National Courts and Genocide: The Kravica Case at the Court of Bosnia and Herzegovina

Abstract

This article comments on the recent Kravica judgment of the Court of Bosnia and Herzegovina, which relates to the mass execution of over 1,000 Bosniaks after the fall of Srebrenica. The judgment relies extensively on the case law of the ad hoc tribunals, at times contributing towards its development. The Court dismissed JCE liability, holding that the accused were mere tools implementing the common plan. It convicted six foot soldiers for co-perpetrating genocide, developing an original test to infer their genocidal intent. Furthermore, it convicted their commander for genocide on the basis of command responsibility, taking a questionable stance regarding his mens rea. The Court declined to state wether the law requires a commander to be simply aware of his subordinate's specific intent, or to share that intent. This approach was justified on the assertion that the higher test was in any event met in this case. In an effort to prove it was so, the Court engaged in an unconvincing reasoning purporting to infer genocidal intent from the commander's omissive conduct.

1. Introduction

On 29 July 2008, the Court of Bosnia and Herzegovina¹ pronounced three separate verdicts, marking the conclusion of the first instance proceedings in the Kravica case, a genocide trial against 11 accused, which had lasted over two and a half years.² The verdicts were rendered separately after the cases against two of the accused had been severed from the rest due to procedural reasons.³ The case concerns the execution of over 1,000 Bosniak men, in the wake of the fall of the Srebrenica enclave, at a location known as the ‘Kravica warehouse’. The Court convicted six accused for co-perpetrating genocide,⁴ and one for genocide on the basis of command responsibility,⁵ sentencing them to prison terms ranging between 38 and 42 years. The remaining four accused were acquitted.⁶

A verdict for genocide demands attention, especially when issued by a national court. What makes this case even more noteworthy is that it is the first trial for the genocide perpetrated in Srebrenica to be conducted by a national court, one of few judgments to convict foot-soldiers of ‘the crime of crimes’, and one of just three cases, to base a conviction for genocide solely on command responsibility.⁷ This article will present a comment on the case. The focus of the analysis is two-fold: firstly the case will be placed in perspective by outlining the objective and subjective elements of the alleged crimes that distinguish it from other Srebrenica cases. Secondly, the most interesting legal issues the Court had to address will be reviewed. These mainly concern the substantive law of genocide and the modes of liability as applied to the crime of genocide.

2. The Trial in Perspective

A. Ratione Materiae

The overall Srebrenica events have attained such a degree of notoriety that it is unnecessary to state them in detail. On 11 July 1995, the UN protected area of Srebrenica capitulated after a six-day offensive of the Army of Republika Srpska (VRS). As the VRS entered the town its inhabitants, numbering approximately 40,000 Bosniaks, fled. About 25,000 individuals, mainly women, children and the elderly, sought refuge in and around the UN compound in Potočari. The remaining 15,000, comprised mostly of men, departed in a column in an attempt to reach the territory controlled by the Army of Bosnia and Herzegovina (ARBiH). The VRS seized control of Potočari and the UN compound, separated off and detained the (largely) military-aged men, and forcibly transferred the women, children and elderly to the territory controlled by the ARBiH. Whilst the men in the column were en route they were subjected to artillery fire, ambushes, capture and detention. The detained men, estimated between 7,000 and 8,000 were executed at various sites between 13 and 18 July 1995.⁸

The acts which are the subject of the Kravica trial took place on 13 July 1995, against the background outlined above. On that day, the co-perpetrators⁹ were deployed at the Sandići meadow where a large number of Bosniaks from the column were induced to surrender and captured. Over 1,000 such captives were taken to the Kravica warehouse where they were executed the same day. The Court established that the co-perpetrators had participated in capturing, detaining, and escorting the Bosniaks to the warehouse. It further found that they had executed most of the captives by firing a machinegun, automatic rifles and throwing hand-grenades inside the crammed warehouse over a time-span of about one and a half hours.

Organized mass executions, such as the one which occurred at Kravica, were the most shocking of the range of atrocities which were perpetrated against the Bosniaks from Srebrenica. The Court found that of all the organized multiple killings of Srebrenica Bosniaks, ‘the murders at the Kravica warehouse probably involved the largest number of victims. The approximation of “more than 1000” has been used in the Indictment and the operative part of th[e] verdict, and the number actually killed by the Accused has been characterized as “the majority” of these.’¹⁰ Taking the widely accepted figure of total Srebrenica victims as ranging between 7,000 and 8,000, basic arithmetic shows that the co-perpetrators in this case, whether alone or together with others who have not yet been brought to justice, have put to death more than 10% of the total victims.

B. Ratione Personae

The accused in this case were no genocide masterminds, but foot soldiers. They were all members of a special police detachment, an elite police unit trained to do ‘complex police tasks’.¹¹ Only two of the accused had a commanding role, namely Trifunović, who was a platoon commander and participated in the acts together with four of his subordinates, as well as Stupar, who as detachment commander was the superior to all but one of the other accused, but was not present at the execution.

The International Criminal Tribunal for the former Yugoslavia (ICTY) has been focusing its efforts to ensure accountability for the Srebrenica genocide against defendants who were in much higher positions than the accused in this case, most of them being military officers at the level of Main Staff, Corps Command, or Brigade Command. The notable exception is the case of Dražen Erdemović, who as a foot soldier tried for participating in a mass execution, is in a position comparable to the accused in this case.

The Court of BiH, which is designed to deal with lower ranking perpetrators compared to the ICTY, has been pursuing defendants at Brigade Command¹² and Battalion Command¹³ levels, as well as rank-and-file soldiers.¹⁴ In the latter category falls the case of Vaso Todorović, also tried for the Kravica massacre.

