Category Archives: Comparative Criminal Law

An Open Invitation to Further Debate (Instead of an Amicus Brief)

In organizing this mini-symposium, I sought to engage expert reactions to my paper from a range of legal systems that have not featured in debates about forms of attribution in ICL. When international courts and tribunals construct(ed) these forms of attribution, they initially drew heavily on Anglo-American jurisdictions, adopting concepts like superior responsibility and joint criminal enterprise from them. Then, judges at the ICC announced a major swing towards notions of criminal responsibility derived from German criminal law, including co-perpetration, theories of control to distinguish perpetration from complicity, indirect co-perpetration and even perpetration through an organization to treat those doing the bloodletting and their masterminds as perpetrators. Throughout this process, nobody appears to have asked experts in systems that adopt a unitary theory of perpetration, which dispenses with all these doctrines, to reflect on the law within their own countries and its potential as a solution to recurrent problems with blame attribution in ICL. This silence has been quite strange, especially when the Nuremberg Tribunal applied a unitary theory of perpetration and several modern ICL judges have argued that the current complexity is unnecessary.

This mini-symposium has broken new ground in this regard, in ways that I hope sets the scene for further scholarly research and debate. I was especially grateful that a range of criminal law theorists from each of the countries I write about in the paper agreed to criticize the paper, and that some very prominent practitioners joined the fray to offer their reflections too. As is evident from this blog’s manifesto, I deliberately seek to create dialogue between theorists and practitioners, so I am thrilled that this discussion has involved members of both groups. Some of the feedback I received was striking—during the course of this online symposium, a senior prosecutor at one international court and a defense counsel for a well-known defendant at another emailed me saying they wholeheartedly agree with the need for a unitary theory. The latter even suggested that I file an amicus brief calling on one particular tribunal to revert to the unitary theory of perpetration adopted at Nuremberg. I politely declined, but decided to open up this final post to whomever wanted to share an opinion one way or the other, provided that it respected the strictures of the blog’s manifesto.

Instead of defending the unitary theory of perpetration or either of the article’s I’ve written about it (see here and here) in this post, I use this opportunity to set the scene for an open online discussion at the base of this post by reiterating what a unitary theory is and by summarizing the excellent posts that appeared in this symposium.

To begin, let me again highlight the three main variations of the unitary theory of perpetration to avoid commentators speaking past one another. The unitary theory of perpetration comes in three principal varieties, although some might contest whether the third species really fits within the genus. The first, known as a pure unitary theory, treats a causal contribution to a crime coupled with the requisite blameworthy moral choice announced in the criminal offence charges as necessary and sufficient elements of responsibility (excuses and justifications aside). On this view, the various forms of participation that exist in current ICL (aiding and abetting, JCE, co-perpetration etc.) are stripped of their autonomous existence and folded into a more capacious single notion of attribution. So, instead of attempting to manufacture fine-tuned rules that define JCE, aiding or any other form of participation in such and such a manner, a unitary theory of perpetration places them all in a big pot, then boils them all down to their shared normative essence. Through this distillation, blame attribution involves deciding whether accused X is responsible for crime Y based on settled core principles that pay no regard to the form participation takes, leaving their moral significance to be assessed post hoc by judges at the sentencing phase of a trial.

The second variant provides more detail without compromising the unitary theory’s core commitments. Known as a functional unitary theory, this iteration provides more guidance while insisting that causation and the mental elements announced within the criminal offense charged are necessary and sufficient bases for establishing wrongdoing across all forms of participation. To ensure that would-be criminals are sufficiently forewarned of their exposure to criminal law penalties, a number of states adopt this variant of a unitary concept—the general part of a criminal code or legislation articulates the different forms of causal connections that might apply within a unitary framework. In this sense, responsibility might involve carrying out the offence personally, instructing others to do so, providing necessary assistance, or furnishing assistance that is readily available elsewhere. Each of these forms of causation is announced within the law so as to inform the public of how they might attract criminal responsibility, but the underlying objective and subjective elements beneath these descriptions remain the same.

Third, some argue that subjecting accomplices to the same range of punishment as perpetrators also constitutes a weak type of unitary theory. In Germany (and the many jurisdictions that follow its example), aiders and abettors are sentenced to a maximum of three quarters of the penalty for the offense they facilitate, whereas the sentence for instigators is taken from the same sentencing range as principals. To a large extent, this discrepancy in maximum sentence drives the need for differentiating between perpetrators and accomplices, even if, as Markus Dubber has observed, “[r]emarkably little effort is spent on justifying this differentiation”.[1] Nonetheless, this differentiated approach, whose purpose is partly to determine the applicable range of sentencing, generates a tendency to look upon systems that formally equate sentencing ranges for perpetrators and accomplices as soft iterations of the unitary theory. France and England, for instance, do just this. For my purposes, though, I do not consider this an example of the unitary theory because it places no restriction on the substantive elements of forms of attribution, whereas truly unitary theories do.

With the stage set, I next situate the various expert responses to this mini-symposium, grouping them into those who also advocated for a unitary theory in ICL, those who were more ambivalent about whether their national experience served as much of a template for ICL, and one who was positively unconvinced.

In the first of these categories, Judge Baragwanath’s excellent post reminded us that there are actually many jurisdictions that fit within variants two and three, even if they might not describe themselves as unitary theories of perpetration. My own country of origin, New Zealand, begins the provision governing parties to offences by stipulating that “[e]very one is a party to and guilty of an offence who,” before articulating different forms of participating in a consummated offence. Judge Baragwanath’s post is so useful because it not only highlights that New Zealand’s criminal law is, in important aspects, unitary, but it also shows how a series of cases in England, Australia and Hong Kong have been struggling with whether to tie mental elements in forms of participation to those in the offense announced in ways that mimic the unitary theory. Despite backsliding in some courts, there is a discernable modern trend in this direction. His post reminds me that the States of New York and California have an even more intense unitary theory of perpetration. In any event, in describing “modes of liability” as “unnecessary,” Judge Baragwanath argues that “international criminal procedure, already complex and expensive, adds to those problems by forcing itself to leap over self-created non-existent hurdles.”

Filippo de Minicis’ post is similarly minded. Filippo is a presently Legal Officer in the Office of the Co-Investigating Judges in the Extraordinary Chambers in the Courts of Cambodia, but he was originally trained in Italian criminal law, which as I show in the article, also discarded a differentiated system of blame attribution in favor of a unitary alternative almost a century ago. Filippo argues that when looking at standards of attribution before ad hoc international criminal tribunals (i.e. in customary international law), there is “little difference in the required actus reus,” and “a sufficient homogeneity on the mens rea side.” Filippo concludes after a decade working with these standards that a unitary theory is both viable and preferable, but he is also circumspect about whether any theory is perfect and, as was the case with New Zealand, shows how Italy’s commitment to the unitary theory is not absolute since Italy too appends a type of common purpose doctrine that approximates to JCE. Reality, it turns out, is complicated. Despite this, Filippo joins Judge Baragwanath as an advocate for the adoption of the unitary theory in ICL.

Other commentators are more ambivalent. Professor Carlos Eduardo A. Japiassú, for instance, highlights how Brazil’s unitary theory has slowly changed over time, shifting from a pure version to a functional one (which he calls “mixed”). While Professor Japiassú also speaks of a certain legal conservativism in Brazil, which I take to imply a lack of desire to shift back to a differentiated system that employs different substantive tests for different stand-alone forms of participation, he ends by concluding that “it remains unclear whether a pure rendition of this theory or a mixed variant like that now applicable in Brazilian Penal Law is a good alternative for International Criminal Law or International Criminal Courts.” Similarly, Professors Iryna Marchuk and Jørn Jacobsen discuss important scholarly criticism of the unitary theory in Denmark and Norway respectively as well as partial retreats from it in recent doctrine, before also questioning their system’s value as an exemplar for ICL.

Finally, in the third category, Judge Albin Eser’s masterful critique exemplifies disagreement with the unitary theory. In many respects, his is a brilliantly concise defense of the structure of blame attribution currently in place in ICL and a deft rebuttal of the arguments in the paper. The series of questions he poses are skillfully listed as issues he would need to be convinced of to accept that a unitary theory is optimal. These start with the argument that different forms of participation actually better track real life, move to the idea that a unitary theory cannot justify why they are addressed at sentencing along, then shows how unitary theorist essentially overlook that these questions will arise at sentencing anyhow. Then, he argues that “the only practical advantage the unitary theory so far seems to offer is a procedural one,” but he sees no procedural advantage here either. Ultimately, he concludes by correctly pointing out that even if we do have a differentiated system of blame attribution in ICL because powerful western states forced it on others, this says nothing about the theory’s conceptual integrity. A unitary theorist would, of course, contest each of these steps, but Eser’s brilliant critique is a wonderful counterpoint.

So, instead of labouring my own perspective any further here, I make space for other scholars, experts and practitioners to weigh in on these debates, which strikes me as a better idea than filing an amicus brief. I have therefore opened this post to comments, and anyone can post their views directly. In order to help ICL practitioners share their views (I recall many hours debating these questions with colleagues in war crimes tribunals), I’d like to offer a procedure through which you can legitimately (I hope) bypass the need for institutional approval to publish. If your institution is agreeable, I will post thoughts and reflections from practitioners anonymously. I would not normally do this through the post, so if want to remain anonymous, please send me your comments by email at Your email message to me should include your title and the institution you work for, but I undertake to keep this information entirely confidential, posting only your thoughts and reactions on this topic. For the rest of you, the post is open.

[1] Markus D. Dubber, ‘Criminalizing Complicity: A Comparative Analysis’, (2007) 5 Journal Int Criminal Justice 984 ff.

Questions From the Unconvinced

Albin Eser is Director Emeritus and Professor Emeritus of Criminal Law, Criminal Procedure, and Comparative Criminal Law with the Law Faculty of the University of Freiburg.  He was formerly a Judge at the International Criminal Tribunal for the former Yugoslavia.

To be satisfied with the unitary theory, you must be content with the most simple solution. In my view this is not the best approach to cope with the variety of social and criminal life. As soon, however, as you wish to pay attention to the varieties of the performances of crimes, you must be prepared to make differentiations – and this may unavoidably be a source for different demarcations. But even if this entails controversies, can this be a reason to sacrifice individual justice (by distinguishing between perpetration and participation) on the altar of (unitary) simplicity?

To become convinced of the latter alternative, I would like to have these questions answered:

  • Regarding the argument that the differentiation models would not correspond to real life, is this more the case if causation is considered the only basis of imputation, thereby ignoring any differences in the manner in which, and in the weight by which, a contribution is made to the performance of a crime?
  • And if there is a factual and social difference, as can hardly be denied if the view is not normatively and holistically preprogrammed, should this difference be paid attention to only at the sentencing stage or should it rather be expressed already in the guilty verdict?
  • And if only on the sentencing level, how and according to which criteria should this be done? This is a question, by the way, which is neither dealt with in your article nor are any suggestions by your proponents of the unitary theory visible.
  • If indeed, however, even the unitary theory cannot avoid taking notice of different types and manners in which a person can be involved in the performance of a crime, why couldn’t and shouldn’t this be made public in the verdict?
  • So the only practical advantage the unitary theory seems to offer so far is a procedural one: that the problem of recharacterization may be avoided. Yet, is this really the case if even according to the unitary theory differences in the crime performance are (to be) made at the sentencing stage? In a murder case, for instance, what could a defendant do if he had defended himself by merely involuntary having told where the victim might be met but finds himself sentenced to life imprisonment because of being proven to have been on the scene and directly involved in the killing? Could his appeal be rejected by simply referring to his causal contribution, thus rendering any differential circumstances irrelevant? Or shouldn’t it rather be in the interest of individual justice to assess the penalty according to the type and weight of his contribution? However, if he thus succeeded in getting his sentence adjusted to his minor contribution, as I think he should, could this be done without differentiations and recharacterizations? So what, in the end, is finally left of assumed procedural advantages of the unitary theory?
  • Regarding your references to political and ideological superpowers smaller countries were able to free themselves from by introducing the unitary theory, should such sovereignty aspects indeed play a role as to whether criminal theory is good or bad?

A Unitary Theory is Both Viable and Preferable

Filippo de Minicis works as a Legal Officer in the Office of the Co-Investigating Judges in the Extraordinary Chambers in the Courts of Cambodia. After graduating in law in 2004, he practiced criminal law in Italy. From 2006 to 2013, he worked first as a defence consultant and then as a Legal Officer in Chambers at the International Criminal Tribunal for the former Yugoslavia.

