1. Introduction

Recent judgments issued by the International Criminal Tribunal for the former Yugoslavia (ICTY) and by the Special Court for Sierra Leone (SCSL) dealing with aiding and abetting international crimes have triggered unprecedented debate on the risks of fragmentation (at times also, and more positively, characterized as ‘pluralism’[1]) within international criminal law. In this paper, I first attempt to identify the terms of the debate by briefly discussing what is meant by fragmentation and how this may be said to apply to the issue of aiding and abetting, with special regard to the cases under consideration. I then move to explain why in my view fragmentation in this field may bring about negative results in terms of the practice of international criminal courts and tribunals, which must be rooted in the protection of the rights of the accused I conclude with some thoughts for a way forward.


2. Fragmentation

a) The issue in international law

Over the past two decades at least, scholars and practitioners have been engaging in heated debates on the risks of fragmentation within international law ‘writ large’. In 2006, the International Law Commission published a Study Group on Fragmentation of International Law finalized by Professor Martti Koskenniemi (Koskenniemi Report).[2]

Fragmentation is often described as the possibility of conflicts arising within and between different rules and legal systems in various areas of international law, leading to differing views about the content of general law. Self-fragmentation would then refer to a subset of this phenomenon, namely when fragmentation occurs over time (or even at the same time) within a single institution. Fragmentation is of course a common occurrence in any legal system, but its consequences for the international legal system – which lacks a proper institutional hierarchy – are particularly problematic, as they appear complex to reconcile.[3]

The Koskenniemi Report opines, with a paragraph that might apply to both fragmentation and self-fragmentation:

‘Differing views about the content of general law create two types of problem. First, they diminish legal security. Legal subjects are no longer able to predict the reaction of official institutions to their behaviour and to plan their activity accordingly. Second, they put legal subjects in an unequal position vis-à-vis each other. The rights they enjoy depend on which jurisdiction is seized to enforce them. Most domestic laws deal with these problems through the instrumentality of the appeal. An authority (usually a court) at a higher hierarchical level will provide a formally authoritative ruling. Such authority is not normally present in international law’.[4]

For the purpose of the present paper, it is worth noting that the Koskenniemi Report specifically highlighted that ‘[a]t an institutional level, the proliferation of implementation organs – often courts and tribunals – for specific treaty-regimes has given rise to a concern over deviating jurisprudence and forum-shopping’.[5] International criminal law commentators have specifically adverted to the fact that ‘the contribution of [international criminal] courts to the gradual specification and precision of legal rules (…) suffers from the major shortcoming that such judicial refinement is decentralized and fragmentary (…). Hence, the possibility frequently arises of a contradictory or “cacophonic” interpretation or application of international criminal rules’.[6]

Various suggestions have been put forward to reconcile these differences, especially those emanating from courts and tribunals – including that of International Court of Justice (ICJ) judge Bennouna, who wrote:

‘The substantive law that each court “produces” can cause disparities in the interpretation and application of international law. For the purposes of coordination, one should consider the creation by the UN General Assembly of a new expert body, modelled on the International Law Commission, which would ensure “the progressive development of international law and its codification”. The task of this expert body would be to identify and analyse the potential divergences in the interpretation and application of general international law, as well as the consequences thereof’.[7]

Whether this would be a valid – or even feasible – solution is beyond the scope of this paper. It is, however, very interesting that a sitting judge of the ICJ perceives fragmentation to be a problem warranting an overall institutional solution under the aegis of the United Nations.


b)  Fragmentation and self-fragmentation in International Criminal Law

Having briefly discussed the general concept of fragmentation in international law, it is necessary to identify where exactly the risks of fragmentation and self-fragmentation may stem in international criminal law. First, however, a caveat is in order. A discussion of this kind necessarily assumes – rather than setting out to demonstrate – the existence of a true system of international criminal law, gradually developed by, and through, the judicial institutions created since the last decade of the 20th century. Indeed, without some sort of harmonious system it would make little sense to even discuss fragmentation, for this phenomenon presupposes the existence of a somewhat coherent ‘whole’ subject to being fragmented. The following reflections are therefore grounded in the findings – by both academics and international judges – on the existence of an international system of individual criminal accountability.[8]

International criminal courts and tribunals make extensive use of customary law and, to a lesser extent, general principles of law to interpret and apply the crimes and modes of liability listed in their constitutive instruments, many of which provisions are the same or similar across institutions.[9] Due to the complexity of ascertaining the content of customary rules and general principles of law, conflicts can easily emerge when different judges confront similar situations and reach different conclusions. This creates fragmentation of international criminal law, i.e., a divergent application of (what would otherwise appear to be) the same rules to different accused.[10]

