MJIL Symposium: A Response to James Stewart by Darryl Robinson

MJIL Symposium: A Response to James Stewart by Darryl Robinson

[Darryl Robinson is an Assistant Professor at Queen’s University, Faculty of Law]

This post is part of the MJIL 13(1) symposium. Other posts in this series can be found in the related posts below.

I am very grateful for James Stewart’s comments on “How Command Responsibility Got So Complicated”. Professor Stewart and I are engaged in similar projects (criminal law theory and international criminal law (‘ICL’)) and immersed in similar literature, so our discussions are always very helpful to me, even though we at times reach different conclusions. Professor Stewart raises several interesting points, and I cannot quite do justice to all of them. I offer the following thoughts on the main points.

As a preliminary point, Professor Stewart rightly notes that people at the Tribunal had done a frenzied review of the relevant literature and so were at least aware of these issues. I take that point very much. Academics are often quick to criticise courts and institutions for their alleged failures to consider this or that issue, when perhaps the relevant actors were in fact deeply aware of it but chose not to elaborate on it given the hundred other priorities they had to attend to. I also sympathise with judges, who are either criticised for failure to elaborate on theoretical underpinnings, or alternatively are criticised for their wordy, theoretical decisions. For precisely these reasons, I ‘emphatically acknowledged’ that the Tribunals were operating in a pioneering phase, dealing with countless questions and constructing doctrinal rules from diverse authorities, and hence could not give detailed consideration to every fine point.[1]

Nonetheless, my point still stands that — with the benefit of hindsight and without criticism of the participants — ICL reasoning has been inadequate in its in wrestling with fundamental principles. Rather than being hasty, I think I have demonstrated this carefully and at length. For example, see the employment of doctrinal arguments as answers to more fundamental principled questions.[2] I do agree that in some instances ‘an absence of judicial reasoning just conceals issues too complex to articulate’ — an institution can’t go into every detail. However, that doesn’t work for the command responsibility-causal contribution problem. Once we pierce through the obscuring doctrinal arguments, we discover a clear internal contradiction with the system’s stated fundamental principles. Accordingly, we have to either fix the contradiction or else furnish our explanation why there is actually not a contradiction. Again avoiding assumptions, I expressly left open the possibility that the Tribunal or someone might indeed have an explanation to show there is no contradiction. But in the absence of such an explanation, we have what seems to be a very clear contradiction.

Secondly, Professor Stewart asks whether ‘failures to punish’ might be said to contribute to crimes where there is some prior agreement that there will not be punishment, or a tacit agreement, or an enabling culture. I heartily agree that a commander can contribute to crimes in such a manner, and thus be party to crimes by command responsibility. I attempted to address precisely this prospect, and I too drew on Blagojevic, but I see that my discussion of this was scattered in different places and was not particularly evident.[3] Now, even in those examples, I would not say that the ‘failure to punish’ the crime contributes to the already-completed crime. Rather, it is the prior signaling that there will not be punishment that contributes. Such a signal (or agreement or culture) is an illustration par excellence of ‘failure to prevent’. ‘Failures to punish’ can of course still make causal contributions in another way. The failure to punish crime A can facilitate or encourage crimes B and C, grounding command responsibility in those crimes. My culpability concern was at all times explicitly restricted to cases outside of such examples, where the commander’s failure to punish did not contribute in any way whatsoever to any crimes.

Thirdly, Professor Stewart expresses concern that I leave the door open to a ‘ratification’ theory, ie. that I am willing to revise my views if a convincing account of culpability without causal contribution is developed. Professor Stewart is concerned that this is inconsistent with the rest of my argument and with the culpability principle. His question is very valuable: it leads me to realise that I might better illuminate my position by clarifying the hierarchy of my commitments. I explain this idea about a hierarchy of commitments in more detail in a new work outlining a liberal account suitable for ICL.

My commitment is not to ‘causation’ per se. My deep commitment is to ‘justice’. The best available arguments lead me to believe that a just system of ICL complies inter alia with the culpability principle. In turn, the best available arguments lead me to believe that the culpability principle requires that accessories must make some sort of ‘contribution’ to the crime in order to be held liable for that contribution. So, the only reason I am concerned about causal contribution is because it is what the culpability principle appears to require. ICL itself appears to recognise this.

If however an argument comes along for a convincingly better way of thinking about culpability — and Kutz, Sepinwall and others are generating possibilities — then I will change my mind about causal contribution about twenty seconds later. My commitment is to justice, and thus to culpability, and only derivatively to causal contribution. Until then, based on the best available arguments, it is clear that command responsibility for the subordinate crimes without causal contribution contradicts the culpability principle. It also contradicts the principle as stated by ICL.

Fourthly, Professor Stewart is concerned that ‘risk aggravation’ — which I regard as one plausible understanding of the necessary ‘contribution’ by an accessory — is too elastic. Several of his points are inextricably interlinked to a different debate, in which Professor Stewart has argued for the rejection of the principal-accessory distinction and an ‘end of modes of liability’ in ICL. His article is a commendable example of sophisticated, thoughtful scholarship, even though I happen not to agree with the most ambitious conclusions (see here and see reactions of others here and here). In the present debate, Professor Stewart argues that causation (as opposed to a more elastic ‘contribution’, such as risk aggravation) is essential because ICL crimes require causation. He argues, for example, that murder requires causation of death and hence one cannot be labeled a ‘murderer’ without it. I agree, but I think there is a difference between being labeled a ‘murderer’ and an ‘accessory to murder’. The first requires ‘causation’; the second merely requires ‘contribution’, facilitating or encouraging the crime of the perpetrator, so that one is indirectly responsible.[4]

For those who accept the principal–accessory distinction, I think ‘risk aggravation’ is an appropriate and plausible reading of the contribution requirement (including the requirement in art 28 of the Rome Statute of the International Criminal Court), from a doctrinal, deontological, and consequentialist point of view. I realise that Professor Stewart, who rejects the distinction, will not find this very convincing. He might object that one is still getting blamed in relation to the same crime. Which is true, but one is blamed for a different role in relation to that crime.