I agree with Professor Stewart’s proposition that a unitary model of blame attribution would be a viable and preferable option for ICL. This opinion is inevitably rooted, at least in part, in my Italian legal training. It is, however, also based on my ten years’ experience as a practitioner in ICL. Because my professional experience is limited to Italy, the ICTY, and the ECCC, my focus will be on participation as applied in those fora.[1] My observations will cover three areas: 1) the issue of the Fascist origin of the Italian Penal Code; 2) the reasons why I favour a unitary system to the current model; and 3) one aspect of the Italian system that Professor Stewart will probably not like.

 1.  Codice Rocco – An essentially liberal code in Fascist attire

Professor Stewart notes that the Fascist origins of the Italian Penal Code (known as Codice Rocco) have led some to dismiss it as a possible model for a unitary theory of participation. Stewart rejects this criticism. I do too. The Codice Rocco is divided in two parts: a general part, which contains the general principles of criminal law, such as causation, culpability, and participation; and a special part, which defines crimes and prescribes penalties. It is in the latter that the Fascist footprint is most evident, both for the nature of certain crimes (e.g. article 265 of the Code prohibits “political defeatism”) and for the order in which they appear, which echoes the hierarchy of values of the Fascist regime. The general part, however, opens by introducing the principle of legality, the prohibition of retroactive application criminal law, and the prohibition on recourse to analogy as fundamental principles of criminal law. These are principles that predate the Fascist ideology and which were part of the legal education of those who drafted the code. They stand in stark contrast to the general principles governing German criminal law during the National Socialist rule, which in substance left to the judge the freedom to incriminate whoever exhibited a behaviour that appeared to conflict with the interests of the National Socialist Party. The Codice Rocco is thus essentially a sufficiently liberal and still modern code, and while there are some articles of the general part which were influenced by the Fascist ideology, the unitary system is not one of them.

2.  A unitary model is possible and may have practical advantages

Italian law does not distinguish between principal perpetrators and accomplices. By contrast, both the ICTY and the ECCC adopt such a distinction. As an Italian lawyer trained to examine criminal liability through a monistic lens, I see the modes of liability applied by the ad hoc tribunals as an unnecessarily prescriptive characterization of different ways of participating in a crime. I hold this view because:

  • Apart from superior responsibility,[2] all modes require that the accused, by ordering, instigating, planning, or aiding and abetting,[3] substantially contributed to a crime, or that her conduct had a substantial effect on its commission.
  • Similarly, for Joint Criminal Enterprise (JCE), the law requires that the accused made a significant contribution to the common criminal plan.
  • With respect to the subjective element, they require that the accused made this contribution with the intent that the crime be perpetrated, or with the awareness of the substantial likelihood that a crime would be committed as a consequence of her conduct.
  • For aiding and abetting, the required mens rea is knowledge that the aider’s conduct will substantially contribute to the commission of the crime.
  • There is thus little difference in the required actus reus,[4] and a sufficient homogeneity on the mens rea side (knowledge, in my view, would also be a sufficient standard to attribute liability through a unitary model – see the next paragraph).

These all appear to be good reasons to consider whether the existence of a variety of modes of liability with an identical or sufficiently similar mens rea requirement is necessary. While some differences, especially if we consider the ICC Statute, do remain in the current universe of modes of liability internationally, they are of a nature that they can be reconciled in a unitary model, especially with appropriate adjustments in the sentencing law and practice. The means with which a person causes or contributes to a crime are not generally determinative of criminal liability strictu sensu. At most, they may be factors relevant for sentencing.

A word about the knowledge standard for aiding and abetting. Leaving aside Professor Stewart’s correct observations that some judgments of the ad hoc tribunals actually require a recklessness standard for aiding and abetting (i.e. knowledge of the probability that her conduct will assist the commission of the crime),[5] the knowledge standard as interpreted by the ad hoc tribunals would suffice, in Italy, to convict an accused for an intentional crime, i.e. one requiring dolus. By way of example, if I want to kill my business associate, and I ask a friend of mine to lend me his gun informing him that I need it to perpetrate the murder, if eventually I commit the murder both my friend and I will answer for it as perpetrators. I see no dogmatic problem with this approach: knowing that our conduct will cause or is likely to cause certain consequences, and still engaging in such conduct, essentially amounts to an intentional act, irrespective of whether those consequences were also desired.

In my experience, it is often practically difficult to distinguish whether a person contributed to a crime with knowledge (in the sense of awareness that one’s conduct will (likely) produce certain consequences) or intent (in the sense of knowledge with the added element of the desire to achieve certain consequences).[6] This, in turn, renders it problematic to legally characterize one’s responsibility as commission through a JCE or as aiding and abetting.[7] Let’s take the example of an alleged JCE the criminal objective of which is the deportation of an ethnic minority. To establish liability through JCE, the prosecution will need to prove that the accused, in concert with others, significantly contributed to the deportation with the intent to commit that crime. In practice, however, it may be difficult (or even impossible) to distinguish this situation from that of someone whose conduct (e.g. the provision of buses to deport the minority) substantially contributed to the deportation with knowledge that the provision of those means would assist that crime, but without the desire to achieve that result: in this scenario, the accused would incur liability via aiding and abetting. In ICL cases, there is seldom a “smoking gun” that allows a simple determination of one’s state of mind and, especially in JCE cases, intent is often inferred from a combination of knowledge and continued participation. This system, however, is far from infallible. Personally, I am not convinced that it is always (or even often) possible to distinguish if a person acted with mere knowledge or intent. Consequently, the law of the ad hoc tribunals features a theoretical distinction that the judges may not be able to apply – beyond reasonable doubt – to the facts under their scrutiny.

Thus, considering that all modes of liability (apart from – to some extent – superior responsibility which I will discuss below) can be distilled as intentional conduct causally linked to the crime, and in consideration of the mens rea issue in relation to JCE/aiding and abetting, I believe that a shift to a unitary model would be both possible and desirable. The superior who orders the pillaging of a village, the officer who plans it, the corporal who incites her troops to carry it out, the mayor of a nearby town who provide trucks to transport the loot, and the foot soldiers who remove valuables from civilian houses all contribute to the loss of property of the villagers. They are all to blame, and their responsibility can be assessed by examining whether they contributed to this loss with the required mens rea. The different degrees of responsibility of all these participants in the crime can be taken into account in sentencing, and, as Professor Stewart has argued, a summary of each convicted person’s criminal conduct could be included in the judgment’s disposition. On a more practical side, this system would also save considerable time in the drafting of international criminal judgments, where addressing multiple modes of liability requires considerable time and human resources. Since the costs of ICL are a constant concern across the international courts, together with the length of some proceedings, this practical advantage should be given just consideration.

A unitary model would not, in my view, affect an accused’s right to know the nature and cause of the charges against her, and consequently her ability to defend herself. While it is certainly indispensable – in fact, it is a fundamental right – for the accused to be informed of how she is alleged to have participated in a crime, international indictments often use rather general formulas such as “by way of her position, the accused ordered, instigated, planned, or otherwise aided and abetted…”. This practice, which very often includes charging suspects with all the available modes of liability, and then adjusting their aim during trial depending on the evidence, offers little information to the accused with respect to the actual nature of her causal contribution to the alleged crimes. On the other hand, it requires a diligent defence counsel to prepare for and address all the alleged modes. Moving away from modes of liability could require prosecutors to draft indictments which describe more precisely the accused’s alleged criminal conduct. It would also put an end to the practice of charging all modes of liability statutorily available, which in turn would allow the defence and the chambers to better focus their work and possibly shorten the duration of the proceedings.

Regarding superior responsibility, I see merits in the suggested separation of the failure to prevent from the failure to punish, treating the former as a form of participation and the latter as a separate crime. Failure to prevent crimes of subordinates which contributed, at least at the level of facilitating, the commission of the subordinates’ crimes may, if the accused acted with the required mens rea, amount to aiding and abetting by omission. In a unitary system, therefore, failure to prevent could be subsumed under a monistic model of participation which makes anyone who contributed to that crime with knowledge or intent responsible for it. As for failure to punish, the creation of a specific offence based on the commander’s dereliction of duty seems a preferable solution. First, it would avoid the theoretical difficulties encountered by judges and scholars when trying to introduce a causal nexus in this mode of liability. Second, on the front of general prevention, superiors would still be incentivized to properly exercise control and call to task criminal subordinates to avoid criminal sanctions, as well as administrative and disciplinary ones which may be imposed as accessory penalties for this type of offence. Finally, considering failure to punish as a discrete crime would be more respectful of general principles of culpability, which are violated when we convict a superior who negligently failed to punish the murder committed by his subordinate for murder (this violation is even more egregious if the superior did not have actual knowledge of the crime, but only “had reason to know” or, even worse, “should have known”).

Eliminating modes of liability, however, would certainly not be a panacea for blame attribution in ICL. In Italy, where there is a unitary system of participation, there have been long, vibrant, and often unresolved debates on different theories of causation and on the system’s adherence to the principle of culpability enshrined in the Constitution. Thus, the adoption of a unitary system in ICL would also presumably shift the debate and practical problems from modes of liability to the main pillars of blame attribution: causation and culpability. This shift would, I suspect, also be a healthy one for international criminal justice.

3.  Culpability distortions in the Italian system

With respect to culpability, Professor Stewart is perhaps too benevolent about the Italian unitary system’s ability to “avoid modes of liabilities acting as a prism that distorts responsibility, instead of assigning it in line with the culpability announced in the crime”.[8]

In Italy, the rules governing participation of a plurality of persons in the commission of a crime are set forth in articles 110 to 119 of the Penal Code. Article 110 essentially states that when a plurality of persons contribute to/participate in the commission of a crime, each of them is responsible for the sanction prescribed for that crime. The judge may then impose different sentences based on the intensity of their contribution and mens rea. To incur liability through article 110, a person must have at least “facilitated” the commission of the crime. As for the subjective element, article 110 requires either dolus or dolus eventualis (the appropriate definition of the latter is probably one of the most debated issues in Italian criminal law but, it suffices to say, dolus eventualis is generally interpreted as a higher standard than the foreseeability one in JCE III).

Article 116 of the Penal Code, however, states that when more people participate in the commission of a crime (for instance a bank robbery), if a different crime is committed by some of the participants (for instance murder), and this crime was not intended by the others, even those who did not intend the different crime will be punished for it, provided there exists a causal nexus between their conduct and the crime. If the crime not intended by some of the participants is graver than that agreed by them, the judge must take it into account at the sentencing stage. In spite of this mitigating factor, this is essentially a form of strict liability which violates the principle of culpability enshrined in the Italian Constitution.[9] The Constitutional Court partly corrected this state of affairs in 1965, stating that for a participant to incur liability for a crime different from the one she intended, the different crime must have been one that was at least a logically foreseeable development of the intended actions, according to common diligence and considering all the circumstances of the case. Thus, article 116 essentially allows a judge to convict someone for a crime requiring intent even when the evidence only establishes a much less culpable state of mind. This is one of the criticisms put forward by many scholars, including Professor Stewart, to JCE III. Those who defend article 116 might do so for reasons similar to those expressed by Professor Damaška in The Shadow Side of Command Responsibility (2001), where he stated that the persons to whom the graver crime is attributed through such system had already consciously decided to participate in a dangerous criminal activity, thus creating a “less dramatic shift in the register of culpability.”[10]

In sum, while I believe that a shift to a unitary model would be possible and could potentially have practical advantages for ICL, I am also aware that in the delicate field of criminal law there is no perfect system of blame attribution. Individual criminal responsibility is a fascinating but complex subject, and this comment of mine, based on years of practice may seem cursory from a scholarly perspective. I am aware that I have, in a non-scientific fashion, taken a position on issues which are the subject of vibrant academic debates, of long hours of legal research, and of thorough academic articles. Mine is the opinion of an Italian practitioner of international criminal justice, who in the past ten years has had the opportunity to test the advantages and limits of the pluralistic model of participation. I thank Professor Stewart for inviting me to share this practical perspective as part of this symposium.





[1] The views expressed in this comment are exclusively mine, and do not necessarily reflect the views of my present or former employers.

[2] While the jurisprudence of the ad hoc tribunals does not require a causal nexus between a commander failure to prevent and/or punish the crimes and their commission, I am aware that this position is criticized by scholars as non-reflective of customary international law. At the ICC, proof of a causal link is required, in terms of an increased risk, in relation to failure to prevent only, although some scholars and judges have expressed the view that causation is also required with regard to failure to punish.