Some correctives already exist, of course. Although the Statutes and Rules of Procedure and Evidence of the ICTY and the International Criminal Tribunal for Rwanda (ICTR) are silent in relation to the value of precedents, one of the ICTY Appeals Chambers’ first judgments stated that ‘in the interests of certainty and predictability, the Appeals Chamber should follow its previous decisions, but should be free to depart from them for cogent reasons in the interests of justice.’ It added that ‘a proper construction of the Statute requires that the ratio decidendi of its decisions is binding on Trial Chambers’.[11] Moreover, the ICTY and ICTR Appeals Chambers – though formally distinct bodies – are composed of the same individuals, acting as ICTY judges or ICTR judges, depending on the case. This has served to ensure a very high degree of consistency between the two institutions. Another example is Article 20(3) of the SCSL Statute, further discussed below, which explicitly states that ‘[t]he judges of the Appeals Chamber of the Special Court shall be guided by the decisions of the Appeals Chamber [rectius, Chambers] of the International Tribunals for the former Yugoslavia and for Rwanda’.[12]

This state of affairs has generally served the international criminal courts and tribunals well. While there have been a few exceptions (especially with respect to the controversial notion of joint criminal enterprise), overall the various international criminal courts and tribunals, and even the ‘internationalised’ ones such as the Extraordinary Chambers in the Courts of Cambodia and the Special Tribunal for Lebanon, have been paying close attention to each others’ case-law, and have ensured a coherent and consistent application of the law. Even the International Criminal Court (ICC) heavily relies, in its reasoning on both substantive and procedural matters, on the case-law of other international criminal courts and tribunals,[13] despite its relatively close system dictated by Article 21 of the Rome Statute.[14] Mindful that the ICC has issued only three substantive judgments and no appellate final judgment as of March 2014, and so caution must be used in predicting how it will interpret its applicable law in the future, this practice does support the proposition that each of these judicial bodies is – or considers itself to be, which is essentially the same for practical purposes – part and parcel of a ‘family’ of institutions, a network of independent (but evidently inter-dependent) courts and tribunals. Discrepancies have, overall, occurred quite rarely.

However, the few important exceptions to consistency demonstrate that fragmentation – and, at times, self-fragmentation – undoubtedly occurs in international criminal law. Such inconsistencies have generated heated debate amongst legal scholars, as well as judges. The 2013-2014 saga related to aiding and abetting is one of these uncommon, but arguably worrisome, instances of fragmentation.


3. Modes of liability in international criminal tribunals

The terms of the judicial ‘debate’ on modes of liability relevant to the present discussion may be summarized as follows. From its seminal Tadiccase, the ICTY has attempted to distinguish principal and accessory liability. Faced with the wording of its own Statute’s Article 7(1) – identical to Article 6(1) of the ICTR and the SCSL Statutes – the judges attempted from the start to develop a framework to ensure that personal culpability (rather than forms of objective or collective liability) be used as a foundation of criminal responsibility.[15] The relevant provision reads:

‘A person who planned, instigated, ordered, committed or otherwise aided and abetted in the planning, preparation or execution of a crime referred to in Articles 2 to 5 of the present Statute, shall be individually responsible for the crime’.

On the basis of the wording of this provision, the judges could essentially have chosen one of two different interpretative paths. On the one side, they could have read Article 7(1) as a whole, following a unitary model and thus considering each of the modes of participation listed (planning, instigating, ordering, etc.) simply as different ways of contributing to the substantive crimes – and that anybody who participated in any way in these crimes would thus simply be ‘responsible’. This approach would have paralleled some domestic criminal codes, such as the Italian one,[16] and some international instruments,[17] under which anybody who participates in criminal conduct is deemed to have committed the crime – or rather, to have ‘participated’ in the crime. The degree of each accused’s responsibility is, in such a framework, considered only at the stage of sentencing, and not when imputing liability.

However, the ICTY judges chose to read Article 7(1) differently, using what may be termed a ‘differentiated’ model. Under this approach, judges are called to make a finding on the specific mode of participation in the crime prior to and, actually, in order to determine responsibility. If a person orders a crime, he will for instance be considered ‘more responsible’ than an individual who has merely aided and abetted the crime. Under this conceptual framework, it is not enough for judges to be satisfied as to a certain (significant) degree of participation in the crime in order to find an accused responsible. To convict, they must identify a precise mode of liability under which the conduct of the accused can be subsumed, or face the prospect of their conviction being overturned.[18]

This framework led the Tadic bench to distinguish between perpetration (which – simplifying the matter to the extreme – included co-perpetration by a group of people with a common purpose, i.e., joint criminal enterprise (JCE))[19] and accessory liability, i.e., aiding and abetting.[20] The analysis of the modes of liability enshrined in Article 7(1) (and Article 6(1) of the ICTR and SCSL Statutes) by the various chambers after Tadic has then been based on attempts to find a customary basis for each one of the modes of liability listed, since the terse language of the Statutes did not provide for other types of guidance. This was a reasonable choice for the judges to ensure that – once the customary nature of the forms of responsibility was ascertained on the basis of precedents and convergence in domestic systems – there would not be any injustice for the accused.[21] However, as further explored below, this approach created the possibility that different judges would come to different conclusions as to the precise content of custom in relation to certain elements of the relevant modes of liability.

