much as half the oral testimony led by the prosecutor departs significantly from pretrial interview statements made available to the defence.

Combs attributes the causes of these weaknesses in evidence to several factors. The level of illiteracy among witnesses frequently exceeds 50 per cent. Support for effective multilingual translations is frequently unavailable or inadequate, so repeatedly the bench is uncertain what exactly the witness is reporting. Linguistic differences and the life experiences of peasants undermine their ability to function effectively in Western-style trial courts; in many oral cultures, direct accusation is considered rude, and people are constrained to formulate accusations in indirect language. In adversarial proceedings, this evidence appears to be evasive and inconclusive. In addition, pre-existing animosities between the groups in conflict make perjury common, since witnesses naturally favour their own side in ethnic or racial conflicts.

Combs outlines the consequences. Trial judges, typically from industrialized nations, adopt what she describes as a ‘lackadaisical attitude toward testimonial deficiencies’ and admit evidence to the record that would raise serious concerns about witness credibility in Europe and America. Furthermore, the vagueness of the factual evidence makes it difficult for the accused to offer alibi evidence about their alleged involvement in crimes, since the case they must rebut, due to the ambiguity of testimony, frequently is inherently imprecise. More worrisome is that the criterion for conviction is effectively lowered from ‘beyond a reasonable doubt’ to ‘preponderance of evidence’. In addition, the high levels of perjury suspected, particularly at the ICTR, are virtually never investigated. Combs notes, in an aside, that all the current ICC cases deal with central African conflicts whose testimonial deficiencies likely reflect the same problems Combs documents in all the current non-European tribunals.

Combs explains the lackadaisical approach of the judges as follows. Since the judges are elected by the Security Council based on their commitment to international courts, they are reluctant to rule out evidence that would be inadmissible in Europe or America since this would doom the courts. According to Combs, they admit such evidence particularly where the accused's official position is evidence of his or her probable involvement in atrocities. Politicians, military personnel, and militia commanders are assumed by the trial chambers to be culpable, even if the evidence is vague or contradictory. The result is a very high conviction rate, based largely on testimony that would be ruled inadmissible by domestic courts.

Critics of the international courts would condemn the enterprise on these grounds alone since the criterion of guilt is the preponderance of evidence. Combs offers three basic solutions to this situation. First, the courts could explore modest procedural reforms to improve fact-finding to raise the reliability of evidence. Second, the courts could undertake major procedural reforms in how evidence is presented, and could formally adopt doctrines of collective culpability to bring them in line with the existing practice. And finally, the international community could revisit the ability of domestic courts to prosecute what have become defined as transnational crimes.

The first remedy consists of modest procedural reforms to improve translations. ‘Many of the colossal interpretation failures that I have recounted stemmed not from intractable linguistic and cultural divergences but from resource restraints’ (2010: 276). Remedy: better translation services and more funds. In addition, the office of the prosecutor has to improve the competence of investigators, and ensure the continuity of the evidence, so that discrepancies between statements and testimony cannot be dismissed as investigator incompetence. This might require audio recording of statements. Combs also stresses the value of on-site visits by the judges to the crime sites to better appreciate the conditions under which the witnesses gathered information. And finally, the courts have to be more aggressive in labelling and punishing perjury.

Combs is not optimistic that modest reforms would have much impact. More serious procedural reforms might replace the adversarial nature of the evidence presentation with the inquisitorial methods under which the judges have a central role in establishing the case against the accused, and come to trial briefed on the evidence, and better prepared to direct the questioning of the witnesses themselves. In principle, the courts draw from both common law and continental procedures; this change would follow more closely the continental approach. In addition, Combs explores how the courts could reduce the impact of testimonial deficiencies by aligning the stated and the actual conviction justifications through more explicit adoption of ‘associational doctrines’ (2010: 321). For example, in the case of Duško Tadić, the court of appeal at the ICTY ignored the fact that there was no direct evidence against the accused that he had killed five captives. It was enough that he was known to have kidnapped them in an illegal operation of ethnic cleansing. Since it was said to be common knowledge that such captives were often killed, it was said that Tadić knew or ought to have known that the captives might be endangered. The court of appeal dealt with the absence of direct evidence ‘by constructing an expansive legal doctrine that permits the imposition of criminal liability on defendants who did not themselves personally commit the crimes in question or intend that such crimes be committed’ (2010: 324). This is the doctrine of joint criminal enterprise. By entering into a criminal collaboration, defendants can be linked to atrocities in which they played no personal role, but which formed part of the common purpose of the enterprise, and which were foreseeable as potential outcomes (2010: 328). While this doctrine has been criticized as a theory of guilt by association, and while it expressly obviates the accused's specific intention, Combs argues that the tribunals have already tacitly adopted the doctrine, even when they do not refer to it explicitly.