[3] While ICTY jurisprudence cited by Professor Stewart in The End of Modes of Liability (p. 48) stated explicitly that proof of causal-effect relationship between the conduct of the aider and abettor and the commission of the crime is not necessary, it is possible that this was not meant to signify that no causal link at all is required between the acts of the aider and the crime. First, as noted by Stewart, the conduct of the aider or abettor needs to have a substantial effect on the commission of the crime. As a practitioner, I find it difficult to distinguish this requirement from causation. I am of the view that the Blaškić Chamber actually meant that the contribution of the aider or abettor need not be a conditio sine qua non. In the Applicable Law Section of the judgment in the case of Popović et al., in fact, paragraph 1018 first prescribes that the aider and abettor’s conduct must have had a substantial effect on the crime, and then clarifies that it is not necessary, however, to prove that the crime would not have been committed absent contribution of the aider and abettor.

[4] While significant is a lesser standard of contribution than substantive, I doubt that a review of the jurisprudence of the ad hoc tribunals would reveal a consistent and coherent practice in characterising the facts emerged from the evidence as a significant or substantive contribution.

[5] The End of Modes of Liability, pp. 38-39.

[6] To incur in JCE liability, a person must, inter alia, intend to further the common criminal plan.

[7] This problem is also noted by Matteo Costi in Introduzione al Diritto Penale Internazionale (terza edizione), Milano, Giuffrè, p. 117.

[8] The Strangely Familiar History of the Unitary Theory of Perpetration, p. 22.

[9] The Italian Constitution was issued in 1948, so when the Codice Rocco was issued in 1930 there was no constitutional impediment to article 116.

[10] In his article, Professor Damaška took as an example the felony murder rule in American law. It seems to me that the situation I described in the example features substantive similarity with this type of common law felony.

Norway: Three Codes, Three (Somewhat) Different Solutions

Jørn Jacobsen is a Professor in the Faculty of Law at the University of Bergen.

James G. Stewart’s argument for a unitary theory to replace the modes of liability in international criminal law is of particular interest from the point of view of Norwegian criminal law theory. Here, one of the most prominent contributors to the discipline, Bernhard Getz, made a similar claim in the latter half of the 19th century. Getz’ famous work on a unitary theory from 1875 is often hailed as a masterpiece (published in 1876 as Om den saakaldte Delagtighed i Forbrydelser – en strafferetlig Undersøkelse: Prøveforelæsning over selvvalgt Thema ved Concurrence om en Professorpost i Lovkyndighed). For good reasons too: he wrote it at only 25 years of age, and it certainly demonstrated a theoretical maturity that was a great surprise to Norway ‘s then fledgling criminal law discipline. After all, one counts Schweigaard’s commentaries from the 1840’s as the starting point for this discipline, and Getz’s work was the first significant theoretical contribution to it.

Getz became a professor the year after he published his book, and went on to have a huge impact on the formation of modern Norwegian criminal law. Getz and his close companion, Francis Hagerup, exhibited their impressive ambitions for criminal law in other manners too. They had international ambitions, and were both active in the AIDP (Association International de Droit Penal – International Association of Penal Law). Even more importantly from a Norwegian point of view, they filled key positions within the Norwegian political and legal order. Professor Hagerup served two separate terms as Prime Minister. Getz, for his part, held key positions in forming Norwegian legislation on criminal law and criminal procedure. He led both the commission preparing the Criminal Procedure Code of 1887 and the Criminal Code of 1902. He then became the first Director of Public Prosecutions in order to implement the Criminal Procedure Code of 1887. Certainly, this was the golden age for Norwegian criminal law – likely the first and only time when a criminal law professor held such prestigious positions in Norwegian society and with regard to the Norwegian criminal code, which was hailed as landmark legislation by central Continental criminal law scholars.

What then about Getz’s unitary theory? In itself, it was a critique of the then existing criminal code, the Criminal code of 1842. Norway’s first criminal code after independence was achieved in 1814. The code was itself a result of constitutionalization, as the Constitution of 1814 sect 94 required a criminal code to be enacted. This criminal code was imprinted by the Continental ideals at that time. The models used were the Code Penal (1810) and in particular the Code of Hannover (1840), which was itself inspired by Feuerbach’s Bavarian criminal code of 1813. Not surprisingly, this code differentiated between contributors to crime. In the code, a separate chapter was dedicated to ‘Participation’ (chp. 5), which was understood as something different from the ‘Perpetrator’. Here, the code had separate provisions for several different forms of participation, such as instigation of crime.

Starting from a concept of causation, Getz heavily criticized the 1842 Code. As there were no conceptual differences between the participator and the perpetrator, there was no reason to differentiate between them – thus the title ‘On so-called Participation in Crime’. The unitary theory was also the starting point for Getz when he embarked on the task of drafting the new Penal Code of 1902, which was celebrated throughout Europe. Here, in line with Getz’s program, there was no separate chapter on participation. In regard to sentencing, however, a section of the code in keeping with Stewart’s approach, assigned differences between different contributors’ importance. Getz clearly aimed at putting his theoretical enterprise into practice. However, the code of 1902 still included complicity as additional elements in a number of specific offenses. The additions do not appear to be consistently included in the code. This also left it for court practice to decide in regard to a number of offences whether complicity gave rise to criminal responsibility. In regard to several offences, the Supreme Court concluded that it did. Complicity therefore remained a central concept in Norwegian criminal law theory and practice even after Getz’s own code.

The lack of complete coherence between Getz theoretical project and his solution as a drafter of the code has often been emphasized in later literature. Unfortunately, Getz passed away at young age in 1901 – a year before his criminal code was enacted. As a consequence, we do not know how he would have responded to this criticism. Moreover, it is a point of intrigue for the current debate that later Norwegian scholars have not followed Getz’s approach on these questions. In particular, G. Astrup Hoel (1941) and Erling Johannes Husabø (1999) have criticized it. Husabø’s critique of Getz, in the most recent and extensive investigation into complicity in Norway, starts out from a different concept of causation and also offers conceptual arguments concerning the relation between ‘Perpetrator’ and ‘Participator’.

In 2015, the Norwegian code of 1902 was replaced by the Penal Code of 2005. At least in part, this legislative shift can be described as return to the original code of 1842. In the code of 2005, there is a general section on participation (sect. 15), which makes participation in crime in general subject to criminal responsibility unless otherwise stated in the offense. This new code does thereby not adopt the opposite solution to Getz’s code of 1902. In the preparatory works, there is no discussion of a unitary theory. Instead, treating complicity as a separate subject seems now to be taken for granted.

Lessons learned? There are two ways to see the rise and fall of the unitary theory in Norwegian criminal law. One way to see it is as an unfulfilled promise, one that was hindered by Getz’s early death, legal culture and other obstacles. The other way to see the Norwegian story is that the theory was flawed in the first place. There is not room for a detailed discussion of the subject here, nor on the particularities of international criminal law and the distinct challenges to modes of liability that this field of law faces. It seems clear, however, that the history of Norwegian criminal law at least is not a particularly strong argument for the potential of a unitary theory internationally.

The Unitary Form of Participation in Danish Criminal Law (and Its Potential Use in International Criminal Law)

Dr Iryna Marchuk is an Associate Professor in the Faculty of Law at the University of Copenhagen.  She has authored the monograph Fundamental Concept of Crime in International Criminal Law: A Comparative Law Analysis (Springer 2014, reviewed in the Journal of Int Crim Justice 15(2), 2017).

In his article, which is a basis for this mini-symposium on criminal participation, James G. Stewart advances the unitary theory of participation that has been utilized in a number of national jurisdictions, including Denmark.[1] He opines that the Danish approach to criminal participation – being representative of the unitary form of participation – can be successfully transposed to international criminal law that has been struggling to find adequate forms of attribution that characterize responsibility for international crimes.[2] Much criticism has been directed towards the interpretation of forms of liability in the ICC that has fallen for Claus Roxin’s theory of participation that distinguishes between principals and accessories to a crime based upon the ‘control over the crime’ theory.[3] Needless to say, the introduction of Roxin’s theory into international criminal law has been subject to passionate debates among academics and practitioners alike who have pondered over its suitability for the needs of international criminal law.[4] The critics of the theory, among them James G. Stewart, have come forward with alternative suggestions as to how criminal responsibility can be assigned to those responsible for international crimes. In this mini-symposium, I was asked to comment on the applicability of the unitary form of participation from the Danish perspective. More specifically, I will focus on the historical evolution of the concept of unitary form of participation in Denmark, its understanding in the theory of criminal law, its applicability in practice, and reflect on the suitability of the unitary form of participation in international criminal law.

The unitary form of participation that does not distinguish between principals and accessories to a crime was introduced into Danish criminal law by Carl Torp who shared the views of the Norwegian scholar Bernard Getz as to the interpretation of participation in a crime. Getz’s theory, introduced in ‘On The So-Called Complicity in Crime’, is rooted in his understanding of causality.[5] In his opinion, every act that is a condition for the violation of law should be equally punishable by criminal law. This, in his view, makes a formal distinction between principals and accomplices to a crime completely unnecessary.[6] The mantle was picked up by Torp, who was inspired by Getz’s theory and advocated for abolishing the distinction between parties to a crime in Danish criminal law, as he considered the perpetrator’s and the accessory’s conduct as being equally unlawful.[7]

More specifically, Torp argued that in reality it is impossible to draw a clear distinction between a principal (gerningsmænd) and an accomplice (meddelagtige) to a crime either on the objective or on the subjective grounds.[8] He criticized the objective theory that distinguished between a principal and accomplice to a crime based on causality.[9] According to the theory, a principal is a person whose action is causal to a crime, whereas an accomplice is a person whose action is merely a condition leading to the commission of a crime. Torp opines that in the chain of causal acts, all of them are equally necessary for the crime to take place and therefore the distinction advanced by the proponents of the objective theory is irrelevant.[10]

Following this, Torp takes on the subjective theory of participation that distinguishes between a principal and an accomplice to a crime based on the person’s subjective state of mind. If a person entertains the intent for committing the crime and embraces the crime as his/her own, he/she is a principal. If the person takes part in realizing other person’s intent and advances the other person’s interests, he/she is an accomplice. Torp opines that the subjective theory does not provide any clear and practically useful distinctions between parties to a crime.[11] In fact, he argues that judges would find it practically unsolvable to draw a clear demarcating line between parties to a crime since the distinction based upon the subjective state of mind simply does not exist.[12]

Torp’s understanding of participation in a crime has been influential in Denmark. This is largely due to the fact that the unitary form of participation was incorporated into the 1930 Danish Criminal Code that introduced § 23, which does not formally distinguish between parties to a crime.[13] This marked the departure from the 1866 Criminal Code that clearly distinguished between principals and accomplices to a crime. In the explanatory notes to the to the 1930 Danish Criminal Code Draft, Torp posed an important question as to “whether all parties to a crime should be punished equally, and whether some should be punished milder than others or should be altogether relieved from criminal responsibility”.[14] He considered the equal treatment of a person who committed a crime and an accomplice fully justified since both of them had triggered conditions for the crime to take place.[15] He opined that the theoretical construction of the participation in a crime belongs to an academic discussion, whereas the law should have a rather neutral expression.[16]

As a result of Torp’s involvement in the legislative preparatory work, the 1930 Danish Criminal Code provides a rather succinct provision on criminal participation that “applies to anybody who by means of instigating, counseling or aiding has contributed to a crime”.[17] At first glance, it seems that the provision distinguishes between parties to a crime; however, the legislative intent was in fact the opposite – to erode any distinction between principals and accomplices. In other words, any contribution to a crime, regardless of whether it is physically committing the crime or providing assistance, is equally blameworthy. The introduction of the unitary form of participation has not been uncontroversial in Danish law. Ross criticized the doctrine for failing to distinguish between parties to a crime, specifically in relation to so-called ‘result’ offences (forårsagelsesdelikter).[18] He submits that there is clearly a difference between a person who fires a gun at close range and a person who supplied the gun or provided the address of a victim. He continues that it is impossible to imagine that an accomplice could have entertained the direct or probability intent as required for the crime of murder.[19] In Ross’ opinion, these considerations render the Getz-Torp theory meaningless.[20]

Although there is no formal distinction between parties to a crime in the Danish Criminal Code, the law gives broad discretionary powers to the judiciary at the sentencing stage that can impose a milder sentence based on the participant’s degree of contribution to a crime or his/her subjective state of mind. According to § 23, a sentence may be reduced in a number of situations: (1) a person who intended to provide a less substantial contribution to a crime; (2) a person who intended to strengthen an already formed criminal intent; (3) a crime was not completed; and (4) intended participation failed.[21] The judges have at their discretion to choose not to impose a sentence for participation in a crime that carries a sentence of imprisonment up to maximum of 4 months and where such participation was due to negligence.[22]