Despite the obvious merits of prosecuting the genocide masterminds, prosecuting accused such as the defendants in Kravica should not be discounted as unimportant. As the Court reasoned, these accused ‘were not the architects of the plan, nor were they the tacticians or commanders who had responsibility for its overall accomplishment. Rather, they were the instruments by which the plan was carried out.’¹⁵ Nevertheless, as the Court properly emphasized, ‘[w]ithout people willing to carry out the genocidal plan by commission of the kinds of [proscribed] acts … genocide could not be committed’.¹⁶

The last passage is a key to understanding the striking disparity of sentences that these accused received, when compared to those imposed by the ICTY and the Court of BiH on Erdemović or Todorović, who participated in similar crimes holding similar positions. Todorović, who was a member of the same unit as the accused in the Kravica case, was present at the warehouse but refused to take part in the executions. His contribution was limited to escorting prisoners and standing guard during the execution, which in the absence of genocidal intent, could amount, at most, to aiding and abetting genocide. Erdemović did participate in shooting, killing approximately 70 men and boys at the farm in Pilica, but he was found to have done so under extreme duress and to have sincere regret. Both pleaded guilty to charges of crimes against humanity and were sentenced to 6 and 5 years’ imprisonment, respectively.

On the contrary, none of the co-perpetrators in the Kravica case was found to have been acting under duress. The Court actually found that they had the opportunity to disengage from the act, as shown by the fact that Stevanović (one of the accused), as well as a protected witness known as S-4 did disengage. Moreover, none of them showed any particular sign of remorse. The six co-perpetrators were enthusiastic implementers of a genocidal plan conceived above their heads. They were found to have acted with genocidal intent.

3. Genocide Law: Theory and Practice

A. Theory

1. The Court's Identification of the Law: The Principle of Legality

The Court of BiH applies municipal criminal law. Bosnia-Herzegovina is a party to the European Convention on Human Rights (ECHR), which is directly applicable and takes precedence over national legislation. Accordingly, the Court is bound to apply the rules on succession of penal laws set forth in the BiH CC and in the ECHR. In this light, it might surprise the reader that the Court applied the current provision criminalizing genocide, namely Article 171 of the BiH CC, which entered into force in 2003, to a crime which was committed in 1995.

The Court found the application of the new law not to violate the principle of legality (Article 3 of the BiH CC, Article 7(1) ECHR) considering that genocide was already criminalized in identical, if not broader, terms under the municipal law applicable at the time of the offence, that is Article 141 of the Criminal Code of the Socialist Federal Republic of Yugoslavia (SFRY CC).¹⁷

The issue of length of punishment was thornier, since under the SFRY CC the offence could have been punished with detention up to 20 years or the death penalty, whereas the BiH CC provides for detention up to 45 years. This issue comes up before every judgment of the Court of BiH, and it is settled practice for the Court to apply the BiH CC. The issue of change of statutory punishment is regulated by Article 4(2) of the BiH CC providing that ‘[i]f the law has been amended on one or more occasions after the criminal offence was perpetrated, the law that is more lenient to the perpetrator shall be applied.’¹⁸ However, Article 4(a) prevents the application of Article 4 if its application would ‘prejudice the trial and punishment of any person for any act or omission which, at the time when it was committed, was criminal according to the general principles of international law’. The Court based its reasoning on this latter provision, thus implying that the SFRY CC was the more lenient but that its application would have prejudiced the trial and punishment in this case, hence concluding for the application of the BiH CC. It further noted that this practice is consistent with Article 7(2) ECHR.¹⁹

It is unfortunate in this respect that the Court did not rely on another — and in this author's view more persuasive — argument, notably that the current law is more lenient because any custodial punishment is always more lenient than the death penalty. Thus the application of the BiH CC would be consistent with Article 4(2), mandating the application of the more lenient law, and would not violate Article 7(1) ECHR, preventing the retroactive imposition of a higher penalty. By recurring to the fall-back position, the Court implicitly disagrees with this reasoning, which is supported by the European Court of Human Rights case law.²⁰

2. The Court's Interpretation of the Law: Special Intent

The Court considered that Article 171 of the BiH CC defines the offence of genocide in virtually identical terms as Article 2 of the 1948 Convention on the Prevention and Punishment of the Crime of Genocide. It held that, as domestic law derived from international law, Article 171 ‘brings with it as persuasive authority its international legal heritage, as well as the international jurisprudence that interprets and applies it’.²¹ On this basis the Court relied extensively on the jurisprudence of the ad hoc tribunals, interpreting the elements of the underlying act, as well as the specific genocidal intent consistent therewith.²²

The cardinal problem in securing genocide convictions is how to establish the special intent of which, save in the case of a confession, there will hardly ever be any direct evidence. The Court recalled international jurisprudence holding that ‘the circumstances and facts surrounding the perpetrator's acts can, as a matter of law, establish genocidal intent beyond doubt’.²³ The international tribunals have tried to list examples of indicia of intent in several cases.²⁴ The Court developed its own test stating that ‘[e]vidence regarding: 1) The general context of events in which the perpetrator acted; 2) The perpetrator's knowledge of that context; and 3) The specific nature of the perpetrator's acts; when taken together, can establish the perpetrator's intent beyond doubt.’²⁵

It is submitted that this test is consistent with the examples indicated by the international jurisprudence, with the added value of being a more precise test, particularly suitable to probing the mens rea of foot-soldiers carrying out acts of genocide.