For the purpose of this discussion, one must bear in mind that in Tadic the judges were focusing on co-perpetration as a way to ensure that the accused would not escape punishment for the killing of five men from the village of Jaskici even though there was no evidence that he personally killed any of them.[22] In doing so, they also discussed as obiter, and simply in order to contrast it with JCE, the regime applicable to aiders and abettors, as accessories to a crime perpetrated by another person (the principal).[23] The recent debate on aiding and abetting at the ICTY and SCSL emanates from these premises.


4. Perisic, Taylor, Lazarevic: specific direction or not?

As hinted at above, the ICTY identified ‘aiding and abetting’ as one of the possible modes of responsibility enshrined in Article 7(1) of its Statute. Indeed, the statutes of all international criminal tribunals have included jurisdiction over aiding and abetting responsibility, without, however, spelling out its constitutive elements.[24] This was more fully undertaken by the Furundžija Trial Chamber, which reviewed the status of customary international law on the point,[25] and was then followed by other cases.

On the basis of this and subsequent case-law, the objective element (actus reus) of aiding and abetting was considered to be, according to international customary law, the provision of assistance to a crime by making a substantial contribution to it, or having a substantial effect on its perpetration. ‘Substantial’ was intended to mean that without the assistance in question, the crime most probably would not have occurred in the same way.[26] For example, the provision of the means to commit crimes, such as the delivery of weapons, has often been considered a substantial contribution.[27] In relation to the subjective element (mens rea), the aider and abettor must have known that the assistance would substantially contribute to the perpetration of the crime.[28] Thus, the aider and abettor must have accepted the risk that – given the circumstances in which assistance was provided – one or more possible crimes could be committed.[29] Moreover, knowledge of the essential elements of the crime was required.[30]

On 28 February 2013 the ICTY Appeals Chamber in Perisic appeared, according to several commentators, to add a new requirement to the objective element of aiding and abetting, at least when the accused is found to be physically remote from the crime(s) actually committed.[31] It stated that the conscious provision of assistance to a broad range of activities– including, but not limited to, international crimes – is not enough to qualify as aiding and abetting, even if that assistance ended up substantially contributing to the commission of the crimes. Only if the contribution was ‘specifically directed’ to the commission of the crimes charged would the conduct of the remote aider and abettor be sufficiently linked to the final crime.[32] This seemed incongruous with cases previously decided, such as Brdjanin and Krstic, where convictions had been entered for aiding and abetting crimes committed at substantial physical distance from the aider and abettor, but no mention of specific direction was made. Nevertheless, the present article does not concern itself with the substance of the matter (i.e., the law of aiding and abetting), but rather with the problems inherent when the Appeals Chamber of the same institution reaches different legal results due to its changing composition.

Momcilo Perisic, Chief of Staff of the Serbian Army (based in Belgrade), had facilitated the provision of weapons and the secondment of personnel to the Bosnian-Serb Army, which used this assistance in the commission of crimes in Sarajevo and Srebrenica. The Appeals Chamber (by a majority of three judges to two) stated that, given Mr Perisic’s remoteness from the crimes, it should have been shown that his assistance was directed specifically and unequivocally to the crimes committed by the Bosnian-Serb Army, not to the Bosnian-Serb war effort in general. The majority of the Appeals Chamber based its legal reasoning on an excerpt from the Tadic Appeal Judgment affirming that the aider and abettor ‘carries out acts specifically directed to assist, encourage or lend moral support to the perpetration of a certain specific crime.’[33]

Just a few months after the Perisic judgment was issued, the SCSL Appeals Chamber rejected the specific direction requirement in the Taylor judgment. Charles Taylor, former President of Liberia, had been accused of aiding and abetting crimes committed by rebel groups in Sierra Leone. Mr Taylor had seized the opportunity provided by Perisic to argue that due to his physical remoteness from the crimes, and because the Trial Chamber had not found ‘specific direction’, he should be acquitted.[34] In other words, since the evidence led at trial had not demonstrated beyond reasonable doubt that Mr Taylor specifically directed his acts of assistance to the rebel groups in question towards the perpetration of the crimes charged, his conviction for aiding and abetting should be quashed. The SCSL Appeals Chamber rejected these arguments. The SCSL in Taylor recognized that Article 20(3) of its Statute requires it to be ‘guided’ by ICTY and ICTR Appeals Chambers’ decisions.[35]However, after an analysis of customary international law and of the SCSL Statute, the Appeals Chamber found that nothing required specific direction for aiding and abetting liability and that it was not obliged to follow ICTY case-law.[36] Having rejected the need for specific direction, the SCSL Appeals Chamber therefore affirmed Charles Taylor’s conviction for aiding and abetting.[37] This legal finding raises the question of the meaning of Article 20(3) for the SCSL Appeal judges, who clearly paid careful attention to it, openly discussing its text and import in the judgment itself, but ultimately came to the conclusion that other courts’ decisions are not binding, but only persuasive, authorities.[38] This was but an honest recognition of the fact that the SCSL Appeals Chamber – institutionally distinct from the Appeals Chambers of the ICTY and ICTR – is the body entrusted with ensuring internal coherence of the law within the SCSL, but has no duty to ensure horizontal harmony with other tribunals. The choice of the term ‘guidance’ by the drafters of the SCSL Statute may hardly be interpreted to mean that the Court is bound blindly to follow ICTY and ICTR precedents, even when its judges believe that customary law does not support such precedents. While the choice to create a completely separate judicial body for Sierra Leone, not even a distinct tribunal within the same United Nations, might be criticized, this preference necessarily entailed the risk of discrepancies in the interpretation of the applicable law, unless at least the founding instruments of the SCSL explicitly required the judges to follow ICTY and ICTR precedents, and created an institutional mechanism to ensure compliance; which they did not.[39]