The scope of § 23 of the Danish Criminal Code is rather wide and far-reaching. As noted by Vestergaard, the Danish Criminal Code – representative of an “extreme variation of Einheitstäterbegriff” – has no counterparts in Europe when it comes to the “brevity and wideness of criminal law provisions on participation in a crime”.[23] One of the key features of participation in a crime under Danish criminal law is that there is no requirement that the principal offence be committed in order to impose criminal liability upon a party to a crime. Also, there is no requirement in Danish law that the person’s action has to be causal to the commission of the crime.[24] It is clear from the wording of § 23 that participation in a crime may take place in the form of (1) instigating, (2) counseling or (3) aiding. The provision should be read as covering all parties to a crime, including a person who commits the crime and those who in any other way contribute to the crime. Instigating is understood as covering any verbal or physical conduct that prompt another to commit a crime or encouraging another person to provide assistance, which is necessary for the commission of the crime.[25] Counseling consists of providing guidance and giving hints. As an example, the supply of information about the building’s layout, the address of a victim, or tax related information fall within counseling as understood by Danish criminal law.[26] Aiding means undertaking any preparatory or executive steps towards the commission of a crime (e.g. assistance with the transport, physical use of force).[27] Passive participation is also punishable under Danish criminal law, in particular in situations when a person has an obligation to act and could have prevented the crime from taking place. Participation in a crime may take place at any time at the preparation or execution stages of the principal offence and in some circumstances after the crime has been committed.[28] As for the mens rea required for participation in a crime, both intent and negligence will be sufficient and the required mens rea standard varies according to the mens rea, which is required for a specific crime. However, each participant is judged based on his subjective state of mind, which means that there are situations when participants in a crime are charged with different offences based on mens rea they have entertained (e.g. one participant is charged with theft, whereas another one is charged with burglary).[29]

Despite the absence of a formal distinction between parties to a crime, normal meaning is being accorded to the words ‘perpetrator’ (gerningsmand) and ‘accomplice’ (medvirkende) in academic literature and judgments of Danish courts. However, there is no consistency in the way the terminology is used since Danish courts are not bound to employ uniform terminology to distinguish between parties to a crime.[30] As an example, if the act of participation is not covered by the offence description in the special part of the Criminal Code, the Danish courts in their judgments are not obliged to refer to § 23 as a ground for imposing criminal responsibility.

This contribution to the symposium is by no means an all-exhaustive account of participation in a crime in Danish criminal law. Rather it is an attempt to outline the legal contours of the unitary form of participation as it has historically developed in Danish criminal law and has been applied in practice. The ultimate question of this symposium is whether a unitary form of perpetration – similar to the Danish example – could prove to be a useful in assigning criminal responsibility to the culprits of international crimes and be transposed to international criminal law.

First of all, I believe that the idea of the symposium itself that draws on expert opinions of scholars from different national jurisdictions is brilliant. This open conversation and exchange of opinions allow us to infer general principles underlying the attribution of individual criminal responsibility across many jurisdictions that could be distilled and applied in the context of international criminal law. Clearly, the ICC’s transposition of Roxin’s theory is problematic, as it seems unfair that the German doctrine was recognized to be well suited to cater to the needs of international criminal law without any attempt on the part of the ICC judges to conduct a thorough comparative legal analysis of the notion of criminal participation in other national jurisdictions. Hence, this invitation to contribute to the understanding and applicability of the theory of criminal participation in different national jurisdictions is particularly valuable, and hopefully will have a catalyzing effect on the interpretation of the concept of participation of crime in international criminal law.

Is there anything useful that we can learn from the Danish example? Could the unitary form of participation work in international criminal law? The fact that the unitary form of participation has been in place in Denmark since 1930 and still remains unchanged is a strong indicator that it is a workable tool. However, should the unitary form of participation find its way in international criminal law, a number of challenges associated with the implementation of the doctrine may arise:

1) Far-reaching scope of the doctrine that would allow penalizing any kind of conduct, which is part of the offence. The doctrine is extremely wide in its scope as technically any contribution to a crime, regardless of how insignificant it is, should be criminally liable. On the one hand, it seems that there would be no harm in embracing rather wide provisions on criminal participation in international criminal law that criminalize any conduct that leads to the commission of core international crimes. On the other hand, it seems very unlikely that the cases of attempted perpetration of international crimes or attempted participation more broadly would actually make it to the international courts and tribunals. The ICC is already experiencing a heavy workload investigating the crimes within the jurisdiction of the Court and preliminarily examining alleged crimes in various parts of the world. It is unlikely that the ICC would be interested in expanding its workload by looking at conduct that does not live up to the required gravity threshold in light of the insignificant contribution to a crime as it is interested in going after ‘big fish’.

2) Dangers of going back to the ‘just convict everybody’ approach in the ad hoc tribunals. One can easily see that the implementation of the unitary form of participation would benefit the work of the prosecutorial divisions of international courts and tribunals. In this case, while assigning criminal liability to the person, the Prosecutor would not have to be preoccupied with evaluating the level of contribution and marrying it with a specific mode of liability. However, the adoption of of the unitary form of participation with its broad and far-reaching scope may lead us to going back to the infamous practices of the ad hoc tribunals that have been criticized for endorsing an extremely broad form of the JCE doctrine that was equated with the ‘just convict everybody’ approach. In practice, of course, the Prosecutor of any international court or tribunal has the ability to filter cases that would be submitted for further consideration by the judges and it is unlikely that they would initiate cases where the suspect’s contribution was rather insignificant or there is a missing causal link between the person’s action and the outcome. That said, one should be aware of possible implications that could accompany the transposition of the unitary form of participation into the terrain of international criminal law.

3) Uncertainty as to the sentencing practice. The dichotomy between principals and accomplices to a crime in the German theory is clearly reflected in domestic sentencing practices. In other words, a principal whose contribution to a crime was essential deserves a higher punishment than an accomplice whose contribution to a crime was less significant. The unitary form of participation as known in Danish law recognizes all parties to a crime as equally criminally liable. Technically, all parties to a crime deserve the same punishment. However, in practice judges may chose to reduce punishment for some parties to a crime in light of specific circumstances. The transposition of the unitary form of participation to international criminal law may create difficulties at the sentencing stage. Should the judges of international courts impose the same punishment upon a rebel commander who supplied one box of Kalashnikov guns as the Head of State who masterminded and directed the implementation of the genocide campaign? If international courts opt for the unitary form of participation, the judges may choose to exercise their judicial discretion and impose punishment depending upon the role and contribution of each party to a crime.

4) No formal distinction between principals and accomplices to a crime but it is still implied. The key feature of the unitary form of participation is the absence of any formal distinction between principals and accomplices to a crime. But is such a distinction truly ignored in practice? As shown by the Danish example, although the law is formulated in neutral terms equally imposing criminal responsibility upon all parties to a crime, in practice judges do in fact distinguish among parties to a crime by imposing different sentences based upon the person’s actual involvement in the crime. In the judgments of Danish courts, words such as ‘perpetrator’ and ‘accomplice’ are attributed their normal meaning and often feature in the text of the judgments. If the form of unitary participation made its way to the ICC, the problems that the Prosecutor and judges are saddled with when matching the suspect’s conduct to the specific mode of liability would disappear. There would be no need to determine whether the suspect was a direct perpetrator or an indirect co-perpetrator based on a rather complex German ‘control over the crime’ theory. However, this does not mean that the judges wouldn’t have to be preoccupied with evaluating the level of contribution to a crime, especially at the sentencing stage.

5) Is the wording of article 25 of the Rome Statute compatible with the unitary form of participation? When the ICC judges introduced Roxin’s ‘control over the crime’ theory to distinguish between principals and accomplices to a crime, some scholars and even the ICC judges themselves were skeptical of the adopted approach. Article 25 of the Rome Statute does not mention the control over the crime approach and only lists available modes of liability. The question that has troubled academics and practitioners is whether modes of liability as listed in article 25 of the Statute are arranged in a particular hierarchical order where clear predominance is given to principal modes of liability. If the unitary form of participation is adopted in the ICC, then it does seem that the wording of article 25 is not ideal. It seems redundant to distinguish between various modes of liability to a crime if the unitary form of participation requires no such distinction. In that case, Article 25 would benefit from a more simplified formulation akin to the provision on criminal liability in the Danish Criminal Code.

6) Is universal application of the unitary form of participation truly possible? In his article, James G. Stewart posits that the unitary form of participation could serve as an effective tool to establish participation in the international crime regardless of the forum where it is charged, be it a national or an international one. Although the theory has the potential of becoming a rival to Roxin’s ‘control over the crime’ theory in the ICC, its universal applicability in the countries that ratified the Rome Statute is rather questionable. If core international crimes were to be prosecuted at the national level, the judges would still employ tools of attribution available in the national jurisdiction and do not really need to invoke the unitary form of participation to arrive at a reasonable outcome.

7) Interpretation problems. One should not discard that Denmark is a small jurisdiction where judges are trained in the same legal tradition. Whereas Danish judges are used to dealing with the unitary form of participation in practice, international judges who are trained in different national jurisdictions will undoubtedly encounter problems in applying the theory in practice. In particular, the judges who come from national jurisdictions where there is a clear divide between principals and accomplices to a crime will have a harder time grasping the applicability of the theory in practice and disagreements are likely to take place.


The implementation of the unitary form participation in international criminal law may prove to be a workable tool. It would undoubtedly make the job easier for the Prosecutor and the judges who would not have to be concerned about making correct determination at the initial stages of trial as to the role played by the suspect/accused in the commission of international crimes and choosing the correct mode of liability. However, it does not mean that the unitary form of participation could be a panacea from all the criminal participation related problems encountered by the ICC. As discussed above, a number of challenges may arise and disagreements may take place, especially at the sentencing stage. The major danger of the transposition of the concept is that it does not repeat the fate of the JCE doctrine (representative of the subjective theory) given its broad and far-reaching scope. My analysis of the unitary form of participation is based on my knowledge of the Danish model, which is an extreme form of Einheitstäterbegriff. I do not discard that there may be some other variations of the unitary form of participation in other national jurisdictions that may better cater to the needs of international criminal law.

[1] Stewart, James G., The Strangely Familiar History of the Unitary Theory of Perpetration (January, 2016). Forthco

ming in Bruce Ackerman et. al. (eds.), Visions of Justice, Essays in Honor of Professor Mirjan Damaška (Duncker & Humblot, Berlin, 2016). See also: Stewart, James G. “The End of ‘Modes of Liability’ for International Crimes.” Leiden Journal of International Law 25.01 (2012): 165-219.

[2] Ibid., pp. 13-18.

[3] Pre-Trial Chamber I, Prosecutor v. Thomas Lubanga Dyilo, “Decision on the confirmation of charges”, 29 January 2007, ICC-01/04-01/06-803-tEN, paras 326-341; Trial Chamber, Prosecutor v. Thomas Lubanga Dyilo, “Judgment pursuant to Article 74 of the Statute”, 14 March 2012, ICC-01/04-01/06-2843, para. 994; Prosecutor v. Germain Katanga, “Judgment pursuant to Article 74 of the Statute”, 07 March 2014, ICC-01/04-01/07-3436, para. 3194. See: C. Roxin, Täterschaft und Tatherrschaft (Verlag de Gruyter, Hamburg, 8th ed., 2006), pp. 277-282.

[4] Jessberger, Florian, and Julia Geneuss. “On the Application of a Theory of Indirect Perpetration in Al Bashir German Doctrine at The Hague?.” Journal of International Criminal Justice 6.5 (2008): 853-869 (arguing in favor of the concept of indirect perpetration as a key mode of liability in ICL); Ohlin, Jens David. “Joint intentions to commit international crimes.” Chicago Journal of International Law 11.2 (2011): 693-753 (suggests a theory of joint intentions as an alternative to the JCE doctrine in the ad hoc tribunals and the control over the crime theory in the ICC); Manacorda, Stefano, and Chantal Meloni. “Indirect Perpetration versus Joint Criminal Enterprise Concurring Approaches in the Practice of International Criminal Law?.” Journal of International Criminal Justice 9.1 (2011): 159-178 (views positively the introduction of the control over the crime theory in the ICC, while at the same time highlighting some possible challenges with respect to the implementation of the doctrine in practice); Werle, Gerhard, and Boris Burghardt. “Establishing Degrees of Responsibility: Modes of Participation in Article 25 of the ICC Statute.”Pluralism in International Criminal Law, Oxford University Press (2014), Forthcoming (2013) (advocates for a systematic understanding of a four level hierarchy of individual criminal responsibility in Article 25 of the Rome State as a starting point of discussion); Ohlin, Jens David, Elies Van Sliedregt, and Thomas Weigend. “Assessing the control-theory.” Leiden Journal of International Law 26.03 (2013): 725-746 (calling on the ICC Appeals Chamber to reconsider or improve the control over the crime theory in the ICC); Granik, Maria. “Indirect Perpetration Theory: A Defence.” Leiden Journal of International Law 28.04 (2015): 977-992 (critically reflects on the suggested Ohlin’s theory of joint intentions and – as a title suggests – defends the applicability of the control over the crime theory in the ICC). See also: Trial Chamber, Prosecutor v. Thomas Lubanga Dyilo, “Judgment pursuant to Article 74 of the Statute”, 14 March 2012, ICC-01/04-01/06-2843, Separate Opinion of Judge Adrian Fulford; Trial Chamber II, Prosecutor v. Germain Katanga, “Judgment pursuant to Article 74 of the Statute”, 07 March 2014, ICC-01/04-01/07-3436, Minority Opinion of Judge Van den Wyngaert, paras 280-281.