Combs also argues more broadly that the international tribunals should be prepared to lower the threshold for conviction—ultimately leading to a higher incidence of mistaken or false convictions than would be acceptable in a domestic prosecution. This is because the international cases are more likely to falter ‘as a consequence of investigatory failures than as a consequence of true evidentiary insufficiency’ (2010: 352). Also,

there is a greater likelihood that he (as opposed to a domestic defendant) committed the same crime but in a different way … [and/or] … there is a greater likelihood than in the domestic context that he committed some other crime … Finally, even if the international defendant committed no crime during the genocide, he may bear some moral culpability for acquiescing in it (2010: 355).

Combs's last remedy is offered in an uncharacteristically perfunctory way in the conclusion. It examines the possibility of replacing the international tribunals with domestic trials based on the international jurisprudence (2010: 367). These courts would not have the same translation problems, perjury would be more easily detected, and the local judges would have a better sense of the local context of the conflicts. However, as Combs points out, many countries rocked by atrocities do not have functioning judicial institutions. In addition, such trials might be open to political interference. Combs concludes that ‘there is little reason to believe that recourse to domestic mechanisms will substantially enhance fact-finding accuracy’ (2010: 372). But surely that depends on the jurisdiction.

On the issue of domestic courts, Combs's view contradicts Smith's evidence from the Balkans. He reports a far higher level of success where there exists an effective judiciary with a modicum of independence. Smith also reports on the successes in Latin American countries whose dictators departed office after legislating amnesties for themselves. Smith notes how governments in Chile and Argentina subsequently have reopened these cases and brought the authors of atrocities before the court, without UN legal services. Smith makes a compelling case for re-examining the role of domestic courts in establishing cosmopolitan justice.

Although Combs is chilly on the prospects of the domestic option, ironically, she does not appear to be convinced of the value of the international alternative. ‘My normative assessment of the tribunals’ fact-finding deficiencies assumes that international criminal trials are valuable endeavors. But just how valuable and for what purposes are questions that I believe to be unanswerable at present’ (2010: 372). This might have been Fritz Bauer's reaction after the Auschwitz trial. But after decades of transnational justice from across the globe, it is a very pessimistic conclusion.

This conclusion suggests another view. I think Combs's work puts the nail in the coffin for transnational tribunals, particularly as they apply to largely illiterate societies. Although they may have been conceived in cosmopolitan society, the UN has exported these hybrid institutions to non-European venues where they are largely dysfunctional, where the witnesses are baffled by the role they are expected to fulfil as actors with stronger alliances to the court than to their own communities; where they are stymied by the perverse attention to minutia about dates, times, and duration that they do not ordinarily heed; and where they are assumed to accept complex understandings of the relationship between guilt and individual autonomy, whether or not this reflects the mentality of their own communities. Where the societies are relatively cosmopolitan (Sarejevo; Belgrade), the domestic courts already appear to have a capacity to function without the UN courts.

Combs's scepticism about the courts is shared by Zahar (2010) who believes that witness memories at the tribunals are typically far removed in time from the original events, and frequently questionable and influenced by suggestions of investigators and prosecutors. ‘Nowhere does international justice feel more experimental and insecure than at the evidentiary coal face of the witness hearing. The observations … go to the larger question of how convincing international fact-finding really is.’ Based on his observations of hundreds of witnesses at ad hoc tribunals, he goes on to say that

the truth about the trials I have experienced, which include cases of all sizes and complexity, is that the evidence can be written up in one direction (guilt), or in a radically different direction (acquittal), depending on the final preferences of the judges … Most of the time the evidence can be led with equal ease in either direction … The factual findings with which I am familiar, and which I have no reason to believe are exceptional, are not findings ‘beyond a reasonable doubt’ but rather ‘reasonable findings in the circumstances.’