The issue of specific direction, however, was not to be put to rest by the Taylor Appeal Judgment. On 23 January 2014, in the Sainovic et al case, a partially differently composed bench of the ICTY Appeals Chamber agreed with the SCSL and reversed the ICTY position adopted eleven months earlier in Perisic. In Sainovic, four accused had been convicted for crimes committed in Kosovo in 1999. The issue of aiding and abetting was relevant to the case of Mr Lazarevic, convicted by the Trial Chamber for his knowing contribution to the crimes of deportation and forcible transfer. The Appeals Chamber, by majority of four judges (Liu, Güney, Pocar, and Ramaroson) found the legal finding in Perisic to be at odds with previous case-law. It noted that, prior to the Perisic Appeal Judgment, ‘no independent specific direction requirement was applied by the Appeals Chamber to the facts of any case before it,’ but rather only ‘substantial contribution as an element of the actus reus has consistently been required.’ Then, after having reviewed the relevant sources, the Appeals Chamber reached the conclusion that ‘specific direction’ was not an essential element of the actus reus of aiding and abetting in customary international law.[40]

As noted by Coco and Gal, the ICTY had generally refrained from analysing whether the provision of assistance was specifically directed to a crime prior to Perisic. In Blagojevic, for instance, the ICTY Appeals Chamber explained that the finding of specific direction ‘will often be implicit in the finding that the accused has provided practical assistance to the principal perpetrator which had a substantial effect on the commission of the crime’.[41] This implies that ‘specific direction’ is actually part of the substantial contribution requirement, and not a separate element.[42] When answering the accused Jokic’s specific ground of appeal on the existence of a specific direction requirement, the Blagojevic Appeals Chamber almost exclusively dealt with the substantial effect of the contribution itself and found that the exercise of routine duties (not ordinarily directed to any particular crime) does not exculpate the accused if he has nonetheless provided a substantial contribution to the crime.[43] In this sense, it could be said that Sainovic does not really intend to depart from the Perisic precedent. It rather purports to ‘correct’ the mistake made there – since, in its own words, ‘the interpretation given in the Perisic Appeal Judgement would appear to be at odds’ with previous case-law, thus leading the new Appeals Chamber to the ‘compelling conclusion that “specific direction” is not an element of aiding and abetting’, and confirming the legal standard ‘constantly and consistently applied’ prior to Perisic.[44] Yet, even the Sainovic judgment does not necessarily signal an end to the debate on specific direction; the issue remains alive, and nothing precludes specific direction from being revived by another bench, or being contradicted by future ICC decisions dealing with accomplice liability.[45] In fact, the ICC– though primarily applying its own Statute, especially in the field of liability – does look at other international tribunals’ application of custom to interpret its own statutory provisions in relation to the elements of the crimes, the modes of liability, and procedural matters. The interaction between the findings by the ICTY and SCSL on the customary nature of aiding and abetting liability, on the one side, and future ICC decisions on accomplice liability, on the other, might engender further judicial debates on the matter.


5. Inconsistencies or Fragmentation?

The expectation that a legal system will provide certainty in order to protect legitimate expectations (‘to keep the scale of justice even and steady, and not liable to waiver with every new judge’s opinion’, as the eminent English jurist William Blackstone wrote in the 18th century)[46] lies at the very heart of the doctrine of precedents. The proliferation of international judicial institutions, especially in the criminal field, undoubtedly tends to foster a decentralized decision-making process. This in turn necessarily subjects each judge to a test of congruence between the decisions reached in one case and those reached in other similar instances, whether within the same institution – such as the ICTY Perisicand Sainovic benches – or by different institutions – for instance, the ICTY and the SCSL. In such circumstances, as discussed above, the lack of formalized rules on precedents outside the case-law itself creates the risk of inconsistencies within the international criminal justice system (assuming that this does exist) that are hardly predictable. While some of these inconsistencies occur within the case-law of one institution (such as the ICTY) and others mar the system across different courts and tribunals, as it were, both these instances can be labeled as ‘fragmentation’, since the same legal concept (in this instance, aiding and abetting) is interpreted differently over time – and thus legal subjects are no longer able to predict the reaction of official institutions to their behaviour and to plan their activity accordingly.[47]

However, this danger should not be overestimated; the very fact that the Taylor and Sainovic cases have been discussed at length as possible examples of fragmentation shows that the system is actually perceived as a whole, and that – when discrepancies arise – these are noted by practitioners and academics as problematic. In a sense, only if and when discrepancies become the norm would one stop talking about the danger of fragmentation; there would not be a system to be fragmented anymore! Until and unless this occurs, the system is still perceived as essentially one, with some uncertainties around the edges.