[5] Getz, Bernard, Om den såkaldte delagtighed i forbrydelser (Christiania, 1875), pp. 1-3 (indledende bemærkninger).

[6] Ibid.

[7] Torp, Carl, Den Danske Strafferets almindelige del (G.E.C.Gads Forlag, København, 1905), p. 559.

[8] Ibid., p. 559.

[9] Ibid., p. 560.

[10] Ibid., p. 560.

[11] Ibid., p. 571.

[12] Ibid.

[13] Danish Criminal Code (Straffeloven), available at

[14] Betænkning angående de af den under 11. august 1905 nedsatte Straffelovskommission udarbejdede Forslag indeholdende Udkast til Love vedrørende den borgerlige Straffelovgivning med Motiver efter Justitsministeriets Anmodning udarbejdet af Carl Torp (København, Trykt hos J. H. Schultz A/S, 1917).

[15] Ibid. See Torp’s full citation in Danish: ”Tanken er formodentlig, at den medvirkende ubetinget maa stilles lige med den, der har foretaget en Iværksættelseshandling. Denne Ligestilling er sikkert berettiget, hvor Spørgsmaalet er om Betingelserne for at opnaa Straffrihed ved Tilbagetræden, fordi den medvirkende ligesom den, der har foretaget en Iværksættelseshandling, har sat Kræfter i Bevægelse, som han ikke mere — i alt Fald ikke fuldt ud — er Herre over.

[16] Ibid. See Torp’s full citation in Danish: ”I Virkeligheden er det Spørgsmaal, som
 Straffeloven bør tage Stilling til, kun det, 
om alle Medvirkende bør straffes lige, eller om nogle, og da hvilke, bør straffes mildere end andre eller være straffri. Den teoretiske Konstruktion bør overlades Videnskaben; og Loven bør derfor saa vidt muligt vælge ganske neutrale Udtryk.” In later academic works, there is also an overwhelming support for the simplification of the rules on the participation in a crime and preference for a practically oriented law in that regard. See: Hurwitz, Stephan, Den Danske Kriminalret almindelig del, 3. Oplag (G.E.C.Gads Forlag, København, 1964), p. 499.

[17] See: § 23 (1). In Danish: ”Den for en lovovertrædelse givne straffebestemmelse omfatter alle, der ved tilskyndelse, råd eller dåd har medvirket til gerningen.”

[18] Ross, Alf, Forbrydelse og Straf: Analytiske og reformatoriske bidrag til kriminalrettens almindelige del (Nyt Nordisk Forlag Arnold Busck, København, 1974), p. 120.

[19] Ibid., p. 121.

[20] Ibid.

[21] See: § 23 (2). In Danish: Straffen kan nedsættes for den, der kun har villet yde en mindre væsentlig bistand eller styrke et allerede fattet forsæt, samt når forbrydelsen ikke er fuldbyrdet eller en tilsigtet medvirken er mislykkedes.

[22] See: § 23 (3). In Danish: For så vidt ikke andet er bestemt, kan straf for medvirken ved lovovertrædelser, der ikke straffes med højere straf end fængsel i 4 måneder, bortfalde, når den medvirkende kun har villet yde en mindre væsentlig bistand eller styrke et allerede fattet forsæt, samt når hans medvirken skyldes uagtsomhed.”

[23] Vestergaard, Jørn, Criminal Participation in Danish Law – Uniformity Unlimited? in Raimo Lahti and Kimmo Nuotio, Criminal Law Theory in Transition: Finnish and Comparative Perspectives (Finnish Lawyers’ Publishing Company, Helsinki, 1992), p. 2.

[24] Ibid. Langsted, Lars Bo, Waaben Strafferettens almindelige del, 6. reviderede udgave (Karnov Group, København, 2015), pp. 245-246.

[25] Langsted supra note 23, p. 242.

[26] Ibid.

[27] Ibid.

[28] Ibid., pp. 246-250. See also: Vestergaard supra note 22, p. 9.

[29] Langsted supra note 23, pp. 250-251. See also: Hurwitz supra note 15, pp. 506-508; Vestergaard supra note 22, p. 10;

[30] Vestergaard supra note 22, p. 4.

The Theories of Blame Attribution In Brazilian Criminal Law

Carlos Eduardo A. Japiassú is Professor of Criminal Law at the Federal University of Rio de Janeiro and Director of the PhD Program at Estácio de Sá University. He is also the Secretary General of the International Association of Penal Law (IAPL).

  1. Introduction

The various possibilities for allocating blame for criminal offending are always important issues in Brazil, especially in recent years, with the emergency of punishing large corruption scandals by the Federal Justice System. The issues are numerous, including whether all perpetrators and accomplices should be punished equally or if they deserve different degrees of punishment according to their culpability. Aiming to answer this question, three main theories have been developed within Brazil, and Brazilian legislators have explicitly adopted some of them. In response to Professor Stewart’s recent article drawing on the Brazilian experience with the unitary theory of perpetration, I provide here an overview of these three theories to reflect more on the potential example Brazil may offer international criminal justice.

  1. The Pluralist Theory

For the so-called Pluralist Theory, each individual who commits a particular wrong shall be punished particularly (delictum sui generis). This maxim means that there must be as many crimes as participants in the criminal activity, as distinct from a single crime that various actors participated in. The Pluralist Theory has not been adopted by the Brazilian legislation as a whole, but, exceptionally, the Penal Code includes certain examples of its use. These include: article 124 (abortion committed by a pregnant woman or with her consent); article 126 (abortion committed by a third person with the pregnant woman’s consent); articles 217-A (rape of a person under 14 years old); article 218 (inducing someone under 14 years old to satisfy someone else’s sexual interest); art. 235 (bigamy) (marriage by an unmarried person to someone already married, knowing this fact); articles 317 (soliciting or receiving bribery) and 333 (offering or giving bribery). 

  1. Dualist or differentiated theory

According to this theory, there is only one single criminal act and forms of participation differ on two distinct levels. Thus, the main activity is carried out by the perpetrator, whereas accessory participation exists on a lower plane, such that it is punishable with a less severe penalty. A system of this sort creates two conceptual positions: perpetrator and accomplice. According to the Dualist or differentiated theory, each of the various modes of liability assigning wrongdoing to a particular actor fall within one or the other of these two families. Convincing an unwitting child to carry out a crime on one’s behalf is a form of perpetrating the crime; providing someone with mere assistance in carrying out the crime is a species of accessorial liability.

In Brazil, the 1830 Criminal Code (articles 4, 5 and 6), as well as the 1890 Penal Code (articles 17 e 18) adopted this dualist theory, distinguishing perpetrators and accomplices. The structure and content of these articles revealed an assumption that multiple actors joined together to bring about a single criminal offence, which they committed by making a main or a secondary contribution. Although Brazilian courts largely applied it prior to 1940 (see below), many lawyers and scholars of the period sharply criticized it. For instance, Justice Nélson Hungria,[1] member of the Brazilian Federal Supreme Tribunal from 1951 to 1961, argued that the differentiated approach was too complicated and exotic. These criticisms led to the adoption of the unitary theory of perpetration in the 1940 Penal Code, which Professor Stewart’s article references.

  1. The Unitary Theory

The Unitary Theory of Perpetration rejects the distinction between perpetrator and accomplice, positing that all participants are responsible for the crime they contributed to.[2] Although this unitary theory has faced many criticisms in Brazil, it came to represent a pragmatic solution to difficulties lawyers, judges and scholars experienced with differentiated modes of liability in this country. Thus, when the Brazilian legislature formally announced a unitary theory of perpetration in article 25 of the 1940 Penal Code, it intended to bypass much of the difficulty the differentiated system was perceived to have produced by relating perpetratorship and causation (article 11).

Nevertheless, it bears stressing that the 1940 Penal Code has always maintained a certain relationship with the differentiated theory, because, when addressing sentencing in article 45 of the Code, the penalty imposed on the organizer of a criminal cooperation is elevated vis-à-vis other perpetrators and/or accomplices. At the same time, article 48, II, of the Code established that the penalty can be reduced if the individual’s conduct was less important or if he or she decided to take part in a crime different from the one which was committed. Moreover, the General Part of the 1984 Penal Code has expressly watered down the Unitary Theory adopted by the 1940 Penal Code by allowing the punishment of offenders in accordance with their culpabilities. The resulting position in current Brazilian criminal law is therefore considered to be in between unitary and differentiated theories.

Therefore, in contemporary Brazilian Penal Law, blame attribution follows a mixed unitary theory. On the one hand, the chapeau of article 29 of the Penal Code reflects the unitary theory by insisting that “whoever, in any way, concurs for the crime is under the penalties attributed to it.” On the other, the Code elsewhere recognizes the differentiated theory to some extent by instructing courts to have regard for the gravity of the penalty. As a consequence, although the concept of perpetrator and accomplice is not articulated anywhere in the Brazilian Penal Code in keeping with the unitary theory of perpetration, it is possible to assign different penalties to each of the participants, in line with the differentiated theory.

  1. Conclusion

 The adoption of the mixed unitary theory in Brazilian Penal Law is well accepted in the country. Today, there are few debates about its propriety or conceptual coherence within the Brazlian legal community, perhaps because the unitary theory provides a pragmatic answer to problems raised in criminal cases. Besides, Brazilian criminal law scholars have not tended to be especially open to innovation. This legal conservatism is no doubt a feature of Brazil’s relative isolation: it is the only Portuguese speaking country in the Americas and most of the population lives close to the Atlantic Coast. These factors help produce a culture that, despite our initial import of the unitary theory all those decades ago, now has difficulties accepting foreign innovations in general and, among penalists, arguments against new trends and changes are common.

Thus, a pragmatic approach to blame attribution plus a certain legal resistance to change may explain the longevity of the Unitary Theory in Brazil to this day. Of course, it remains unclear whether a pure rendition of this theory or a mixed variant like that now applicable in Brazilian Penal Law is a good alternative for International Criminal Law or International Criminal Courts. That said, considering the different approaches of the Ad Hoc International Criminal Tribunals and of the International Criminal Court, it may be useful to consider the Unitary Theory as a simpler tool to answer this question.

[1] Nilo Batista, Concurso de agentes. Rio de Janeiro: Lumen Juris, 2005.

[2] Roberto Lyra, Comentários ao Código Penal, Vol. II. Rio de Janeiro: Forense, 1958, p. 317.

A Unitary Theory of Perpetration? New Zealand and a Touch of the International

Sir David Baragwanath is a Judge at the Special Tribunal for Lebanon, and was President of that court between 2011 and 2015. Before this, he sat as a judge of the High Court and the Court of Appeal of New Zealand. He was made a knight companion of the New Zealand order of merit. It is a great honor to host such a distinguished compatriot here.


This paper considers concepts of unitary theory of perpetration in relation to New Zealand domestic law and touches on aspects of international criminal law.  It argues:

  • that New Zealand law adopts in s 66(1) of its Crimes Act 1961 a dilute unitary approach (from which a divided Supreme Court in Ahsin v The Queen [2014] NZSC 153 has recently created a partial exception) and in s 66(2) a mixed unitary and non-unitary approach. It may be expected that New Zealand will prefer the recent UK Supreme Court and Privy Council decisions in R v Jogee [2016] UKSC 8 [2016] UKPC 7, which accord with the language of s 66(2), to the decision of the High Court of Australia in Miller v The Queen [2016] HCA 30 which preferred the approach of R v Chan Wing-Siu [1985] AC 168(PC) which, having been as accepted for 30 years as stating English law, was rejected in Jogee;
  • that the decision of the Special Tribunal for Lebanon in Interlocutory Decision on the Applicable Law STL-11-01/I 16 February 2011 (STL Casebook 2011 pp167-8, 177 and 182) paras 227, 248-9 and 259 n 381) in relation to terrorism (at para 249) conforms with Jogee by rejecting JCE III for crimes of specific intent;  other cases cited (para 239) adopt the more rigorous approach of a form of  JCE III which conforms with Chan Wing-Siu but not with the Statute of the STL.

