LJIL Symposium: James Stewart Responds to Jens Ohlin

[James G. Stewart is Assistant Professor of Law at the University of British Columbia]

Jens Ohlin, with George Fletcher and in his own right, has been a pioneer in bringing criminal theory to bear on international criminal justice. His earlier work warned us that our dogmatic insistence on ascertaining international criminal law in pre-existing sources of public international law risked undermining the inherently criminal nature of this adjudicative process and the fundamental notions of criminal law that must apply as a consequence. As is the case with the other critics who have written for this blog, my article is counterfactually dependent on his earlier groundbreaking work.

I think it appropriate to start by placing Ohlin’s comments in context. His admirable defense of the differentiated model of blame attribution presently in place in international criminal justice does not take into account that arguably the most prominent theorists even within his own jurisdiction, from Michael Moore to Sandy Kadish and Larry Alexander, all view complicity as conceptually superfluous. This does not respond in any way to Ohlin’s comments, but I do think it important to table the growing body of authoritative academic argument against the differentiated model international courts have unquestioningly absorbed. In many respects, my article is an attempt to do just that.

On another preliminary note, I fear that Ohlin’s criticisms might miss the real essence of the paper. Most importantly, he does not address the normative substance of “modes of liability” in international criminal justice. Both the title to his response (“Names, Labels, and Roses”), and the content of his remarks under that heading imply that the issue is just one of nomenclature, as if there were no normative significance to convicting someone of genocide for recklessly assisting the crime. But the major argument in my paper is that in its extremities, complicity violates the same standards that commentators have used to criticize the overreach of other “modes of liability” within the discipline, and that consequently, this mode of liability too is sometimes unjustifiably harsh or simply unprincipled.

Ohlin does not engage with this core claim, or the broader implications I draw from it for blame attribution writ large. While the tendency in criticisms of “modes of liability” in the field of international criminal justice thus far has been to focus on a single mode, I have argued that all “modes of liability” as a genus fall afoul of the basic conceptual principles used to chastise individual instances. This arises since “modes of liability” invariably employ static mens rea standards that are insensitive to the dynamic characteristics of mental elements across the many international crimes (recklessness as a requirement for complicity clashes with the purpose necessary for genocide or intention required for deportation). By failing to engage with the substance of this argument, Ohlin’s criticisms wrongly imply that all is well with international “modes of liability.” But I know he does not think that.

I come then to his criticisms. To begin, Ohlin argues that the unitary theory for which I advocate simply returns us to JCE as espoused in Tadić, but I have to disagree as a matter of both history and substance. On the historical level, the unitary theory returns us to Nuremberg, where most defendants were simply convicted because they were “concerned in”, “connected with”, “inculpated in” or “implicated in” international crimes. This, I believe, was (apparently again by chance not good design) a preferable approach to the indecipherable complexity that came soon thereafter to which he makes reference, that aside from erecting standards that are painfully difficult to comprehend, convict defendants for crimes they do not deserve. A unitary theory would correct not emulate this.

As this last comment suggests, there are major substantive differences between a unitary theory and joint criminal enterprise. For instance, a unitary theory would not permit JCEIII as announced in Tadić, since JCEIII acts as a prism through which the law escalates an accused’s liability beyond that set out in the elements of most international crime with which she could ultimately be convicted. By contrast, the unitary theory makes mental elements common for all participants in the crime. If some commonality between a unitary theory and JCE remains in that both the unitary theory and JCE convict the mastermind and the foot-soldier of the same crime, I content myself in the knowledge that both types of perpetrator contributed to the harm having made the moral choice designated as warranting the stigma the offense denotes. In contrast, the status quo Ohlin defends tolerates misapplications of responsibility.

Ohlin then claims that only the differentiated system ensures that convictions are “more governed by reasons and less by ad hockery.” As I mention above, sentencing guidelines within a unitary theory take most of the sting out of this argument. Moreover, the problem remains that, whether or not these “reasons” accord with defensible principles is entirely random. If the mental element for complicity is recklessness (alas, it varies depending on jurisdiction), then a match between this mental element and the requirements of the crime is entirely haphazard (some crimes allow recklessness, others do not). The real ad hockery then lies in allowing modes of liability to randomly skew culpability designated within crimes depending on chance couplings between crime and mode of liability.

Ohlin then argues something I think is very difficult to sustain—he makes peace with a world where bases for criminal responsibility mean nothing to relevant audiences. He illustrates his argument that it matters little that JCE or “indirect co-perpetration” mean nothing to lay audiences by appealing to other legal terms in property, anti-trust and intellectual property, citing the American rule against perpetuities as evidence of legal concepts nobody understands. But this parallel with other areas of law misses the communicative aspirations that are particular to criminal justice. At least since Joel Finberg’s expressive theory of punishment (and long before in non-English speaking systems), criminal justice has sought to express disapprobation, to convey denunciation, to communicate stigma. Indeed, these expressive theories are particularly in vogue as justifications of international punishment. If communication matters, one would think it important to make the message intelligible. This, lamentably, is not the case with “modes of liability” as they presently stand.

Finally, we come to Ohlin’s best point. Given my reductionist inclinations, why not also reduce all international crimes to a single broadly defined offense (Ohlin calls “Atrocity”), then allow judges the ability to parse out responsibility for “Atrocity” as they see fit within the sentencing phase. This neat reductio exposes concerns for legality when responsibility is so abstract that it is no longer adequately specific to forewarn defendants of their potential liability. But what Ohlin overlooks is that there is a competing conceptual need to guard against the propagation of “modes of liability” that have no meaningful deontological relationship with the criminal offenses they harness to stigmatize defendants. This competing objective militates in favor of a mirroring between elements in modes of liability and those in a crime, but by this process of conceptual harmonization, all normative difference between “modes of liability” dissolves.

This does not answer Ohlin’s criticism, but it does raise new questions: what is an appropriate balance between the need for specificity and the importance that modes of participation track elements of crimes? In other words, by what conceptual framework do we assess the trade-off between specificity and culpability? Perhaps here too we uncover another central philosophical issue that the discourse has overlooked in its piecemeal admonition of particular “modes of liability,” for flaws all share. My own sense, for what it’s worth, is that the unitary theory of perpetration marks a happier medium by ensuring that modes of attribution are theoretically coherent, while allowing the great diversity of international crimes to designate the many variations of harm with specificity. Nonetheless, I very much look forward to further discussions.

My thanks again to Jens Ohlin and all others for their helpful criticisms. I have learnt much from them.