To be sure, different voices and interpretations are not always harmful; dissenting voices – or dissenting chambers – can show progress and ways of thinking ‘outside the box’, better aimed at securing the objectives of international criminal justice. Similar to what occurs in domestic criminal justice, certain notions develop gradually. Evolution of case-law is not necessarily negative, as it may allow the law to adapt to societal changes, as long as the rights of the accused are preserved.[48] But there is no doubt that inconsistencies come at a cost, both at a systemic level and at a personal level (for the accused). While there is no formal rule of stare decisis in international (criminal) law, there are at least two good reasons for international courts and tribunals to be cautious about sudden changes in their jurisprudence, and to consistently endeavour to build on previous case-law when arriving at legal findings.

In the field of international criminal law the benefits of predictability and ‘legal security’ (to use the expression of the Koskenniemi Report) are arguably even more important than in other fields, for they necessarily concern the fundamental human rights related to a fair trial and the nullum crimen principle. It would indeed appear essential for any accused before an international court – before any court of law, really – to be able to reliably foresee his or her fate by perusing the law and previous similar cases. Furthermore, with a growing recognition of the importance of victims’ participation in international criminal proceedings, it is important to note that predictability of verdicts also impacts this additional important ‘constituent’ of the system of global justice.

Predictability and legal security are, moreover, arguably very important for the international community as a whole, which is expending substantial resources in building what is clearly meant to be a coherent system of international accountability of the individual. Coherence of the system would be lost if its participant actors came to expect future inconsistencies due to significant sudden changes in the case-law. This would be the case even if, in practice, such inconsistencies only seldom occurred; the mere expectation that drastic changes in the case-law may occur in the future would create uncertainties, thus greatly undermining confidence in the system. While the value of predictability must always be balanced against other important rights and interests – especially when a wrong legal finding may create gross unfairness to an accused – any legal system rooted in the rule of law must be premised on a certain amount of coherence.

This is plainly a problem that goes well beyond the issue of aiding and abetting at the ICTY and SCSL (and, potentially, in the future, at the ICC). This controversy may actually point to a significant shortcoming of the current structure of international criminal justice, where no higher authority has the responsibility to ensure nomofilachia. While at first sight this might be considered positive for the judges involved in actual decision-making at the various courts and tribunals, who are allowed much wider (because ultimately unchecked) discretion in their interpretation of the law than their domestic counterparts, the reality is much more complex. Having a safeguard of last resort over the wrongful application of the law would likely reassure judges who are dealing with complex facts and (often) novel legal issues. These judges might actually feel supported if their choices – often controversial – could be referred, in one way or another, to a higher legal authority, with narrow competence and more general expertise.


6. Conclusions – A more careful way forward for international criminal judges?

The reflections in the preceding paragraphs suggest that if international criminal institutions are aware of previously decided cases, and remain at the same time flexible in reaching specific conclusions on the facts under consideration, they will enhance their own effectiveness and, hence, their legitimacy.[49] Each international criminal court or tribunal is of course its own institution, independent from the others – but formal (the ICTY and ICTR Appeals Chambers; the provisions of Article 20(3) of the SCSL Statute) and informal (cross-citing of legal authorities, comparative analysis by judges and scholars) mechanisms exist to ensure interdependence and, if possible, consistency among such institutions. There is no doubt that all of these judicial bodies form part of the family of international criminal justice, created as they are for similar purposes and often acting in the name of the United Nations or other substantial groups of States.

Consistency is generally good judicial policy – though of course it should not be a primary concern of international criminal judges who are called upon to judge on the individual criminal responsibility of the specific accused indicted and brought to trial before them. However, domestic authorities are often called to take into account other legal policy considerations, when individualizing penalties, for instance, or (in the systems that allow this) in deciding to proceed with all or some of the possible charges. Due to the uncertainty that this creates, and the potential consequences for the accused and for the victims (and for the whole system of international criminal justice), there would appear to be good reasons for international judges to adopt an extremely careful approach when contemplating departure from precedents, even if these are not found to be formally binding.