The New Zealand statute

New Zealand criminal law is codified by the Crimes Act 1961 which states:

Part 4 Parties to the commission of offences

66 Parties to offences

(1) Every one is a party to and guilty of an offence who—

(a)  actually commits the offence; or

(b)  does or omits an act for the purpose of aiding any person to commit the offence; or

(c)  abets any person in the commission of the offence; or

(d)  incites, counsels, or procures any person to commit the offence.

(2) Where 2 or more persons form a common intention to prosecute any unlawful purpose, and to assist each other therein, each of them is a party to every offence committed by any one of them in the prosecution of the common purpose if the commission of that offence was known to be a probable consequence of the prosecution of the common purpose.


In Alan Reed and Michael Bohlander’s Participation in Crime: Domestic and Comparative Perspectives,[2] Julie Tolmie of the University of Auckland gives a valuable overview of the New  Zealand perspective.

S66(1), as far as it goes, applies what James Stewart terms “a watered down unitary concept”: “criminal legislation [that] … articulates the different forms of causal connections that might apply within a unitary framework” and, unlike any example of  his pure unitary theory”, “does … forewarn would-be  criminals of their exposure to potentially serious criminal responsibility” and so not “violating  the principle of legality.”[3]  Both the “person who actually commits the offence” and all who “aid…, abet, incite, counsel or procure” the offence are party to and guilty of it.  All are factual contributors, in different ways, to the very same crime.  Each personally commits an act which in fact is causally connected to achieving the crime.

S66(2) provides a further pathway to liability for an offence by the one, who in the language of s66(1)(a), actually commits the offence,  on the part of all others who fall within the scope of s 66(2).

S66(2) constitutes in part a unitary offence – a single offence, to which all liable for it are regarded as parties. But in part it is not unitary: liability of Parties B1-B? for the offence committed by Party A does not derive from their factual contributions, in different ways, to the very same crime.  It is not necessary that, as in the case of s 66(1), each personally commits an act which in fact is causally connected to achieving the crime.   Instead S 66(2) imputes to Parties B –B? a legal vicarious liability for the offence Z of A, on account of (1) their being party with A to a common purpose (say bank robbery) which they share with him ( to commit that offence ), coupled (2) with their knowledge that his committing a different offence (say murder) is a probable consequence.

It should be added that in Ahsin the New Zealand Supreme Court, reversing Court of Appeal authority, held that s 66(2) imposes liability both for the robbery/ murder case just mentioned and also for a single offence the object of some common purpose. I respectfully disagree with that conclusion, based on a purely verbal reference to the words “common purpose” that can literally embrace the case of a single crime.    But the context does not permit that conclusion.    The “common purpose” is to commit a crime.  Once that element is established and the crime is committed the case under S66(1) is complete.  There is simply no reason to have s66(2) apply to that crime.  Its function is to impose liability for a second and distinct crime where the elaborate further mens rea condition is proved.  Ahsin unnecessarily complicates the clear and simple language of the statute.

There is a striking contrast with such non-unitary jurisdictions as Germany and Lebanon, whose Criminal Code distinguishes sharply between “the perpetrator of an offence” (Criminal Code Art 212) and others, such as “an accomplice”(Art 213) and “an instigator” (art 217).

Cassese writing extrajudicially pointed out that in international criminal law

…although it is not strictly required by any of the statutes of the modern international criminal courts, the judicial practice has been to classify every charge and conviction by mode of liability. Thus, the discussion of the modes of liability – whether it is, for example, commission, aiding and abetting, ordering, or command responsibility – is a focus of almost every judgment. [4]

Yet unless the law so requires, this complication of  trial and adjudication, by the introduction to international criminal law of concepts of “modes of liability” not stipulated by legislation or required by any general principle of law, is unnecessary.  Its use may illustrate the old habit of lawyers to create imaginary goblins.  Familiar examples are the notion that there is some such thing – Kant’s ding an sich­, a real entity – as “possession”, or “a company”, beyond the legal result created by rules of law.  Each has been demolished by masters of jurisprudence; [5] it is time for the same to be done to modes of liability in jurisdictions that do not stipulate for them.  Otherwise international criminal procedure, already complex and expensive, adds to those problems by forcing itself to leap over self-created non-existent hurdles.

The following draws on Edmonds v The Queen [2011] NZSC 159 at [21ff] where William Young J described the history of the New Zealand legislation and provided a comparison with the laws of England and Wales.

The New Zealand Act derives its s66(1)(b)-(d) from s71 of the 1879 draft Code prepared by the Stephen Commission and s8 of the (current) Accessories and Abettors Act 1861 (UK), but there has been divergence between the English and New Zealand jurisdictions in relation to mens rea.  The New Zealand aiding and abetting cases require the Crown to establish that the secondary party intentionally helped or encouraged the principal offender with knowledge of the essential matters constituting the offence, including the principal’s mens rea.   Some English cases on aiding and abetting suggest that recklessness or foresight, rather than knowledge, is sufficient. [6]

As to s66(2), William Young J observed:

[23] The general common law principles of joint enterprise liability correspond closely to s 66(2).

Chan Wing-Siu, Jugee and Miller

A footnote then cites Chan Wing-Siu v R [1985] AC 168 (PC) and suggests that its author, Sir Robin Cooke, then President of the Court of Appeal of New Zealand and later a Law Lord, would have had s66(2) well in mind.  In fact its language, requiring that “the commission of that offence was known to be a probable consequence of the prosecution of the common purpose”  is more exacting than that required by Chan Wing-Siu.  There:

it was submitted for the appellants that it was not enough if an appellant foresaw death of grievous bodily harm as a possible consequence of the joint enterprise: that the jury ought to have been directed that it must be proved that he foresaw that one of those consequences would probably  result.

The former was held to suffice.

The citation from Edmonds continues:

It is, however, important to recognise that the common law status of joint enterprise liability principles means they are more susceptible to judicial development than s 66(2).

Given the leap in Chan Wing-Siu from the “probable consequence”of s 66(2) to the lesser “possible consequence” test adopted, the observation was correct.

After Chan Wing-Siu v R had been followed for three decades by English law, including R v Powell [1999] AC 1, [1997] UKHL 45, in R v Jogee [2016] UKSC 8 [2016] UKPC 7 a joint judgment of the UK Supreme Court and the Privy Council rejected Chan Wing-Siu v R.  It held that Party B is not guilty unless he acted with the same mental intent the law requires of Party A (Party B’s intent may be conditional –  agreement that if the occasion arose A’s crime would be committed: at para 94).

The High Court of Australia in Miller v The Queen [2016] HCA 30 has since declined to depart from the Chan Wing-Siu v R principle which it had previously endorsed. Recently, in HKSAR v Chan Kam Shing FACC No. 5 of 2016 , delivered on 16 December 2016, the Final Court of Hong Kong preferred the Cooke and High Court of Australia decisions to the UKSC and PC.

In Ahsin v The Queen [2014] NZSC 153 the majority of the Supreme Court at [92] and William Young J (concurring in part) at [240] cited Chan Wing-Siu v R without discussing the point later discussed in  Jogee and in Miller.  Elias CJ [20] and William Young J [245 ff] dissented from the view of the majority that an offence by an accessory party is complete when assistance is given, considering that the assistance or encouragement “must continue at the time of the commission of the offence.”   Such distinction was relevant to the debate in that case as to whether absence of  “withdrawal” is an element of the offence, to be excluded by the prosecution (minority view), or a substantive defence, requiring evidence of withdrawal before such issue is left to the jury (majority [120); Elias CJ [21]; William Young J [244 ff].   The majority opinion on these points entails a breach of the unitary principle in relation to s66(1)(b): unlike others charged under s66(1), those charged under that provision appear to be entitled to a defence of withdrawal which does not negate assistance previously provided: William Young J at [253].


The Special Tribunal for Lebanon is empowered by its Statute to apply to party liability either the criminal law of Lebanon (Article 2), with its non-unitary approach, or a formulation of international criminal law (Article 3).  In Interlocutory Decision on the Applicable Law its Appeals Chamber has held that the principle favor rei entitles the accused to select whichever is more favourable to the defence.

Article 3 of the Statute provides for individual responsibility if an accused either infringed Article 2 or:

… Contributed in any other way to the commission of the crimes set forth in Article 2 … by a group of persons acting with a common purpose, where such contribution is intentional and is either made with the aim of furthering the general activity or purpose of the group or in the knowledge of the intention of the group to commit the crime.

The emphasised passage does not justify application by the STL of a Chan Wing-Siu approach  There is some support however for such approach in cited decisions at international law.

In “The End of ‘Modes of Liability’ for International Crimes” (2012) 25 Leiden Journal of International Law 165 Professor Stewart mounted a particular challenge to JCE III.    I add to his reference to Antonio Cassese’s essay [7] the discussion of international criminal law in the judgment of the Appeals Chamber of the  Special Tribunal for Lebanon in Interlocutory Decision on the Applicable Law STL-11-01/I 16 February  2011 (STL Casebook 2011 pp167-8, 177 and 182) paras 227, 248-9 and 259 n 381) , requiring Party B to be aware either of the criminal intent of Party A or of the substantial likelihood that Party A will commit the crime.

There may be noted however the citation of “JCE III” decisions requiring only that the secondary offender “(i) was aware that the resulting crime was foreseeable  as a possible consequence of the execution of the JCE, and nonetheless (ii) willingly took the risk that the incidental crime might be committed and continued to participate in the enterprise with that subjective awareness”(p174 para 241).  That is pure Chan Wing-Siu. But the Appeals Chamber did not apply such principle to the case of specific intent crimes such as terrorism, holding:

249 … the better approach is not to allow convictions under JCE III for specific intent crimes like terrorism… He must have the required specific intent for terrorism; he must specifically intend to cause panic or to coerce a national or international authority [according to the STL AC’s definition of terrorism].

This is not the occasion to debate the topic of Chan Wing-Siu type JCE III, which is not permitted by the STL Statute.  Or whether a Chan Wing-Siu approach might be warranted in the case of a commander such as General Yamashita (cf In re Yamashita 327 US 1 (1946)).


The STL AC’s approach to international criminal law coincides with the NZ watered down unitary principle of s 66(1) and its mix of unitary and non-unitary approach in s 66(2).  It conforms with Jogee in the case of terrorism;  other cases of JCE III cited in the decision do not.


[1] KNZM  former Judge of the Court of Appeal of New Zealand, President of the New Zealand Law Commission, New Zealand Member of the Permanent Court of Arbitration;  Appellate Judge and past President the Special Tribunal for Lebanon, Overseas Bencher The Inner Temple

[2] Ashgate 2013

[3] James G. Stewart, Complicity, in Oxford Handbook of Criminal Law (Markus Dubber & Tatjana Hörnle eds., OUP 2014),  at 539-540.

[4] Cassese’s International Criminal Law (3ed Oxford 2013) 162

[5] “it is impossible to define a legal concept, and […] the task of legal writers should be rather to describe the use of a word like ‘possession’ in the particular legal rules in which it occurs. ‘Possession’ in the legal sense has no meaning at all apart from the rules of law in which it is used as a tool of legal thought.  “The Concept of Possession in English Law” by Professor D.R. Harris QC in Oxford Essays in Jurisprudence (ed AG Guest, Oxford 1961) 69 at 70 drawing on Bentham and HLA Hart.

Lord Hoffmann in Meridian Global Funds Management Asia Ltd v The Securities Commission (1995):

… a reference to a company “ as such” might suggest that there is something out there called a company of which it can be meaningfully said that it can or cannot do something.  There is in fact no such thing as the company as such, no ding an sich, only the applicable rules. To say that a company cannot do something means only that there is no one whose doing of that act would, under the applicable rules of attribution, count as an act of the company.

 [1995] UKPC 5, [1995] 2 AC 500 at [10]

[6] Eg R v Rook [1993] 1 WLR 1005 (CA) criticized by AP Simester “The Mental Element in Complicity” (2006) 122 LQR 578.

[7] Nn 25 and 56

New Symposium: The Strangely Familiar History of the Unitary Theory of Perpetration

To date, I have largely used this blog to host debates about other people’s scholarly work. In this instance, I wanted to host a discussion about an article I authored for a Festschrift in honor of Yale Professor Mirjan Damaška, which is entitled The Strangely Familiar History of the Unitary Theory of Perpetration. I thought to invite a range of judges, expert practitioners from international criminal law (“ICL”) institutions as well as scholars from countries that adopt the theory of blame attribution I advocate for to comment on the idea of abandoning “modes of liability” in ICL entirely. Somewhat strangely, the long debates about these questions in the field have mainly involved academics from dominant Western countries, but none of the world’s leading experts from jurisdictions that adopt the unitary theory of perpetration have had an opportunity to engage with the debate about whether we should have forms of participation in ICL or do without them as per their own national systems. Both the article and this symposium are an attempt to bring these perspectives to the fore without, of course, prejudging how these particular commentators will see the issues in question or respond to my treatment of them in the article.

