An Assessment of Judge de Prada’s dissent, in the Karadžić Appeals Judgment, on genocide in seven Bosnian municipalities

2 May 2019

On 20 March 2019, the Appeals Chamber of the International Residual Mechanism for Criminal Tribunals (IRMCT) issued its long-awaited judgment on the appeals filed by Radovan Karadžić and the Prosecution against the ICTY Trial Judgment rendered on 24 March 2016.

Much of the Appeals Judgment in Karadžić affirmed the trial chamber’s findings and convictions save for: reversing Karadžić’s convictions related to the Overarching Joint Criminal Enterprise (“JCE”) – to permanently remove Bosnian Muslims and Bosnian Croats from Bosnian Serb-claimed territory – to the extent that they were based on certain Scheduled Incidents; and setting aside the sentence of 40 years of imprisonment in place of a sentence of life imprisonment. On both of those issues Judge de Prada dissented.

Judge de Prada also dissented to the Majority’s finding that Count 1 – the charge relating to genocide in seven municipalities in 1992 – was properly decided by the trial chamber.

The two aspects of Count 1 that the Prosecution appealed, that are of particular interest and are considered in this article, were the trial chamber’s findings that:

  • certain crimes, some of which constituted the actus reus of Count 1 genocide (such as some crimes of persecution in count 3 and murder/execution in counts 4, 5, and 6), were not part of the common plan of the Overarching JCE to permanently remove Bosnian Muslims and Bosnian Croats from Bosnian Serb‐claimed territory – and that Karadžić did not intend for these crimes (“Excluded Crimes”) to be committed; (Prosecution Ground 1) and
  • it had insufficient evidence (after assessing the pattern of crimes as well as the specific statements and conduct of Karadžić and other members of the Overarching JCE) to find that Karadžić and other members of the Overarching JCE possessed ‘specific’ intent to destroy (i.e. genocidal intent), as charged. (Prosecution Ground 3)

On both Grounds 1 and 3, the Majority found, Judge de Prada dissenting, that the Prosecution had failed to demonstrate any error in the trial chamber’s assessment. On both, the appeals chamber’s reasoning was that it could not conclude that a reasonable trier of fact would have inferred, as the only reasonable inference, that either: (a) some or all of the Excluded Crimes formed part of the common plan; or (b) the conduct and statements of Karadžić and other JCE members reflected an intent to destroy the Bosnian Muslim and the Bosnian Croat groups, as such, in the seven Count 1 municipalities.

Basis of Judge de Prada’s dissent

Judge de Prada’s dissent on the lack of a positive finding on specific intent [at paras 337-341] was, in many ways, predicated on his earlier dissent on the Majority’s finding that certain crimes (which also formed the actus reus of genocide) were excluded. [paras 325-337] In his view Karadžić must have had mens rea for the Excluded Crimes and that they, therefore, should have been within scope of the Overarching JCE on the basis that the test for the assessment of evidence relating to inferences was not met due to: (a) a failure by the Majority in providing a reasoned and, therefore, verifiable opinion (pursuant to Art 23(2) ICTY Statute and 21(2) IRMCT Statute and the Rules of both) of their evidentiary assessment i.e. an inference that the Excluded Crimes were within scope was not the only reasonable inference; [paras 806-808] and (b) the mistaken belief of the Majority that the availability of an, or any, alternative inference drawn from the evidence puts into question the only reasonable inference. In the view of de Prada, the alternative must be reasonable enough to cause reasonable doubt and here it simply was not “truly contradictory or incompatible”. [paras 800-808, 813]

On specific intent, Judge de Prada’s dissent was based on his assessment that the Majority misunderstood the legal test for specific intent and that the evidentiary standard did not accord with the required reasonableness standard due to, principally, a lack of reasoning. [paras 830-834] Judge de Prada cited the extensive literature on the ambiguity inherent in the concept of specific intent – explicitly citing the tension between the purpose-based interpretation as opposed to the knowledge-based interpretation. [paras 835-836] (See further, Ambos’s piece critically considering the debate). He preferred to settle on a combination of a purpose-based intent requirement rooted in ‘objective bases’ which consider: the underlying ‘genocidal acts’; and ‘the effective contribution of the perpetrator to the collective destruction of the protected group’ in instances where they were not a principal perpetrator. [paras 837-840] His rationale is that a purely intent-based requirement is impossible to fulfil and the concept ‘almost inapplicable’; retaining only symbolic character.

Based on his clarification, Judge de Prada found that the trial chamber limited itself to a formal approach to assessing the evidence without deeply analysing the ‘elements and peculiarities of the present case’. He found that many of the ‘alluded indicators’ used as evidence of specific intent were already present and appeared as proven in the findings of the Trial Judgment itself (at paras 2595 to 2624 and 2634 to 2903). Yet, in his view, those indicators were not sufficiently considered and the trial chamber, regrettably, established an evidentiary test on the basis of purely quantitative criteria. [paras 840-842] Judge de Prada stated that: “another result could have been achieved from a correct assessment of the’ ensemble of the evidence together with more objective and pragmatic criteria regarding the requirements of intent “to destroy in whole or in part”. For example, taking into account the presence of the alluded genocidal indicators and the certainty of knowledge on the part of the accused that his acts or omissions were contributing to the collective destruction of a group.” [para. 843]

Significance of Judge de Prada’s dissent

Judge de Prada was right that the Karadžić Trial Judgment provided limited reasoning for reaching both of its findings relating to Grounds 1 and 3; explaining that it was not satisfied that the inference was the only reasonable one but not explaining precisely why (at paras 2605, 2610-2612, 2613 and 2625-2626). Most trial judgments of the ad hoc tribunals tend to provide limited reasoning on the application of complicated areas of the law like mens rea for genocide partly because: the sheer volume of evidence precludes detailed reasoning either on the factual or legal findings; and agreement on elaborate reasoning of nebulous concepts like specific intent (see my earlier piece on definitional issues relating to genocide here) is very difficult to obtain – on conclusions, for obvious reasons, it is rather easy. Furthermore, appeals chambers tend to presume that a trier of fact must have taken all relevant factors into account when assessing evidence even if it did not do so expressly because limited reasoning allows it to do little else unless, of course, the findings were plainly without foundation. The Majority here [at paras 669-673] did much the same.