There is undoubtedly some merit in Judge Shahabuddeen’s statement in his declaration appended to the OricAppeal Judgment:

‘[a] decision to reverse turns upon more than theoretical correctness; it turns upon larger principles concerning the maintenance of jurisprudence, judicial security and predictability. Included in those principles is, I believe, a practice for a judge to observe restraint in upholding his own dissent. […] I do not assert that a dissenting judge can never form part of a subsequent majority upholding his earlier dissent, but I think that the preferred lesson of the cases is that he is expected to do so with economy’.[50]

Commentators have been discussing other means to ensure a degree of consistency. These could take the form of new supra-national bodies, along the lines of Judge Bennouna’s proposal in the first paragraphs of this paper (though not specifically aimed at international criminal tribunals, of course). Or, they could be internal mechanisms, such as a sort of ‘en banc’ or ‘Grand Chamber’ procedure summoning the collective expertise of the whole (or the vast majority) of the appellate body of the judicial institution in question.

As regards the ICC, discussions in this area will likely emanate from Article 39(2)(b)(i) of the Rome Statute (i.e., ‘[t]he Appeals Chamber shall be composed of all the judges of the Appeals Division’), which according to some would provide the basis for an inclusive approach aimed at involving all appeal judges in appellate decision-making. However, there is no easy solution. Judges may be conflicted due to previous activities within the same court (as has already been the case with the ICC Appeals Chamber, whose judges have previously sat on Pre-Trial Chambers dealing with the same situations), and thus might need to be replaced by other judges. More importantly, international judges’ terms of office expire, and they are routinely replaced by new judicial appointees, leading to a reshuffling of the Pre-Trial, Trial, and Appellate sections of the ICC.

Therefore, there is no sure way for a court to ensure a benign consistency aimed at protecting the rights of the accused and all of the other interests involved, other than striving for judicial self-control in making changes to the existing case-law. External structures would too likely evolve into ‘strictures’, stifling the healthy development of the law to deal with unforeseen circumstances. More to the point, institutional mechanisms for achieving greater consistency will only realistically work if the overall judicial culture of the decision-makers is imbued with a healthy sense of restraint and careful assessment of the law as already applied previously, whether within the same institution or elsewhere.


[1] E van Sliedregt, ‘Pluralism in International Criminal Law’ (2012) 25 Leiden J Intl L 847; Teitel and Howse, ‘Cross-Judging: Tribunalization in a Fragmented but Interconnected Global Order’ (2009) 41 NYU J Intl L Politics 959. The rationales for the creation of a multitude of different courts to handle international claims are also explored in C McLachlan, Lis Pendens in International Litigation (Martinus Nijhoff 2009) 305 ff.

[2] ILC, ‘Fragmentation of International Law: Difficulties Arising from the Diversification and Expansion of International Law’ UN Doc A/CN.4/L.682 (13 April 2006), as corrected UN Doc A/CN.4/L.682/Corr.1 (11 August 2006) (by Martti Koskenniemi).

[3] Koskenniemi Report (n 2) para 26.

[4] Koskenniemi Report (n 2) paras 52-53 (emphases added).

[5] Koskenniemi Report (n 2) para 489.

[6] A Cassese, P Gaeta et al, Cassese’s International Criminal Law (OUP 2013) 29.

[7] M Bennouna, ‘How to Cope with the Proliferation of International Courts and Coordinate Their Action’, in A Cassese (ed), Realizing Utopia: The Future of International Law (OUP 2011) 287.

[8] Most recently, see the references to international criminal law as a ‘body of law’ and a ‘branch’ of public international law in Cassese, Gaeta et al. (n 6) 3-21; see also R Mulgrew, Towards the Development of the International Penal System (CUP 2013). On the relevance of external precedents in this ‘system’, see also A Zammit Borda, ‘Precedent in International Criminal Courts and Tribunals’, (2013) 2 Cambridge J Intl Comparative L 287. For the assumption that international criminal liability necessitates a system of international criminal justice, see also: US v Otto Ohlendorf (1950) 6 Trials of War Criminals before the Nurnberg Military Tribunal under Control Council Law No. 10, at 462 (‘because they are accused of having offended against society itself, and society, as represented by international law, has summoned them for explanation’); ICC Statute, Preamble (‘Mindful that during this century millions of children, women and men have been victims of unimaginable atrocities that deeply shock the conscience of humanity, Recognizing that such grave crimes threaten the peace, security and well-being of the world (…), Determined to put an end to impunity for the perpetrators of these crimes and thus to contribute to the prevention of such crimes, (…)Resolved to guarantee lasting respect for and the enforcement of international justice’, etc).

[9] On the relevance of custom for, eg, war crimes, see Prosecutor v Tadic (Decision on Defence Motion for Interlocutory Appeal on Jurisdiction) ICTY-94-1-A (Appeals Chamber, 2 October 1995) para 94 and Prosecutor v Galic (Judgment) ICTY-98-29-A (Appeals Chamber, 30 November 2006) para 85. Among scholars dealing with the interaction between the principle of legality and the use of customary international law, see K S Gallant, The Principles of Legality in International and Comparative Criminal Law (CUP 2009) 11-30; L Gradoni, ‘Nullum crimen sine consuetudine: A Few Observations on How the International Criminal Tribunal Has Been identifying Custom’,  in Agorae Papers of the ESIL Inaugural Conference (Florence, 13-15 May 2004), available at <www.esil-sedi.eu/sites/default/files/Gradoni_0.PDF>.