I begin by introducing the discussion’s relevance for international law. To do so, I reiterate an argument I recently made about the significance of these issues for global governance. Modes of liability, or forms of attribution as they are probably better labelled, can be fairly arid, technical, technocratic concepts in the theory of criminal law that are not normally of great interest to international lawyers. But I want to depict them in a way that highlights their great regulatory potential on an international plane. If one thinks of all of the harms in the world on the one hand, then all of the actors operating globally on the other, modes are attribution are those devices that exist between these two sets, reaching into the ocean of actors to tie them to particular atrocities. One can therefore understand how these concepts can have huge implications for global regulation, even though they are cast in fairly technocratic language that can be quite alienating to international lawyers. Of late, there is seemingly a rising recognition of this fact for a variety of global issues, including counterterrorism, foreign assistance, and business.

Against this backdrop, let me introduce the unitary theory. A unitary theory of perpetration is one that does not espouse different legal standards for different forms of participating in crime. So, whereas modern international courts and tribunals employ different legal tests to differentiate aiding and abetting from joint criminal enterprise, superior responsibility and indirect co-perpetration, a unitary theory of perpetration condenses all of these standards into a singular unified standard that only requires a substantial causal contribution to the consummated offense together with the blameworthy moral choice announced in the crime with which the accused is charged. It is worth noting, however, that there are pure, functional and sentence-based variants of this unitary theory (for discussion, see here, pp. 8-10), which come with different contours. For present purposes, however, the key aspect of the unitary theory I want to emphasize is that the formal legal elements of blame attribution remain constant across the different relationships actors bear to atrocity.

Initially, international courts employed a unitary theory of perpetration in practice. Although the Nuremberg and Tokyo Charters explicitly enumerated different forms of attribution, the Nuremberg Tribunal itself often just considered whether an accused was “concerned in,” “connected with”, “inculpated in” or “implicated in” international crimes. As many leading commentators now accept, this approach entailed a functional unitary theory of perpetration, namely, a system of blame attribution that declined to disaggregate modes of participation into formal legal concepts like aiding and abetting, superior responsibility or JCE, instead holding the substantive elements of blame attribution constant across the various roles different actors might play. In the modern era, however, ICL absorbed dominant Western doctrine to supplant this history, first from the Anglo-American system then from Germany. Whatever one might say about these shifts as matters of customary international law, it is striking that neither set of practices was informed by the experience of states throughout the world that had abandoned modes of liability. This article and mini-symposium introduce that missing comparative experience.

Conceptually, this article is the sequel to a more conceptual piece I authored some years ago entitled The End of Modes of Liability for International Crimes. In that earlier article, I had argued that a conceptually coherent concept of complicity involves its disappearance into a more capacious single notion of perpetration, and that by the same analytical method, all modes of liability in international criminal law should suffer a similar fate. Having worked on these issues for many years as a practitioner before coming to the theory, my sense was that practically speaking too, the unitary theory of perpetration offered a way out of a difficult legal morass for practitioners. In my experience, standards for blame attribution are sometimes harsh, often unprincipled, in a constant state of flux and inconsistent with the expressive aspirations of the field across diverse cultures. At the very least, then, my hope was to invite robust scholarly defenses of the system in place. Moreover, I was particularly motivated to undermine the justification, which I heard a lot in practice, that the existing approach in ICL is defensible because several large Western states adopt it. To my mind, that argument is not sound.

Several prominent scholars, whose work I respect, have since offered helpful defenses of the differentiated system in response to my earlier argument (see Werle and Burghardt, Jackson, Steer). Although these excellent initial works certainly advance the debate, I am also convinced that the comparative experience I attempt to offer in The Strangely Familiar History of the Unitary Theory of Perpetration represents another important piece of the puzzle that has not figured in these debates before now. In the hope that others will pick up on aspects of these discussions to defend the differentiated approach or deepen thinking about the unitary theory, I am excited to host a range of prominent judges, one practitioner, and a host of leading scholars from each of the jurisdictions I discuss to participate in this mini-symposium (see list of commentators here). I am honored to have leading experts speaking for their own hitherto neglected legal traditions.


The Historical Importance of the Kouwenhoven Trial

Last month, a Dutch Court of Appeal convicted Dutch businessman Guus Kouwenhoven for complicity in war crimes that were perpetrated in Liberia and Guinea a little less than two decades ago. Kouwenhoven was a timber merchant, who also shipped weapons to the Liberian President Charles Taylor in clear violation of UN arms embargoes, which Taylor used to brutally terrorize civilian populations. The Kouwenhoven judgment is now available in English (see here). I know something of the backstory behind the case through the work of Global Witness, have benefited from Dieneke De Vos’s very helpful overview of the judgment (see here), and learned from Ruben Carranza at the International Center of Transitional Justice’s discussion of the trial’s significance as part of broader aspirations for economic accountability in transitional justice (see here). In what follows, I consider the self-consciously provocative claim that Kouwenhoven is among the most important war crimes cases in the history of international criminal justice.

There are several reasons why the Kouwenhoven case might come in at the top of a ranking of war crimes trials, if one were prepared to undertake such a strange exercise. No doubt, readers will object that the assumptions underlying my prioritization go largely unannounced, are highly contentious, leave much out that others might place greater weight on and presuppose the propriety of criminal law trials as a response to atrocity in the first place. All this I concede, but to guard against the possibility that this precedent’s potentially tremendous implications for this and associated fields might escape detection by relevant audiences, I here set out a brief series of reasons why the case may well live up to the grandiose billing I  assign it. Also, because comparing this case to the Hissène Habre trial in Senegal, the Justice Case at Nuremberg and the prosecution of war crimes in civil wars for the first time at the ICTY would involve unnecessary political insensitivity, I avoid all comparison by merely highlighting the Kouwenhoven case’s claim to the title.

To begin, the Kouwenhoven case brings accountability to the worst elements of the arms trade in ways that are almost entirely unprecedented. I will not labor the characteristics of the dark side of the global trade in weaponry, except to point out that authors like Andrew Feinstein paint a harrowing picture of it. Before now, the very worst weapons transfers to regimes bent on mass violence have taken place in a regulatory vacuum characterized by the almost perfect absence of all forms of accountability. Violations of UN-imposed arms embargoes, for instance, have stimulated almost no legal scrutiny. I say “almost” because a study I undertook together with a team of researchers some years ago of all UN arms embargo violations as documented by UN Panels of Experts since 1993 discovered that of 502 alleged violations, only 1 led to legal accountability for sanctions violations.[1] Against this backdrop, the Kouwenhoven case is the first that holds a nefarious arms vendor responsible for complicity in African atrocities.

In so doing, the case complements other regulatory initiatives. As many will know, the signing of a UN Arms Trade Treaty has sought to make the human rights and humanitarian law records of end-users relevant to the legality of transferring weapons to them. The adoption of the treaty is salutary, long overdue, and aside from its own regulatory effect, it helpfully draws attention to the tremendous social upheaval caused by what Harold Koh once described as “a world drowning in guns.” At the same time, the very idea that a new treaty should be needed to make the human rights and international humanitarian law records of end-users relevant to the legality of transfers will likely leave criminal lawyers and moral theorists slightly perplexed; the pre-existing notion of complicity already achieves that purpose. By enforcing this pre-existing norm, the Kouwenhoven case employs expressive condemnation to transmit the underlying moral principle across surrounding initiatives.

In addition, the contours of the version of complicity the Kouwenhoven trial employs helps overcome the ubiquitous but overly restrictive debate about the doctrine in Business and Human Rights.[2] For many years, litigation brought predominantly under the auspices of the now (nearly?) defunct Alien Tort Statute debated whether an accomplice must share a principal perpetrator’s purpose to carry out an international crime (Kouwenhoven would have to have positively wanted his weapons to bring about international crimes in Liberia and Guinea), as compared with a knowledge standard (which made cognition rather than volition the touchstone for complicity, thereby criminalizing indifferent implication in atrocity for profit). Understandably, Business and Human Rights largely absorbed these standards. As I have argued elsewhere (see here), however, this binary was never a complete articulation of the universe of available standards for complicity globally. The Kouwenhoven case confirms this proposition by applying a less stringent, dual test.[3] Although the case only indirectly relates to human rights, it is important because of the lessons it holds for that adjacent field.

This brings us to the theory of accomplice liability. Like all notions of complicity, the iteration the court adopts in the Kouwenhoven case is theoretically contestable,[4] and one can only anticipate that it will be a central point of Kouwenhoven’s appeal. Nonetheless, whatever debate the standard generates before and beyond the Dutch judiciary, the Kouwenhoven trial is important because it not only brings these even more accountability-friendly variants of complicity to the fore, it also reiterates the pressing importance of theoretical debates about the shape the doctrine should take globally. In earlier work, I entitled a section of an article on this topic “Towards a Moral Theory of Accomplice Liability,” precisely because cases like Kouwenhoven should be grounded in defensible first principles that I am tempted to think should be universal. That a businessman is sentenced to 19 years in prison via application of this doctrine amplifies the need for critical engagement with these ideas, including the need for further thinking about assigning complicity different meanings from one jurisdiction to the next.[5] In both these respects, the case is again critically important.

Significantly, it also involves the Dutch prosecuting their own national for participating in African atrocities, thus marking a move away from the “victor’s justice” paradigm that has long characterised international criminal justice. The term “victor’s justice” evokes the one-sided justice dispensed at Nuremberg despite no shortage of Allied offending, but it remains a recurrent theme in critiques of modern international criminal justice too. The longevity of the critique is understandable given ICL’s uneven record of enforcement at the international level, but what of enforcement locally? Although Kouwenhoven certainly does not offer a silver bullet through the power politics that made internationalising trials necessary in the first place, it certainly marks an important and under-theorized turn away from victor’s justice. In particular, it sounds the emergence of a new element of what Kathryn Sikkink calls a “Justice Cascade” – some states are beginning to take responsibility for war crimes committed by their own nationals, including their businesspeople. From an historical perspective, the shift seems momentous – although the circumstances are very different, the reluctance to assume this responsibility after WWI was so pronounced that it nearly caused a revolution in Germany and a return to war in Europe (see here).

Relatedly, the Kouwenhoven case promises to dilute perceptions that ICL is, to paraphrase Rwandan President Paul Kagame, a neo-colonial tool. In my view, that claim was always overstated, but it is hard to overlook the spectacle of indicting the most senior Congolese and Sudanese political leaders for pillaging property that is epiphenomenal to African armed conflicts without addressing corporate responsibility for the same offense that drives resource wars. In the past, the ICC prosecutor’s explanation for the unique focus on Africans has been that Africa produces the most serious atrocities, taking human suffering as a metric. Yet, as I have argued elsewhere, there is a fallacy in this justification: it assumes that only Africans are responsible for atrocities in Africa. To entertain this assumption is to overlook the long history of commercial implication in and, in especially egregious circumstances, instigation of atrocity in Africa, of which Kouwenhoven is an exemplar. By holding him accountable, this trial goes some way in addressing the fallacy, undermining the neo-colonial critique, and improving the field’s wider claims to legitimacy.

The case achieves this improvement by, at least in this instance, simultaneously overcoming a strangely regressive element of modern ICL. As many have shown, the awareness of corporate implication in atrocity was so prominent in the minds of the architects of post-WWII justice that they very nearly staged a second prosecution of “industrialists” after the main Nuremberg trial. While this business-specific second trial never came to pass, businesspeople were tried within the Nuremberg trial itself as well as in trials staged in occupied zones throughout Europe. Whatever the political aspirations for these processes at the time, for better or worse, their example has not formed a discernable part of the stratospheric ascendance of modern international criminal justice. While public recognition of corporate implication in atrocity grew dramatically after the Cold War, the rebirth of modern ICL over the same period overlooked the commercial angle its earlier incarnation viewed as painfully obvious. The Kouwenhoven trial is a partial correction for this curious regression, which again makes it all the more important in relative terms.

In the end, it is too early to discern the Kouwenhoven case’s overall significance, let alone pronounce on its relative importance even if one were prepared to try ranking the unrankable. With time, it may well prove to be a distracting fig-leaf over a colossal systemic problem, a weak apology for the one-sidedness of ICL’s enforcement globally, or the exception that reinforces the rule of very near total absence of accountability for the worst forms of corporate malfeasance in war. For now, though, it appears more likely that the Kouwenhoven trial is an important milestone in an imperfect, vulnerable, and under-theorized cultural shift in the long struggle to stem atrocity. In case it needs saying, no one should confuse this shift, if it is that, as being a panacea for all the woes of commercial interests in mass violence or delude themselves that criminal justice is anywhere near a sufficient substitute for ethical and political engagement with these problems. Likewise, no one has claimed that cases like this are beyond reproach or that they perfect the system.