B. Practice

1. Murder and Intent

The findings related to the underlying act are of no particular interest since the Court had unquestionable evidence to conclude that the six co-perpetrators either carried out the actus reus of the crime or made a decisive contribution thereto, with the intent that the prisoners be killed.²⁶

2. Dolus Specialis

The application to the facts of the test for dolus specialis is more interesting. The Court identified the protected group as Bosnian Muslims, or Bosniaks, who were objectively identified as a ‘constitutive nation’ in the constitution and were subjectively stigmatized as a national and religious group. Bosniaks from Srebrenica constituted ‘a part of the protected group’ for the purposes of the definition. Mirroring ICTY analysis in Krstić, the Court held that although the numerical factor alone does not make the Srebrenica Bosniaks a substantial part of the group, the evidence establishes that ‘this population was a particularly prominent and significant part of the group of Bosniak people, particularly by July 1995 [due to its] immense strategic and symbolic value’; accordingly the Bosniak population of Srebrenica was a ‘substantial’ part of a protected group.²⁷

The Court then analysed ‘the general context of events in which the perpetrator acted’. It looked at the broader Srebrenica events (referred to by the Bosnian Serbs as the ‘liberation of Srebrenica’), identifying a first phase of military takeover and a second phase of eradication of the Bosniaks through forcible transfer of the women, children and the elderly, and detention and murder of the men. It examined how the operation was organized and coordinated ‘from the top’, evident from its complexity and the amount of resources involved (in particular it noted that the order to use the special police force had come from Radovan Karadžić, and that their tasks were ordered and monitored from the top of military and civilian structures). Based on the foregoing, the Court concluded that ‘there was a plan to destroy a protected group in part, perpetrated against the Bosniaks in Srebrenica by the Bosnian Serb forces, and implemented by forcibly transferring the women children and elderly and killing the males’.²⁸ The Court found that on the basis of such plan genocide was in fact perpetrated,²⁹ and that the execution at Kravica was consistent with that plan.³⁰

Relevant to ‘the perpetrator's knowledge of that context’ the Court stated that the accused were not the genocide masterminds, but merely ‘the instruments by which the plan was carried out’.³¹ As such, they ‘were not privy to the entire plan or to each of the acts that were contemplated to carry that plan out’.³² Nonetheless, the Court found that they were already ‘aware of the rudiments of the genocidal plan that was conceived by their superiors’ before reaching the Srebrenica area.³³ Their knowledge was confirmed and reinforced through what unfolded before their eyes in Srebrenica during the time leading upto the killings.³⁴ By that point they were inevitably aware of the ongoing operation to eradicate the Srebrenica Bosniaks through forcible transfer of the women and elderly and murder of the men, an operation which would have resulted in the destruction of the Bosniaks from Srebrenica as group. In short, they were aware of all the factual elements that made the operation a genocidal operation.

Then the Court turned to ‘the specific nature of the perpetrator's acts’. It built upon the brutality of the manner and character of the execution to assert that ‘the Accused did not simply intend to kill the victims; they intended to destroy them’.³⁵ By using this term the Court apparently intended to underline the perpetrators’ cruelty and persistence in inflicting lethal damage to the victims, even to their corpses once they were already lying dead in the warehouse. Albeit the use of this term might lead to erroneously conflate the mens rea requirement towards the group (i.e. the intent to destroy) and that towards the victims of the underlying act (in this case, the intent to kill them), this feature of the perpetrator's acts is nevertheless an important element in highlighting their enthusiasm in performing the odious task they had been assigned.

By looking at the acts of the accused in the light of the context of the ongoing genocide operation, and of their knowledge of that context, the Court was satisfied beyond doubt that when committing the execution, the accused possessed the required genocidal intent. Six accused were thus found to be co-perpetrators of ‘one of the genocidal acts committed against the Srebrenica Bosniaks during the period between 10 and 19 July 1995’.³⁶

4. Modes of Liability

A. Introduction

All the accused had been charged with co-perpetrating the crime of genocide as participants in a joint criminal enterprise (JCE). Two of them were also charged on the basis of command responsibility. The judgment dismissed the JCE theory pleaded by the Prosecution and entered genocide convictions on the basis of co-perpetration (for the six accused who had participated in the massacre) and of command responsibility (for the detachment commander). This section will address the legal and factual issues which the Court had to face in identifying and applying the appropriate modes of liability.

B. JCE and Co-perpetration

1. Theory

Co-perpetration and JCE are doctrines designed to address group criminality; while the former is embedded in many civil law systems, the latter appears to rely more on common law concepts.³⁷ The co-existence of these two doctrines in any system of criminal law is doomed to be uneasy as, due to their similarities, overlaps, and unclear mutual boundaries, the two doctrines are likely to compete in practice. This has proven to be the case at the ICTY where critics of JCE have argued that it should be replaced by co-perpetration, although eventually the Appeals Chamber in Stakić has dismissed the latter as a mode of liability applicable before the Tribunal.³⁸ This uneasy co-existence also unfolds in BiH whose Criminal Code, according to the prevailing interpretations, acknowledges both doctrines.

Co-perpetration is regulated by Article 29 of the BiH CC, which reads: ‘[i]f several persons who, by participating in the perpetration of a criminal offence or by taking some other act by which a decisive contribution has been made to its perpetration, have jointly perpetrated a criminal offence, shall each be punished as prescribed for the criminal offence.’³⁹

The above definition of co-perpetration is relatively wide, since it also includes those who have not performed the actus reus of the joint crime, but have acted in other ways so as to make a decisive contribution to the collective perpetration of the crime. Its wideness becomes apparent if we compared it with another definition of co-perpetration, identified by Antonio Cassese, as the situation in which more persons all materially take part in ‘the physical carrying out of the prohibited conduct accompanied by the requisite psychological element’.⁴⁰

Trying to apply the two notions in the case at issue confirms their different scope. If we were to apply Cassese's definition, then only the accused who fired their weapons or threw grenades in the warehouse could be held responsible as co-perpetrators. Conversely, the Court relying on Article 29 qualified as co-perpetrators also those accused who did not shoot but who stood at the doors of the warehouse during the execution in order to prevent the prisoners from escaping. Given its wide scope of application, the definition of co-perpetration in the BiH CC substantially trespasses into the realm of JCE.