[10] Much of the inconsistencies are however at the procedural level due to the different procedural provisions, but also to what has been called a ‘zone d’ombre entourant le processus interprétatif’. J de Hemptinne, ‘Hybridité et autonomie du Règlement de procédure et de preuve du Tribunal pénal international pour l’ex-Yougoslavie’, in M Delmas-Marty, E Fronza and E Lambert-Abdelgawad (eds), Les sources du droit international penal (Société de législation comparée 2004) 135-155.

[11] Judgment, Prosecutor v Aleksovski (Judgement) ICTY-9416/1-A (Appeals Chamber, 24 March 2000), paras 107-113, 123. The ICTR Appeals Chamber adopted the same position shortly thereafter in Prosecutor v Semanza.

[12] Interestingly, and contrary to the text of the Article, at least one SCSL Trial Chamber held that ‘the provision equally applies to triers of fact at first instance’: Prosecutor v Brima et al. (Judgement) SCSL-04-16-T (Trial Chamber, 20 June 2007) (Brima Judgment) fn 1269.

[13] See, for instance, the discussion on enlistment and conscription of children in Prosecutor v Lubanga (Judgment) ICC-01/04-01/06 (Trail Chamber I, 14 March 2012) paras 603-631 (where the ICC Trial Chamber recognizes that ‘SCSL’s case law therefore potentially assists in the interpretation of the relevant provisions of the Rome Statute’).

[14] The ICC Statute requires the Court to apply, in the first place, the Statute, elements of Crimes and the Rules of Procedure and Evidence. Only in the second place ‘where appropriate’ applicable treaties and custom, and then general principles of law derived from domestic legislation (art 21(1)). The Statute then adds that the ‘Court may apply principles and rules as interpreted in its previous decisions’ (art 21(2)). The wording of this provision, a clear attempt at compromise between two opposing approaches, shows that no agreement was reached among the drafters as to whether to bind the ICC to its previous holdings. Nor there is any explicit provision on the force of appellate decisions vis-à-vis first instance decision-makers. See O Triffterer, Commentary on the Rome Statute of the International Criminal Court (2nd edn Hart 2008) 711; W Schabas, The International Criminal Court – A Commentary on the Rome Statute (OUP 2010) 394-397.

[15] See Prosecutor v Tadic (Judgment) ICTY-94-1-A (Appeals Chambers, 15 July 1999) (Tadic Appeal Judgment), para 186.

[16] See art 110 of the Italian Penal Code (‘Where more persons participate in the same crime, each of them is liable to the penalty established for that crime (…)’) (unofficial translation by the author).

[17] In Control Council Law No. 10 – issued by an international body, the Allied Control Council in Germany after World War II, but giving rise to domestic military trials in the various zones of occupations – for example, art 2(2) provided: ‘Any person (…) is deemed to have committed a crime as defined in paragraph 1 of this Article, if he was (a) a principal or (b) was an accessory to the commission of any such crime or ordered or abetted the same or (c) took a consenting part therein (…)’.

[18] For a brief analysis of the benefits and pitfalls of both approaches, see A Cassese et al., International Criminal Law: Cases and Commentary (OUP, 2011) 323 ff. Along the same lines, see also H Vest, ‘Problems of Participation – Unitarian, Differentiated Approach, or Something Else?’ (2014) 12 J Intl Criminal Justice 295.

[19] Tadic Appeal Judgment (n 15) paras 195-225; ICTY, Judgment, Prosecutor v Brdjanin (Judgment) ICTY-99-36-A (Appeals Chamber, 3 April 2007) paras 415-437.

[20] The academic debate about modes of liability in international criminal law is of course immense. See, among others, J G Stewart, ‘The End of “Modes of Liability” for International Crimes’ (2012) 25 Leiden J Intl L 165; J D Ohlin, ‘Joint Intentions to Commit International Crimes’ (2011) 11 Chicago J Intl L 693; S Manacorda and C Meloni, ‘Indirect Perpetration Versus Joint Criminal Enterprise: Concurring Approaches in the Practice of International Criminal Law?’ (2011) 9 J Intl Criminal Justice 159.

[21] The issue of whether the analysis carried out by the ICTY, ICTR, and SCSL did always satisfy the requirements for a finding of the customary status of certain notions, such as aiding and abetting and joint criminal enterprise, goes beyond the scope of the present article.

[22] TadicAppeal Judgment (n 15) para 185.

[23] ibid para 229. The propensity of some international judges to add dicta to their legal findings arguably heightens the risks for fragmentation, thus leading to possible conflicts.

[24] See art 6(1) ICTR Statute and art 6(1) SCSL Statute, containing the same provision.

[25]Prosecutor v Furundzija (Judgment) ICTY-95-17/1-T (Trial Chamber, 10 December 1998) (Furundzija Trial Judgment) paras 193-226 (actus reus) and paras 236-241 (mens rea).