Nice (2001) also identifies many of the imperfections of the ad hoc tribunals but expresses more optimism for their capacity to overcome them.

Another recent critical account from the courts is provided by Thierry Cruvellier who reported on the Arusha courts for five years. In Court of Remorse, Cruvellier (2010) captures the disorganization and incompetence that marked the creation of the Arusha court in its early years. When Canadian Louise Arbour replaced Goldstone, she found that the office of the prosecutors ‘lacked strategy, discipline, and coherence’ (2010: 15). The registrar and the deputy prosecutor were forced to resign. Arbour succeeded in bringing a series of indictments against twenty-eight defendants associated with Colonel Bagosora, considered at the time the architect of the genocide, but the court refused to permit such a massive joint trial. Bagosora was always treated as the mastermind behind the genocide, the head of the 1991 army committee that supposedly started planning the massacres, the individual who left the Arusha accords in 1993 promising to ‘unleash the apocalypse’, and who ignored the murder of the ten Belgian peacekeepers working under Romeo Dallaire. Bagosora was a retired general and chief of staff. When his superior left on a trip in April, he assumed command of the army for three days. The court dismissed his culpability for all events associated with the genocide, save for these three days. The 1991 army committee did not plan the genocide, and was staffed by several generals who openly opposed it, including Leonidas Rusatira. The witnesses who reported the ‘apocalypse’ quote were discredited. Although he was sentenced to life imprisonment, on appeal (December 2011) Bagosora's penalty was reduced to thirty-five years and the question of who implemented the genocide is still an open question. The mountains of material gathered by investigators over seventeen years were largely irrelevant. Cruvellier (2010) calls this ‘the brainless genocide’ and suggests that it is indicative of the incompetence that has marked the court from the start.

Cruvellier's book is a searing condemnation of the tribunal, which documents failures of disclosure, inconsistent sentencing philosophies, inept investigations and prosecutions, and an unwholesome over-reliance on questionable eyewitness evidence, often given anonymously or in camera, and often accepted naively. He reports how senior military figures who appear as witnesses in the defence of other military figures are threatened themselves with indictments. He is critical of the under-reporting of the procedures by major news agencies as well as the legal harassment of journalists sceptical of the courts who have been threatened with contempt. Cruvellier leads one to conclude that the standards of legal practice are not higher at the international level than in the common law courts, but significantly lower.

Are current judicial responses to mass murder ‘cosmopolitan’? It does not appear so. The international community may have begun the process effectively at Nuremberg in 1945. But the effectiveness of the courts (tribunals, hybrids, and the ICC) has become increasingly questionable the more they proliferate. Cost is relevant, but Combs's evidence is more worrisome, because she establishes that justice provided by the tribunals outside of Europe proceeds with lower standards of evidence than would be acceptable in most Western jurisdictions. In contrast, in the hybrid courts, the UN is imposing legal traditions that are foreign to the indigenous populations. The unstated implication is that these tribunals will accelerate the ‘modernization’ of peasant societies in terms of a legal code and procedure that are euro-centric. The West is imposing law based on Western conceptions of fairness, culpability, and procedure, irrespective of the level of development of such societies. Ironically, the West makes these changes by directing financial support into The Hague, not Kigali, Belgrade, or Sarajevo—cities that could use the investment to develop indigenous legal solutions. The sovereign impunity for crime at the domestic level has migrated from the national context to the level of superpowers. Now the superpowers can hold the sovereigns of minor nations accountable for their crimes, while escaping control themselves. This is epitomized by the failure of the superpowers to join the Rome Treaty in support of the ICC. This is not the world of Kant as much as the world of Nietzsche. The genealogy of good and evil is power.

The interests of victims of mass atrocities might be better served by prosecution of perpetrators under national criminal law, by internal truth and reconciliation commissions, by diplomacy and pardons, and by inter-state acts of compensation such as those arranged at the International Court of Justice (ICJ). I shall explore some of the possibilities of these alternatives in the next chapters.