Unlike co-perpetration, JCE is unwritten in the BiH CC. Nonetheless, the prevailing interpretation maintains that, as at the ICTY, JCE is included in the word perpetrating contained in Article 180, which mirrors Article 7 of the ICTY Statute.⁴¹

Whilst the acceptance of JCE in BiH is no longer seriously questioned and the Court of BiH has followed the ICTY jurisprudence in its application, any attempt to soundly define the relations between co-perpetration and JCE have thus far failed.⁴² Any deeper analysis of this relation would go beyond the scope of this article; for our purposes it is sufficient to have established that both doctrines are applicable in BiH and that their relationship remains unclear. The Court in Kravica thoughtfully analysed the two doctrines with relation to the facts of the case, unfortunately without assisting in clarifying their relation.

2. Practice

(a) JCE: charged and rejected

The Indictment alleged that all the accused participated in a JCE whose purported membership includes many notorious military and police leaders (such as Ratko Mladić, Radislav Krstić and Ljubomir Borovčanin) and extended to ‘many other individuals and military and police units who took part in the operations of the forced transferring (sic) and killing of Bosniak men.’⁴³

The common purpose of the alleged JCE encompasses the entire plan for the ‘liberation’ of Srebrenica, being defined as ‘to forcibly transfer women and children from the Srebrenica enclave to Kladanj on 12 and 13 July 1995; to capture, to detain, to summarily execute by shooting, burying, and reburying thousands of men and young boys, Bosniaks from the Srebrenica Enclave, aged between 16 and 60, in the period between 12 July 1995 and around 19 July 1995.’⁴⁴

The specific actions in which the accused are alleged to have participated, all but one of which relate to the Kravica operation, are spelled out in the Indictment.⁴⁵ It seems to be implied that those actions would have constituted the alleged contribution of the accused to the implementation of the common plan which, as the Indictment notes, resulted in the execution of more than 7,000 Bosniak males from Srebrenica.⁴⁶ Read as a whole, the Indictment states that by participating in the JCE and performing the specified actions, the accused co-perpetrated genocide, pursuant to Articles 29 and 180(1) of the BiH CC.⁴⁷

Albeit the overall wording is not unambiguous, it is possible to interpret the Indictment to the effect that the Prosecution intended charging the accused with all of the crimes resulting from the implementation of the genocidal plan, and not simply with the crimes committed in the Kravica operation in which they participated. The Court certainly held this wider interpretation when it discussed and rejected JCE.

The Court noted that the JCE pleaded by the Prosecution was identical in membership and purpose to the one pleaded by the ICTY Prosecution in Blagojević and Jokić. It was so despite the changed pleading practice reflected in the Indictment against Popović and others, where only the military and civilian leadership are described as members of the JCE, whereas the units who are alleged to have physically perpetrated the crimes are said to have merely participated in its implementation.⁴⁸ In the words of the Court, the ‘Prosecutor has effectively proposed that almost all VRS and [police] personnel that were deployed in the Srebrenica area between 12 and 19 July, from the highest echelon of the officer corps to the lowliest common soldier, together were members of a single joint criminal enterprise.’⁴⁹ Relying on authoritative findings by the Appeals Chamber in Brđanin, the Court held the view that a finding of participation in a JCE entails that each and every JCE member is responsible for each and every crime committed through the JCE.⁵⁰ Accordingly it noted that pursuant to the Indictment the accused ‘are alleged to be criminally responsible for all crimes committed following the fall of Srebrenica’.⁵¹ The Court showed its scepticism by noting that ‘[t]he path from the Accused to officers like General Mladić traversed … the whole length of the chain of command’, and by emphasizing ‘the wide disparity between the scope of the purported JCE and the scope in fact of the Accused's acts’.

Having already established that a genocidal plan existed, that the accused had at least some knowledge thereof, and that through their own acts they had contributed to the implementation of the plan, the decisive point in the Court's analysis was the existence of a plurality of persons acting ‘jointly’ or ‘in concert’. The Court, again relying on the Brđanin Appeals Judgment, correctly stated that ‘[a]s a matter of law, knowing participation in the implementation of the plan of a JCE is not sufficient to establish membership in that JCE.’⁵² The key factor is rather whether the persons shared that criminal purpose, that is to say that they ‘had joined together to realize that criminal purpose’.⁵³ ‘Joint action requires some degree of reciprocity, mutuality, or bi-directionality’, which was not shown by the evidence, hence it was not established that the accused acted ‘jointly’ or ‘in concert’ with the members of any JCE.⁵⁴

Consequently, even assuming arguendo the existence of the JCE alleged by the Prosecution, the accused could not be found to have been members thereof and thus cannot be held responsible for the crimes committed pursuant to that JCE. As the Court eloquently stated:

[t]hose who conceived and directed the criminal plan that was implemented following the fall of Srebrenica are criminally responsible for all crimes that ensued. The common soldiers of the VRS and the [police], on the other hand, are responsible for the crimes they participated in, and no more. To conclude otherwise would be to assign collective responsibility to all soldiers for the crimes of their superiors, a notion absolutely repugnant to national law, international criminal law, and the law of war.⁵⁵

The Court did not shy away from acknowledging that the ‘Accused's actions implemented in part the common purpose or plan of the JCE’. Nevertheless, recalling the language used in Krajišnik, it held that they were mere tools of the JCE who had been procured to commit crimes.⁵⁶

Had the Prosecution alleged a JCE with a more narrowly-defined purpose (i.e. the plan to kill over 1,000 Bosniak males at the Kravica warehouse) and membership (i.e. the members of the 2nd special police detachment who took part therein), which I will call ‘Kravica JCE’, then it is likely that the JCE allegation would have succeeded. It is assumed that the broader JCE allegation was motivated by a wish to ensure that the facts of the trial were not seen in isolation and that their significance to the bigger picture was properly noted. This could have been obtained by alleging that the Kravica JCE was inter-linked to the wider ‘Srebrenica JCE’. Such an approach of using separate but inter-linked JCEs has previously been advocated in this journal.⁵⁷

(b) Co-perpetration

Having dismissed JCE, the Court found the six accused who participated in the killings at the warehouse guilty of co-perpetrating genocide. The Court found that each one of them through their individual acts had decisively contributed to the perpetration of the criminal offence.