In these respects too, the Kouwenhoven trial is important because it invites rigorous, inter-disciplinary scholarly argument for and against this type of accountability.


[1] Judgment of Replacement Rol N° 4465-09 (19 January 2012) (Chile’s Supreme Court) <> accessed 2 February 2012. Arguably, the American case against Viktor Bout offers another example, although technically, he was tried for his willingness in a sting operation staged by the FBI rather than for his complicity in any of the many atrocities his weapons smuggling enabled.

[2] My preliminary thoughts about the complicity-debate in Business and Human Rights are online, here.

[3] My kind thanks to Dieneke Vos, who generously provided the following translation of the relevant passage in the Kouwehnoven judgment: “In accordance with established jurisprudence, to convict for complicity, it must be proven not only that the accused person’s intent was directed at the contribution or facilitation of that crime in accordance with article 48 of the criminal code, but also that his intent – whether or not conditional – was directed at the crime committed by a third person, in this case the commission of war crimes.” In other words, under Dutch law complicity requires what the court calls “double intent”: intent to contribute means, and intent relating to the specific crime. The court adds that the accused person’s contribution need not have been indispensable or causal. It is sufficient for the contribution to have “actually encouraged or facilitated the commission of the crime”

[4] For an overview of competing theories as well as recurrent conceptual problems in the theory of complicity, see here. For my attempt to ascertain the meaning of complicity in the ICC statute, see this blog post and the various expert posts it draws upon.

[5] I have argued elsewhere that a multiplicity of complicity standards that can couple with international crimes throughout the world is not desirable. See here. For my (qualified) response with Asad Kiyani to objections that this argument unjustifiably tramples on important social and cultural values that are reflected in diverse criminal doctrine, see here.

Judicial Rejection of “Specific Direction” is Widespread

I hadn’t thought to use this blog to write individual posts on new judgments or decisions in international criminal law but at the instigation of some friends, I’m persuaded to offer some very short reactions to a number of interesting blog posts in the past week on “specific direction.” The posts include commentary by Marko Milanović, Kevin Heller, Dov Jacobs and Jens Ohlin. Despite my initial reticence to re-engage with this topic in the blogosphere, it struck me that offering some thoughts on these ideas would also be an appropriate topic to end the year on and a convenient pretext for me to wish readers happy holidays and a prosperous new year.

By way of background, the latest discussions of “specific direction” in complicity stem from a judgment by the ICTY Appeals Chamber in the Stanišić & Simatović case (hereafter “Stanišić), which again rejected “specific direction” as a relevant aspect of the actus reus for aiding and abetting (see paras 94 – 109). As most readers will know, a differently constituted Appeals Chamber at the ICTY had adopted the controversial “specific direction” standard two years ago in a case called Perišić (paras 17 –74), before the same body (differently constituted) reversed itself in a very detailed judgment called Šainović (paras 1617 – 1651). But as I point out below, the judicial treatment of this question since Perišić is actually a lot thicker than this brief history would suggest: rejection of “specific direction” is far wider.

To review, I was opposed to “specific direction” when it first emerged in Perišić. I have always thought that “specific direction” as announced by the ICTY in that case was a misreading of casual language in Tadić. I won’t rehearse everything I wrote opposing “specific direction” at the time, but I do want to reiterate my empirical findings that the concept had no grounding in customary international law, comparative criminal law or previous discussions of the theory of complicity by leading experts (see here). In addition, I also wrote a blog post on Opinio Juris responding to Kevin Heller’s thoughtful defense of the doctrine. To the extent that experts at the national level have considered this problem, this is the established orthodoxy on the topic.

To complete this (overly) long introduction, I should say that I’ve always insisted that specificity is one of the many difficult questions in the theory of complicity (see here). For this reason, I’m looking forward to reading Sasha Greenawalt’s new draft article on the topic in the new year, which discusses “specific direction” by drawing on much of the voluminous and difficult literature on the theory of complicity (sorry for the delay getting to this Sasha). I very much admire Sasha’s work and I’m very pleased to have colleagues engaging with this thorny literature, especially if they reach different conclusions to mine. As I’ve mentioned public recently, I believe that a major public debate on complicity will be a great benefit to the world, even if no consensus ever emerges about its contours.

With all these preliminaries said and done, I set out below a very short list of thoughts about the most recent emanation of the “specific direction” debate in the blogosphere. I hope some of my reactions are helpful:

  1. I would like to suggest that our discussion of “specific direction” would be far clearer if we dropped the word “direction” out of “specific direction,” calling this the specificity issue in complicity or some other clearer label depending on what we mean. As I say, “specific direction” was very casual language first employed in Tadić that didn’t really mean terribly much before Perišić (see here). The definition the Perišić court gave the concept really does the specificity problem a disservice – recall that according to the Perišić definition “specific direction” entails: (a) an assessment of whether the recipient of the aid is “an organisation whose sole and exclusive purpose was the commission of crimes” (Perišić Appeal Judgment, para. 52); (b) whether the aider “endorsed a policy of assisting” (Perišić Appeal Judgment, para. 52); and (c) a distinction based on whether the aider is present at the scene of the crime or not (Perišić Appeal Judgment, para. 39, 70). I maintain that these positions are clear misreadings of complicity and that they only detract from: (a) whatever there is that’s genuinely problematic about the specificity problem in complicity; and (b) attempts to account for that residual difficulty in the theory of blame attribution writ large. I also believe that, because advocates seldom define what they mean by the term “specific direction”, we frequently talk past one another.
  1. My main contribution to the discussions in the blogosphere, however, is to place a far larger number of cases on the table for discussion. In particular, I think it bears noting that many courts have now rejected “specific direction,” such that a differently constituted court in Stanišić could not really have taken us back to Perišić even if it had decided to readopt the controversial concept. The list of cases that have rejected “specific direction” now includes:
  • The ICTY Appeals Chamber in Šainović (paras 1617 – 1651);
  • The ICTY Appeals Chamber in Popović (para 1758)
  • The ICTY Appeals Chamber in Stanišić (paras 94 – 109)
  • The ICTR Appeals Chamber in Nyiramasuhuko et al (see para 44 of Judge Agius’s Separate Opinion)
  • The Charles Taylor Appeal Judgment (see here, paras 466 – 481).
  • The ECCC in Case No 002/01 (see paras 707 – 710)

I got the sense from some of the commentary in the last week or so that the rejection of “specific direction” here again in Stanišić was somehow a farce given the composition of the bench in this case or the lack of reasoning substantiating the position. To my mind, the first of these arguments plays down that a variety of courts, at both trial and appellate levels, have rejected the standard. Thus, the supposition that a differently constituted court would have just reinstated Perišić as a norm in ICL as a field is, I think, unconvincing.

  1. To expand on this observation, I plot here the number of judges across all courts and tribunals who have voted for and against “specific direction,” from Perisic onwards. By my rough count, at least 20 different judges have had opportunity to pronounce on “specific direction” if one includes the Perišić court and everyone since in the different cases I list in 2 above. Three judges endorsed the concept in Perišić, and now Judge Afande has on entirely different grounds, but that still leaves a full 16 judges who have voted to have it overturned, some multiple times. By the by, this includes Judge Khan in Nyiramasuhuko, which means that even in Stanišić, the Agius/Afande coalition would likely have been inadequate to reinstate the standard had the judicial changes many lament not taken place. Nevertheless, even if Khan had been on the case and helped reinstate “specific direction” in Stanišić, the resulting judgment would still be at sharp odds with the vast majority of judicial thinking on the topic. Four times more judges think it is incorrect than are willing to endorse it.
  1. Looking through these more recent cases post Perišić, I read one additional judgment (not in my list in 2 above) that I think warrants mention. The ICTR’s Ngirabatware Appeals Judgment was presided over by Judge Meron and included Judge Liu, but it also involved three other judges who were entirely new to the issue. Logically, counsel for the defense argued, drawing on Perišić, that “the Trial Chamber erred in failing to determine whether the ‘specific direction’ requirement of aiding and abetting had been satisfied in his case.” (see para 145). The Appeals Chamber unanimously rejected this argument, despite their finding that “the Interahamwe used at least some of the weapons Ngirabatware distributed […] during the attacks and killings” (see para. 148) (my emphasis). Again, I’m not sure what “specific direction” means in its best light, but if it operates to deny complicity where conduct has a dual use, then surely Ngirabatware’s conduct was not “specifically directed” and he should have been acquitted. I suspect that people may argue about this given Ngirabatware’s intentions (separate from “specific direction”), but I wanted to highlight the case to suggest that even the minority of judges who advocate for this controversial standard in complicity are less than clear about when it applies and how.
  1. On the issue of substantive reasoning, I don’t necessarily share the concern about the absence of deep reasoning in the Stanišić Judgment. I take this position because the prior decision in Šainović was surely amongst the most meticulously researched judgments in the history of this discipline, drawing on the criminal law of an enormous number of states (see Šainović Appeal Judgment, paras 1617 – 1651). I’d written a doctorate that addressed the comparative law and theory of accomplice liability in ICL, then four years of further research on the topic thereafter, but still there were many sources in this judgment I’d never even heard of before. I can’t imagine what it took to acquire and analyse all these legal materials in such a short period of time, but however one views “specific direction” as a normative concept, I think we have to acknowledge that this depth of research and justification goes far beyond what criminal courts normally offer. In fact, to the best of my knowledge, it is without parallel on any other issue in ICL before or since. Accordingly, I didn’t see terribly much reason for the Stanišić Appeal Judgment to reinvent the wheel on this; it had been fully addressed previously. To be clear, no court has offered a compelling theoretical explanation of this problem, but I believe that task falls to academics.
  1. This brings me to Judge Afande’s apparent endorsement of “specific direction” by way of dissent in Stanišić, wherein he is the first and only judge to defend the concept since Perišić (Judge Tuzmukhamedov questioned the need to address it on the facts in Šainović but he did not opine on the propriety of the doctrine itself and Judge Agius has included a paragraph in two judgments maintaining his earlier reasoning without further argument). Although I respect Judge Afande’s attempt at finding a third way through a difficult legal problem and competing dissenting positions, it strikes me that: (a) the account he offers is no longer speaking about “specific direction” as espoused by Perišić (see my point 1 above); (b) his methods for this novel third way are at times highly suspect, like the use of dictionary definitions to cut through all previous debates; and (c) the resulting position is hard to reconcile with any of the different schools of thinking about blame attribution in the theory of complicity (see here). Perhaps others will defend Afande’s reasoning, but personally, I have some difficulty imagining that it will prove adequately convincing to bring so many judges around to readopting “specific direction” given the widespread judicial disagreement with it now. My guess is that only the in-depth work of scholars will be able to do that at this point, to the extent that they are able to establish that the problem of specificity (or whatever more accurate label we can attribute to the problem) cannot be accommodated anywhere else in conventional theories of blame attribution.
  1. Finally, a brief word about the changes to the composition of the bench in Stanišić. As my colleagues have pointed out, two of the judges were replaced with others before the Stanišić appeal hearing. I know very little about this backstory and have consistently steered clear of the various political controversies surrounding “specific direction”, but I confess that I don’t fully understand the complaint that the composition made the decision predictable or arbitrary. Judge Meron, himself an advocate of “specific direction,” appointed one judge for and another against “specific direction.” I see no scandal. The orders doing so were perfectly hum-drum, speaking about “the appeal management and case distribution needs of the Tribunal.” Isn’t this entirely normal and very banal? As we now know, it’s unlikely to have changed anything in the concrete case (Khan appears to be against “specific direction”) or the field as a whole (the vast majority of ICL judges clearly oppose “specific direction.”) As for predictability, I’m not sure what the problem with that is either. Many would argue that predictability is highly desirable in a criminal trial. Moreover, judges the world over almost always come with a known set of legal commitments. Think of the US Supreme Court. The parallel isn’t entirely direct, but I wanted to raise these points because I’m not convinced that this situation at the ICTY is quite as bad, surprising or irregular as some of the previous commentators have suggested. To my mind, the only thing that would be objectionably arbitrary is if, by chance, the very few international judges prepared to endorse “specific direction” again found themselves on a single appellate bench.

In any event, I hope some of the foregoing is helpful. I’ve written this uncomfortably quickly, so I hope readers will correct my errors, if there are any. Once again, I encourage scholars of all stripes to engage with this issue and the very many other difficult problems in the theory of complicity. As I mention, I believe they are part of an important struggle to lead ethically decent lives in a world that is at once highly inter-connected and very dysfunctional.

Happy holidays to one and all.