Judge de Prada is also correct – notwithstanding a Defendant’s presumption of innocence and the requirement for the Prosecution to prove its case beyond doubt – that any mere inference cannot be a serious source of doubt to a principal inference unless it is truly capable of raising a doubt. The problem is that the Majority (who may not necessarily disagree with de Prada’s reasoning) come to a different result; for them the trial chamber was entitled to find that Karadžić did not have mens rea for the Excluded Crimes on the basis that it was not the only reasonable inference – why that is exactly so is unclear and unverifiable as de Prada’s alternative assessment and inference (and conclusion) [at paras 814-819] makes clear. His alternative conclusion was that the Excluded Crimes were a means of achieving the common plan (as that plan developed), that Karadžić’s unique position at the apex of executive power lent support to the carrying out of these crimes; and that those crimes endured over a number of months – relevant factors that were, in his view, plainly ignored. The difficulty, in circumstances in which no detailed elaboration of the trier of fact exists as to how exactly it reached its conclusions, is that we will never know whether any alternative was the only reasonable one.

On specific intent, Judge de Prada is attempting to bring in line the specific intent requirement closer to an ordinary and practical understanding of what ‘intent to destroy’ would mean for a high-level perpetrator. That is by allowing for a purpose-based test and requiring that the context and circumstances in which the actus reus are carried out are thoroughly considered. That is not controversial and both the Karadžić and Mladić trial chambers cited the relevant case law requiring that the pattern of the underlying crimes ought to be thoroughly considered (at para. 2592 Karadžić, at para. 3435 Mladić); whether or not it was, in fact, is the issue. Unfortunately, Judge de Prada then departed to conflate part of the mental requirement for mode of liability under JCE 1 (“significant contribution”) with the test for specific intent. For JCE 1 all participants must share the specific intent and must ‘significantly contribute’ to the common plan (which involves the commission of a crime). Specific intent itself, however, has no requirement for ‘effective contribution’. Reading-in that requirement could both weaken (if it displaces completely the purpose-based test) or perversely strengthen (in other forms of perpetration other than JCE 1) the mens rea required. Certainly, conduct of the perpetrator ought to be a consideration in whether it is possible to infer that a perpetrator had specific intent but using the language of modes of liability (and specifically co-perpetration) would not assist in clarifying an already obscure mens rea concept for genocide. Treating relevantly different perpetrators (based on status, role, and capability) differently can be achieved through the mode of liability charged.

Implications, if any, of the dissent going forward

Judge de Prada’s dissent may not seem particularly remarkable to long-standing academic scholars of genocide. It is of significance, however, as it sheds light on the very real tensions any judge seriously grappling with applying the mens rea element of genocide to a real factual matrix faces but rarely explicitly acknowledges. Practically, the dissent brings to light the shortcomings of a loosely reasoned judgment (echoing Judge Flavia Lattanzi’s dissent lamenting the inadequately reasoned opinion of the Majority in the Šešelj trial judgment) and the consequences of applying an inherently ambiguous legal concept without a clear attempt to remove the ambiguity.

The broader significance relates to the form and content of a judgment relating to mass atrocity crimes containing a clear and unequivocal dissent. (See further, the excellent piece on radical dissents in ICL by Neha Jain). De Prada’s dissent acts as a public and transparent means of displaying judicial discourse on the limits of the law in such cases, the difficulties in interpretation of (mis-)understood legal concepts and providing a counter-narrative to hotly disputed law (if not facts). In an age where the international institutions purporting to dispense international criminal justice are themselves under attack any introspective analysis of procedural or substantive shortcomings (if that is what they are) by the institutional actors, sworn to impartiality, themselves is to be welcomed for its legitimising function; that can only serve to dispel the myth of monolithic institutions dispensing, without much thought, selective justice and help shape future international criminal legal discourse.

This article was written by Aarif Abraham, an instructing pupil barrister at Garden Court North Chambers.


Image courtesy of United Nations International Criminal Tribunal for the former Yugoslavia (UNICTY) – The first session of the ICTY held in Nov 1993.



Injunctions in the Court of Protection – what are the limits?

The extent of the Court of Protection’s power to grant injunctions has not always been clear, to judges and lawyers alike. Practitioners will recall that...


International Women’s Day 2023

As part of our celebration of this year’s International Women’s Day we are highlighting Anna Morris – an exceptional barrister who has recently been appointed...


What is it like to be a GCN Practice Manager?

We asked our practice managers, Christian Wolfenden and Michael Johnston to share their experiences of the role and how it differs from other places they...


High Court Allows Judicial Review Claim on Universal Credit Overpayment Recovery

R. (on the application of K) v Secretary of State For Work and Pensions [2023] EWHC 233 (Admin) Anyone with any experience trying to negotiate...

Sign up to our mailing list

Our mailing list is dedicated to professionals with an interest in our work.

Sign up