The specific relevant acts were not limited to killing the captives, but included marching them to the warehouse and preventing their escape during the execution. Considering that all the co-perpetrators participated in escorting the captives, but once at the warehouse only four of them participated in the shooting, whereas the other two confined themselves to standing guards at the warehouse doors, the Court needed to elaborate on the gravity of this latter act. Article 29 in fact requires of the co-perpetrators who do not carry out the actus reus of the crime, that their contribution to the joint crime be decisive. The Court found that such act was ‘of equal importance and effect as if [the two specific accused] had fired at the captives. Covering the openings that were the only chance for the captives to survive, at a time when a mass execution was underway at … the warehouse, constitutes a decisive contribution to the act, thus achieving the planned intention — the intention to have no survivors in the warehouse.’⁵⁸

Despite having rejected JCE, the Court referred to both Article 29 and 180(1) of the BiH CC when convicting the co-perpetrators. This author believes that the reference to 180(1) is nugatory and possibly misleading as the Court did not specify what part of that Article it was relying upon (perpetration? perpetration by way of JCE?). A Court which has been resolute in rejecting JCE with well-founded arguments should have been equally resolute in dismissing any reference to the Article purportedly including that notion, or in specifying the intended use of such reference.

C. Command Responsibility

1. Theory

Two accused (Trifunović, platoon commander and Stupar, detachment commander) were also charged with command responsibility. Having found that Trifunović was a co-perpetrator, the Court did not need to consider whether he incurred command responsibility. Conversely, Stupar was not present at the execution site and the Court based his conviction solely on command responsibility.

Command responsibility in the BiH CC is regulated by Article 180(2), which mirrors Article 7(3) of the ICTY Statute. Preliminarily the Court satisfied itself of respect for the principle of legality by noting that command responsibility was already applicable in municipal law, moreover it constituted customary international law, at the time of the crime.⁵⁹

Relying on international precedents, notably ICTY jurisprudence, as persuasive authority in the interpretation of Article 180(2), the Court outlined the elements of command responsibility as follows: The Court then turned to analyse the peculiar connotations assumed by the mens rea of command responsibility vis-à-vis genocide. The question at issue was whether the law requires a commander to share genocidal intent in order to find him responsible for genocide on the basis of command responsibility.

1. The commission of a criminal act of the type set out in the applicable sections (which include genocide, war crimes and crimes against humanity).

2. The existence of a superior/subordinate relationship between the Accused and the perpetrators who carried out the criminal act.

3. The superior knew or had reason to know:

   • the subordinate was about to commit the crime; or

   • had committed the crime.

4. The superior failed to take reasonable and necessary measures to:

   • prevent the crime; or

   • punish the perpetrator of the crime.⁶⁰

The Court looked at the jurisprudence on the point, starting from a Trial Chamber decision in Stakić holding that ‘it follows from … the unique nature of genocide that the dolus specialis is required for [command responsibility] as well’.⁶¹ That precedent has remained isolated, and has been contradicted by later jurisprudence. The Court indeed noted that a similar finding made by the Trial Chamber in Brđanin with respect to extended JCE had been reversed by the Appeals Chamber, which stated that it was an error to conflate ‘the mens rea requirement of the crime of genocide with the mental requirement of the mode of liability’ and that extended joint criminal enterprise and command responsibility ‘do not require proof of intent to commit a crime on the part of an accused before criminal liability can attach’.⁶² The Court further noted that the Trial Chamber in Brđanin, relying on the foregoing finding, set the commander's mens rea requirement simply to the level of knowledge: ‘the superior must have known or had reason to know of his or her subordinate's specific intent’.⁶³

Although admittedly obiter dicta, the analysis in Brđanin is comprehensive. The Trial Chamber also reviewed International Criminal Tribunal for Rwanda (ICTR) jurisprudence noting that, having found several accused guilty of genocide on the basis of superior responsibility, the ICTR has never expressly required that they share dolus specialis. Admittedly this case law is not conclusive because, in all but one of the cases, the accused were also found responsible for genocide through a form of primary liability, hence a finding of genocidal intent had been made.⁶⁴ The only accused to be convicted solely on superior responsibility was Samuel Imanishimwe in Cyangugu, however his conviction was overturned on appeal, for unconnected reasons.⁶⁵ Besides reviewing the case law, the Trial Chamber held that, as a matter of statutory interpretation, there is no inherent reason why command responsibility ‘should apply differently to the crime of genocide than to any other crime in the Statute. Superior criminal responsibility requires the Prosecution to establish that a superior knew or had reason to know of the criminality of subordinates. In the case of genocide, this implies that the superior must have known or had reason to know of his or her subordinate's specific intent’.⁶⁶

The Court in Kravica found that ‘none of the various pronouncements by Trial Chambers of the ICTY and ICTR are particularly persuasive authority on this issue, as their conclusions were not necessary to resolve the guilt of the accused’.⁶⁷ Whilst it is true that there are no final findings on the point which are not obiter dicta, this consideration is not in itself sufficient to dismiss them as unpersuasive, as the Court does.

Any possible doubt left after reviewing the case law could have been dissipated by making further analysis into the nature of command responsibility. Despite prior uncertainties, the literature and especially case law tend to now agree that command responsibility is a sui generis form of liability whereby the commander is not held responsible for the crime of his subordinates, but because of such crime, and for his own dereliction of duty.⁶⁸ Accordingly, the conduct and mens rea required of the superior are remarkably different from those required of the perpetrators.⁶⁹ In fact, when considering that the commander's culpability lies in the dereliction of his duty to prevent or punish genocide committed by his underlings, it appears illogical to require that he had to hold a specific intent in neglecting his duties.