Fifth, Professor Stewart suggests that inquiry into causation by omission is needed. His suggestion is intriguing and I hope and expect to take it up. This article was about unraveling the discourse, exposing the first thorn that created the need for the subsequent contortions and compensations. However, as noted in the paper, the philosophical debate over whether omissions can have causal impact at all is of course highly relevant. Thus I noted that the question is contested and debated, and I explained why I did not go into that particular debate in this article although I did outline my underpinnings. I expect to return to the question in a book-length treatment of some principled problems for ICL.

Lastly, I can offer one final minor clarification. Professor Stewart agrees with what he reads as my ‘tone of disdain at the extraordinary position’ adopted by the Tribunals. Written language is an indeterminate and imperfect medium, so I just want to clarify that ‘disdain’ is not quite what I wished to convey. Internal contradictions and culpability departures are of course to be taken seriously. So, perhaps I should at least confess to ‘chagrin.’ But the tone I am striving for is one of dispassionate, fair, careful inquiry. The Tribunal position is ‘extraordinary’ in a quite literal sense — ie, not in accordance with the ordinary understanding of culpability. Thus, if successor commander liability is to be entertained, a convincing alternative account of culpability would be needed.

I am very grateful to Professor Ilias Bantekas, Professor Jens Ohlin and Professor James Stewart for their insightful comments and reactions.

[1] Darryl Robinson, ‘How Command Responsibility Got So Complicated: A Culpability Contradition, Its Obfuscation and a Simple Solution’ (2012) 13 Melbourne Journal of International Law 1, 25.

[2] Ibid 25–9.

[3] Ibid 14 n 48, 18 n 67–8, 48 n 225.

[4] This position is developed in ibid above n 1, 12–15, 41–5 (drawing inter alia on the work of Gardner, Kadish, Dressler, Ashworth, Ambos, Hart and Honoré, and ICL jurisprudence on principals and accessories).




Print Friendly, PDF & Email
Articles, International Criminal Law
Notify of
John C. Dehn

In my humble opinion, the failure to punish aspect of command responsibility is an artifact of the natural law concepts underlying both the law of nations and one of its most significant branches, the law of war. Allow me to quickly explain. At its origins, the law of nations was based in natural law.  Thus, violations were deemed intrinsically wrongful and punishable by “collective” remedies of reprisal or even war. As the law of nations developed, it allowed the substitution of individual for collective remedies.  States could obviate a state-to-state remedy by providing or imposing one against an offending individual.  For these reasons, the U.S. Congress was given the power to punish “offences against the law of nations,” and (arguably) exercised that power through the Alien Tort Statute, granting a federal civil remedy to offended aliens (See Anthony Bellia and Brad Clark’s excellent Chicago Law Review article, The ATS and the Law of Nations).  It also did so through other statutes like the Neutrality Act, which imposed criminal punishment for those who from within U.S. territory intiated or waged war against a foreign sovereign with which the U.S. was at peace. Winthrop articulated the relationship of individual to collective punishment… Read more »

Jens David Ohlin

Just a quick clarification of my citation to Yamashita.  Yes, it is a failure to prevent case.  But that’s my point.  If one argues that command responsibility is akin to a separate offense, based on derelication of duty and not requiring causation, that conclusion fails to accord with cases like Yamashita that treat command responsibility as a form of vicarious responsibility. 
Hence my suggestion that going forward we could recognize both, depending on which prong of the doctrine you are talking about.  So failure to prevent requires causation and is a full mode of liability, but failure to punish is a separate offense based on dereliction of duty, and not requiring causation.


Yamashita was not a form of strict or vicarious liability as such — check the facts re: notice from his JAG officer, etc. John — not merely for failure to punish b/c he was on constructive notice of torture, etc.  The U.S. Theatre Judge Advocate Review of his conviction should be checked.  See 57 Military Law Review 99, 181-82 & n. 310 (1972).  See also the survey of several convictions for dereliction of duty in that article.

John C. Dehn

Jordan, all I said was that Yamashita involved a failure to prevent rather than a failure to punish.  I did not intimate that it was a case of strict or vicarious liability.  The facts were convoluted and, in fact, I agree with the dissent that there was a significant question as to whether Yamashita had any reasonable means to prevent or remedy the violations under the circumstances.  However, the Court did not believe it had the power to look behind the conviction.  It would only review whether the commission acted within the traditional scope of its jurisdiction.

What I did say strongly implies that the failure to punish aspect of command responsibility is a form of vicarious liability anchored in evolving law of nations principles applied to law of war violations.

Thanks for the clarification, Jens.  It seems to me, as is obvious by now, that the theoretical origins of failure to punish command responsibility might be inconsistent with a “cause in fact” theory of culpability.