All posts by Filippo De Minicis

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.