Through the foregoing analysis the Court could have certainly provided a negative answer to the question whether a commander needs to share genocidal intent. Nevertheless, it chose to avoid answering the question, effectively shying away from its job of iuris dicere. It stated: ‘it is not necessary in this case to decide whether specific genocidal intent is a necessary element of liability for genocide under command responsibility in every case. It is sufficient in this case that the Panel finds beyond reasonable doubt that Stupar in fact had the specific intent to destroy the Bosniaks from Srebrenica as such’.⁷⁰

This author finds it unfortunate that the Court refrained from stating its interpretation of the law on an essential point for its decision. Moreover, by taking this stance the Court possibly incurred more problems than those it was trying to avert since, in order to satisfy itself of Stupar's own genocidal intent, it had to reach certain findings which, as explained below, are unconvincing or in contrast with other parts of the judgment.

2. Practice

The Court found that Stupar, as Commander of the 2nd Šekovići detachment, was the superior officer of the five accused who were primary perpetrators of the crime, and that he had effective control over each of them. Remarkably, the judgment considered Stupar's command responsibility only with respect to the five subordinates who were convicted in the same verdict with him, and not for Mitrović (who had been tried with them, but whose case had been severed before the pronouncement of the verdict) nor for other possible participants who were not tried but were sufficiently identified by the evidence. As it is well settled that a commander can be held responsible for the acts of subordinates who are not on trial, the reasons for this approach, which are not stated in the judgment, are puzzling.

In assessing Stupar's alleged responsibility for failing to prevent genocide, the Court concluded that ‘[a]lthough the evidence establishes beyond doubt that Stupar knew of the genocidal plan … and that he knew that his unit was part of that mission, there is not sufficient evidence that he knew in advance … that the particular task of some of the members of his detachment would be killing the men in the Kravica warehouse.’⁷¹ On that basis, the Court could not conclude that Stupar ‘knew or had reason to know that this specific crime would be committed by his subordinates.’⁷²

Here the Court misapplied the ‘had reason to know’-test, which it had correctly articulated earlier. The law does not require that the information available to the accused be specific: it is enough that he has ‘some general information in his possession, which would put him on notice of possible unlawful acts by his subordinates’⁷³ and would thus justify further enquiry.⁷⁴ The Court's prior findings that all detachment members (and logically, the Commander in particular) must have had some understanding of the criminal nature of their mission before arriving in Srebrenica might, alone, meet the test. If one adds the findings the Court makes concerning the ‘controlled lawlessness’ that Stupar witnessed on the ground before the executions,⁷⁵ then one cannot see how the ‘had reasons to know’-test was not fulfilled. It seems in reality that the Court required the commander to have specific advance knowledge of the subordinates’ crime, which is not what the law requires.

Despite the foregoing, the Court found abundant evidence that Stupar learned of the crime very soon after its commission.⁷⁶ It was found that he had the authority to punish his subordinates through disciplinary actions, which he failed to do.⁷⁷

The most controversial part of the reasoning is the one devoted to finding that Stupar acted with genocidal intent when failing to punish his subordinates. The Court made extensive findings on the awareness that Stupar would have acquired on the ground in the run-up to the execution, findings which starkly contradict his purported lack of ‘reasons to know’ of possible unlawful acts by his subordinates. The relevant section reads:

Stupar, who was present in the Srebrenica area at least since 12 July, would have had the same opportunity as the other Accused to observe the implementation of the plan to eradicate the Bosniaks who lived in the UN Safe Area by forcible transfer and by killing. In addition, given his position in the command chain … he would have known much more about the plan than the other Accused. … The tasks on both days for the Detachment Stupar commanded was the killing of the 8000 Bosniak men, 1300 of whom were armed. Stupar as commander of the Detachment would have known that this was happening by ambush, shelling, and induced surrender because he would have seen it, and his men and his artillery platoon were involved in it. He would have witnessed the controlled lawlessness whereby acts of violence and abuse and plundering were carried out with impunity by the soldiers on the surrendering Bosniaks.⁷⁸

The analysis continues finding that, based on the same factors which led the Court to conclude that the co-perpetrators had genocidal intent, Stupar too would have become aware of such intent on the part of his subordinates.⁷⁹ Then the Court looked at Stupar's conduct in the days following the execution, noting that he and his troops engaged in the capture of the remaining groups of Muslims men who were trying to flee.⁸⁰

The Court stated that Stupar's awareness of his subordinates’ dolus specialis would rule out any possibility that his failure to punish was motivated by a belief that ‘the offenses [were] trivial or insignificant’.⁸¹ Likewise, his subsequent actions would rule out that it could have been an ‘oversight or negligent omission’.⁸² In the Court's reasoning, ruling out these options is tantamount to finding that Stupar possessed genocidal intent, as if other options such as sheer indifference or supineness were not possible. It is submitted that this reasoning is faulty and shows how, even giving due weight to the extraordinary circumstances surrounding the Srebrenica events, it is much harder, maybe impossible, to infer a specific intent from an omissive conduct than it is from a commissive one.

Possibly aware of the weakness of its reasoning, the Court tried to strengthen it by noting the importance of Stupar's omissive behaviour for the implementation of the overall genocidal plan, since it reinforced the message ‘that there was nothing [the soldiers and policemen] could do to the Bosniaks in violation of national or international law that would not be tolerated with impunity’.⁸³ This message was an important message for the success of the genocidal plan given the number of executions which had to take place in the days that followed, and the number of men needed to carry them out.⁸⁴

Be it as it may, this fact in its objectiveness cannot reinforce Stupar's mens rea, in the absence of a finding that Stupar was aware of the importance of his omissive behaviour for the successful outcome of the plan, that is, of the impending executions, which neither he nor his men participated in. Such a finding is absent and, had the Court made it, it would have set Stupar's awareness of, and contribution to, the plan at a level sufficiently significant to call into question the rejection of JCE liability for this accused.