[26]Prosecutor v Tadic (Judgment) ICTY-94-1-T (Trial Chamber, 7 May 1997) para 688.

[27]Prosecutor v Ntakirutimana (Judgment) ICTR-96-10-A and ICTR-96-17-A (Appeals Chamber, 13 December 2004) para 530.

[28]Prosecutor v Mrksic (Judgment) ICTY-95-13/1-A (Appeals Chamber, 5 May 2009) para 49 and Prosecutor v Oric (Judgment) ICTY-03-68-A (Appeals Chamber, 3 July 2008) (Oric Appeal Judgment) para 43.

[29]Brima Judgement (n 12) para 776; Furundzija Trial Judgment (n 25) para 246.

[30] Furundzija Trial Judgment (n 25) para 246; see also Prosecutor v Krstic (Judgment) ICTY-98-33-T (Appeals Chamber, 2 August 2001) para 140.

[31] On the academic debate in the ‘blogosphere’, see also: M Milanovic, ‘The Self-Fragmentation of the ICTY Appeals Chamber’ (at <www.ejiltalk.org/the-self-fragmentation-of-the-icty-appeals-chamber/>); S Vasiliev, ‘Consistency of Jurisprudence, Finality of Acquittals, and Ne Bis in Idem’ (at <http://cicj.org/?page_id=1608>); K J Heller, ‘The SCSL’s Incoherent – and Selective – Analysis of Custom’ (at <http://opiniojuris.org/2013/09/27/scsls-incoherent-selective-analysis-custom/>); A Fielding, ‘Charles Taylor Appeal: Why Its Rejection of ‘Specific Direction’ Doesn’t Matter’ (at <http://beyondthehague.com/2013/09/30/charles-taylor-appeal-why-its-rejection-of-specific-direction-doesnt-matter/).

[32] Prosecutor v Perisic (Judgment) ICTY-04-81-A (Appeals Chamber, 28 February 2013) paras 25-44.

[33] Tadic Appeal Judgment (n 15) para 229. In Perisic, judges Ramaroson and Liu contested the existence of the specific direction requirement, while judges Meron, Agius and Vaz voted to adopt it. However, Judge Ramaroson joined the majority on the acquittal because in her view the evidence was insufficient to confirm the conviction of Mr Perisic.

[34] Prosecutor v Taylor (Judgment) SCSL-03-01-A (Appeals Chamber, 26 September 2013), paras 467-471 (Taylor Appeal Judgment).

[35] See supra, Section 2.b).

[36]Taylor Appeal Judgment (n 34), paras 472-475.

[37]ibid para 481.

[38] ibid para 472.

[39] An institutional mechanism would appear necessary in this respect, due to the fact that SCSL judges – just like the judges of all other international courts and tribunals – ‘shall be independent in the performance of their functions, and shall not accept or seek instructions from any Government or any other source’ (art 13(1) SCSL Statute), a principle they must reconcile with the ’guidance’ provided by the ICTY and ICTR Appeals Chambers.

[40] Prosecutor v Sainovic (Judgement) ICTY-05-87-A (Appeals Chamber, 23 January 2014) (Sainovic Appeal Judgment) paras 1617-1651. Judge Tuzmukhamedov dissented on the issue of specific direction.

[41] Prosecutor v Blagojevic (Judgment) ICTY-02-60-A (Appeals Chamber, 9 May 2007) para 189.

[42]ibid, as confirmed in Sainovic Appeal Judgment (n 40), para 1625.

[43]See A Coco and T Gal,‘Losing Direction: the ICTY Appeals Chamber’s Controversial Approach to Aiding and Abetting in Perisic’(2014) 12 J Intl Criminal Justice 345.

[44] Sainovic Appeal Judgment (n 40), paras 1621, 1649-1650.

[45] The relevant provision (according to which ‘for the purpose of facilitating the commission of such a crime, aids, abets or otherwise assists in its commission (…)’) might prompt ICC judges to ‘revive’ the specific direction requirement – in this case however it would not be a case of fragmentation, but rather of interpreting of applying the text of the ICC Statute as different from custom. No unfairness to the accused or other party and participant would ensue.

[46] W Blackstone, Commentaries on the Laws of England, vol 1 (1765, facsimile version Legal Classics Library, 1983) 69.

[47] Koskenniemi Report (n 2) para 52.

[48] See, for instance, the renowned finding by the European Court of Human Rights (in C R v UK (1995) 21 EHRR 363) according to which ‘[t]he gradual clarification of the rules of criminal liability through judicial interpretation from case to case complies with the legality principle provided that the resulting development is consistent with the essence of the offence and could reasonably be foreseen’.

[49] See, mutatis mutandis, G Acquaviva and F Pocar, ‘Stare Decisis’, Max Planck Encyclopedia of Public International Law (OUP online edn).

[50]Declaration of Judge Mohamed Shahabuddeen, Oric Appeal Judgment (n 28) para 14.