5. Conclusion

The completion of the first Srebrenica trial in BiH is to be hailed as a great success of the local judiciary in ensuring that justice is dispensed for the Srebrenica atrocities. This is all the more important in light of the impending closure of the ICTY. When compared with other trials for Srebrenica, the Kravica case is inevitably characterized by the fact that the defendants were mere tools at the bottom of the chain of command. Nevertheless, as the trial revealed, they were ruthless and enthusiastic tools, keen to implement the genocidal plan conceived by their superiors. Bringing this kind of actors to justice is important for closing the impunity gap. However, as the case teaches, the Prosecution must be careful in confining charges against similar defendants to what they can realistically be blamed for.

A review of the judgment shows that the Court of BiH uses the case law of the ad hoc tribunals as highly persuasive authority. This is done very competently in areas where the case law is well developed, as shown by the analysis of the substantive law of genocide and of JCE. The section on genocide law sets out an original and sound test for genocidal mens rea. The section regarding JCE, albeit not setting any innovative interpretation, applies principles which were only recently outlined in the case law; it is an important piece in the ongoing effort to map the precise contours of JCE liability.

Conversely, when the Court ventures in relatively unchartered waters, its steps become somewhat less confident. The Court's uncertainty and consequent refusal to state its interpretation of the law on the mens rea requirement of command responsibility for genocide is regrettable. It is so not only because the Court had to embark on unconvincing reasoning in order to infer a specific intent from an omissive conduct, but also because it missed an opportunity to contribute to developing an area of the law where precedents are still scarce.

1 Panel composed of judges Hilmo Vučinić, presiding, Paul Melchior Brilman and Shireen Avis Fisher, and Dženana Deljkić-Blagojević as legal adviser. The Court of Bosnia and Herzegovina [hereinafter ‘Court of BiH’, or ‘the Court’], has jurisdiction over the crimes defined in the Criminal Code of Bosnia and Herzegovina [hereinafter ‘BiH CC’], which include war crimes, crimes against humanity and genocide. For additional information, cfr. www.sudbih.gov.ba.

2 First Instance Verdict, Stupar and others (X-KR-05/24), Court of Bosnia and Herzegovina (First Instance Panel), 29 July 2008 [hereinafter ‘Stupar and others’]; First Instance Verdict, Mitrović (X-KR-05/24-1), Court of Bosnia and Herzegovina (First Instance Panel), 29 July 2008 [hereinafter ‘Mitrović ’] First Instance Verdict, Stevanović (X-KR-05/24-2), Court of Bosnia and Herzegovina (First Instance Panel), 29 July 2008.

3 Decision on Severance of Proceedings, in Stupar and others, at 321–326.

4 They are: Milenko Trifunović, Aleksandar Radovanović, Slobodan Jakovljević, Branislav Medan, Brano Džinić and Petar Mitrović.

5 Miloš Stupar.

6 They are: Velibor Maksimović, Dragiša Živanović, Milovan Matić and Miladin Stevanović.

7 German courts have convicted several low- to middle- level perpetrators of genocide in Bosnia. The convictions include that of Nicola Jorgić, Maksim Sokolović and Djuradj Kusljić. The only other Tribunal which has convicted anyone for genocide solely based on command responsibility is the ICTR, in the cases of Samuel Imanishimwe (his conviction was reverted on appeal) and Aloys Ntabakuze (convicted by the Trial Chamber; the appeal is pending). See infra notes 63 and 65.

8 For a more detailed account of the events see Stupar and others, at 26–51.

9 I will use this term to refer to the six accused who were found guilty of genocide for co-perpetrating the execution at the Kravica warehouse.

10 Stupar and others, at 118.

11 The only exception is Milovan Matić, allegedly a VRS member. Since he has been acquitted, his different position is irrelevant.

12 Cfr. Indictment, Trbić (KT-RZ-139/07), Prosecutor's Office of Bosnia and Herzegovina, 20 July 2007.

13 Cfr. Indictment, Pelemiš and another (KT-RZ-74/08), Prosecutor's Office of Bosnia and Herzegovina, 20 November 2008.

14 Cfr. First Instance Verdict, Božić and others (X-KR-06/236), Court of Bosnia and Herzegovina (First Instance Panel), 6 November 2008; First Instance Verdict, Todorović (X-KR-06/180-1), Court of Bosnia and Herzegovina (First Instance Panel), 22 October 2008; Indictment, Tomić (KT-RZ-143/07), Prosecutor's Office of Bosnia and Herzegovina, 18 August 2008; Indictment, Vuković (KT-RZ-53/08), Prosecutor's Office of Bosnia and Herzegovina, 26 August 2008.

15 Stupar and others, at 117.

16 Ibid. (emphasis added).

17 Ibid., at 185.

18 Emphasis added.

19 Stupar and others, at 185–187.

20 Cfr. Partial Decision as to the Admissibility of Application no. 76965/01 by Hisham Ibrahim Karmo against Bulgaria, European Court of Human Rights (First Section), 9 February 2006.

21 Stupar and others, at 53.

22 Ibid., at 54–58.

23 Ibid., at 58.

24 See e.g. Judgment, Jelisić (IT-95-10), Appeals Chamber, 5 July 2001, § 47, listing: ‘the general context, the perpetration of other culpable acts systematically directed against the same group, the scale of atrocities committed, the systematic targeting of victims on account of their membership of a particular group, or the repetition of destructive and discriminatory acts’. Also see Judgment, Brđanin (IT-99-36), Trial Chamber, 1 September 2004, §§ 971–989: ‘the court looked at four factors: (a) the extent of the actual destruction; (b) the existence of a genocidal plan or policy; (c) the perpetration and/or repetition of other destructive or discriminatory acts committed as part of the same pattern of conduct; (d) the utterances of the accused’.

25 Stupar and others, at 58–59.

26 Ibid., at 55. On the qualification of the contribution as co-perpetration, see below.

27 Ibid., at 60–61.

28 Ibid., at 102.

29 Ibid.

30 Ibid., at 117.

31 Ibid., at 63.

32 Ibid., at 117.

33 Ibid., at 63 and 105–112.

34 Ibid., at 112–117.

35 Ibid., at 121 (emphasis added).

36 Ibid., at 117; Mitrović, at 107.

37 A. Cassese, ‘The Proper Limits of Individual Responsibility under the Doctrine of Joint Criminal Enterprise’, 5 JICJ (2007) 109–133, at 115.

38 Ibid.

39 Emphasis added. Note that the article's heading is erroneously translated into English as ‘accomplices’, whereas a more accurate translation would be ‘co-perpetrators’.

40 A. Cassese, International Criminal Law (2nd edn., Oxford University Press, 2008), 188–189.

41 Verdict, Vuković and another (KRŽ-07/405), Court of Bosnia and Herzegovina (Appellate Division Panel), 2 September 2008, at 6, note 1. No importance is given to the use of the word perpetrating instead of committing.

42 See: Ibid., at 5, stating that JCE and co-perpetration ‘are mutually exclusive and their coexistence is not possible’. The same judgment states that it is unclear if JCE is a crime or a mode of liability but if the latter is true then the difference with co-perpetration lies ‘particularly in the field of mens rea, since joint criminal enterprise implies common intent on the level of subjective element … while co-perpetration implies the principle of limited responsibility’, at 6. Compare with: Verdict, Rašević and another (X-KRŽ-06/275), Court of Bosnia and Herzegovina (Appellate Division Panel), 6 November 2008, at 27, stating that the ‘distinction between co-perpetration and participation in a JCE is that a larger degree of contribution (that is, more decisive) is required for co-perpetration’.

43 Amended Indictment, Stupar and others (KT-RZ-10/05), Prosecutor's Office of Bosnia and Herzegovina, 24 June 2008, at 4.

44 Ibid., § 4.

45 Ibid., at 2. The only act that does not relate to the execution is described in paragraph (a), alleging that the accused ‘secured the road and kept it closed or open for traffic in line with the [forcible transfer] plan’; this allegation was dismissed by the judgment.

46 Ibid., § 4.

47 Ibid., at § 1 and at 4, § 2.

48 Stupar and others, at 172, note 387.

49 Ibid., at 172.

50 Ibid., at 173, note 389. The Court relies on § 445 of the Brđanin Appeals Judgment: ‘A coherent application of such a notion could make each one of the [relevant physical perpetrators], as members of the JCE, responsible for each one of the crimes that the Trial Chamber found were committed throughout the territory of the ARK during the Indictment period.’ The Court's interpretation of this authority is believed to be correct, subject to the caveat that responsibility for crimes within the JCE plan would attach automatically to all JCE members, whereas responsibility for foreseeable crimes would only attach to those members who had foreseen them.

51 Ibid., at 172–173.

52 Ibid., at 174. The Court relies on § 410 of the Brđanin Appeals Judgment.

53 Ibid. (emphasis added).

54 Ibid., at 175.

55 Ibid., at 173.

56 Ibid., at 175.

57 K. Gustafson, ‘The Requirement of an ‘Express Agreement’ for Joint Criminal Enterprise Liability: A Critique of Brđanin', 5 JICJ (2007) 134–158, at 145–149.

58 Stupar and others, at 134.

59 Ibid., at 135–140.

60 Ibid., at 140.

61 Decision on Rule 98bis Motion for Judgment of Acquittal, Stakić (IT-97-24-T), Trial Chamber, 31 October 2002, § 92.

62 Decision on Interlocutory Appeal on Defence Motion for Acquittal, Brđanin (IT-99-36-A), Appeals Chamber, 19 March 2004, § 7.

63 Brđanin, supra note 24, §§ 720–721; also Judgment, Blagojević and Jokić (IT-02-60), Trial Chamber, 17 January 2005, § 686. For a further precedent, altough not yet available at the time of the Kravica judgment, see Judgment and Sentence, Bagosora and others (ICTR-98-41-T), Trial Chamber, 18 December 2008, §§ 2127–2154. See in particular §§ 2135, 2139, 2144, 2149, 2154 where only awareness of the principals’ intent is discussed in the context of finding Théoneste Bagosora and Aloys Ntabakuze responsible as superiors for genocide. This precedent is of particular importance because Ntabakuze was found guilty for genocide only through command responsibility.

64 Brđanin, supra note 24, §§ 717–719.

65 Judgment and Sentence, Ntagerura, Bagambiki and Imanishimwe (‘Cyangugu’) (ICTR-99-46-T), Trial Chamber, 25 February 2004, § 654.

66 Brđanin, supra note 24, § 720 (footnote omitted).

67 Stupar and others, at 162–163.

68 For the case law see: Declaration of Judge Shahabuddeen, Judgment, Orić (IT-03-68), Appeals Chamber, 3 July 2008, §§ 18–26; Judgment, Halilović (IT-01-48), Trial Chamber, 16 October 2007, §§ 42–54; Judgment, Krnojelac (IT-97-25-A), Appeals Chamber, 17 September 2003, § 171. For the literature see Cassese, supra note 40, at 147 and 243. Cassese argues that it is not sound to hold a commander who fails to punish genocide by his subordinates responsible for the crime of genocide since his culpability arises from the ‘intentional, reckless, or grossly negligent breach of his supervisory duties’. Cassese allows for the possibility to hold a commander as participant in the crime when he willingly fails to prevent the commission of the crime.

69 Cassese, ibid., at 243.

70 Stupar and others, at 163.

71 Ibid., at 156.

72 Ibid.

73 Ibid., at 153.

74 Ibid., at 154.

75 Stupar and others, at 164. On these findings, see further below.

76 Ibid., at 154.

77 Ibid., at 161.

78 Ibid., at 164.

79 Ibid., at 165. Incidentally, this finding meets the mens rea test articulated in Brđanin and Blagojević, which the Court never expressly adopted.

80 Ibid., at 166–167.

81 Ibid., at 165.

82 Ibid., at 167.

83 Ibid.

84 Ibid.