Blog has moved, searching new blog...

Saturday, August 24, 2013

Why the Vienna Convention should not be applied to the ICC Rome Statute: a plea for respecting the principle of legality

Dapo Akande and Kevin John Heller are engaging in a really interesting discussion on how the application of the Vienna Convention on the Law of Treaties (VCLT) might allow us to interpret the Rome Statute to include the use of chemical weapons as a discrete war crime. Dapo argues that it can and Kevin, while agreeing with Dapo, regrets that this result would be attained at the expense of states’ consent and explicit desire to exclude such a provision in the statute.

What I find interesting in those posts is that they both take for granted that the VCLT in fact can be applied to the Rome Statute. Of course, any first year international law student will tell me that this is obvious and unquestionable: the Rome Statute is a treaty and therefore, the VCLT applies. However, I’m not entirely sure I agree.

I've always found the question of the applicable rules of interpretation to international criminal law statutes to be an underdeveloped aspect of the literature on the work of the tribunals. The applicability to the statutes of the VCLT, or at least the rules contained in it, has rarely really been questioned. The case law of the ad hoc tribunals is full of judgments and decisions, which either explicitly or implicitly refer to those rules, despite the fact that as UNSC Resolutions, the Statutes of these institutions should not necessarily be looked through that lens. In a recent decision, the Appeals Chamber of the Special Tribunal for Lebanon even went as far as to claim that the VCLT applies to “any internationally binding instrument, whatever its normative source". As for the ICC, Judges have, most would say logically, applied the VCLT to the Rome Statute as a treaty.

This situation is understandable. As international lawyers, the VCLT is our default go-to document to look for rules of interpretation of international documents. But I believe this fundamentally ignores the specific nature of international criminal law and the central role of the principle of legality. This is why, in an upcoming book chapter, the first draft of which can be found here, I suggest my own, somewhat unorthodox (according to everyone I've spoken to about them) views on the applicability of the VCLT to the Rome Statute in the first place. In a nutshell, what I argue in my Chapter is that the requirements of the principle of legality in ICL would warrant against the application of the broad and ultimately discretionary rules of interpretation of the VCLT.

For one, they have been thoroughly been misused in the past, with unacceptable references to the "object and purpose" approach to essentially introduce morality as a way to circumvent strict legality. Indeed, you often see broad references to the “end of impunity” or various variations on the protection of human dignity as part of the “object and purpose” of the Rome Statute to justify expansive (some say progressive) interpretations of the applicable law. Also, I think that a rule such as that of subsequent practice of States (article 31(3)(b), VCLT) would often not be compatible with the non-retroactivity of criminal law. 

Second of all, and more technically, I put forward 2 series of arguments against the application of the VCLT to the Rome Statute.

The first one relates to the clear existence in Statute (compared to other international criminal tribunals so far) of lex specialis rules of interpretation, in terms of in dubio pro reo, strict intepretation and the prohibition of analogy (article 22). This therefore would exclude the lex generalis rules of the Vienna Convention. 

The second argument is a bit different. I think that the "nature" of a document depends not on the document itself, in an absolute and abstract way, but on the entity applying it and the entities it is applied to. In other words, the Rome Statute might sometimes be considered as a treaty and sometimes not. When it is applied by the judges of the ICC, it is an internal application of the Statute and it is therefore not applied qua treaty, but rather as internal rules of the organization. On the other hand, if two States were to engage in a dispute on the interpretation of the Rome Statute (for example in relation to duties to cooperation or duties to surrender accused), then the Rome Statute would apply qua treaty between them, and the VCLT would arguably be a valid point of reference.

This second approach, of the possible dual nature of an international document, is not unheard of. For example, the question arose in the the Kosovo Advisory Opinion, which I think completely fumbled the question of whether the constitutional framework was relevant international law for the dispute. The ICJ said that it was, based on the fact that it was formally an UNMIK Regulation, adopted pursuant to powers granted by the UNSC. I must admit I initially agreed with the ICJ, but on further reflection I do believe that because in that context it was meant to be an internal legal document not aiming at having international legal effect, it was not relevant international law at all (see my LJIL article for further discussion on this point).

To clarify, I don’t suggest that my proposal removes by magic any difficulty in interpreting the Rome Statute. There will always be cases of ambiguity, real of perceived, that will probably require a balance of interests between different possible interpretations. I just want to reintroduce one interest that is somewhat often forgotten in these debates: that of the accused and more generally, the application of the principle of legality. These interests should come first in the discussion, not last as is often the case. For example, in the above mentioned STL Appeals Chamber decision, there is a lengthy discussion of all the different rules of interpretation contained in the VCLT, and only at the very end is it mentioned that, if nothing else works to solve an ambiguity, then the interpretation most favourable to the accused should be adopted. This, for me, is the wrong logic. The first rules to go to are the ones which favor the defendant.

Applied to Kevin and Dapo’s conversation, this doesn’t mean that I would necessarily disagree with them, just that I would approach things differently.

For me, Dapo’s excellent interpretation needs to pass an additional test, that of being foreseeable by the defendant. I’m also not sure it is not in violation of the prohibition of expansion by analogy.

As for Kevin’s points about the importance of State consent, I think that it is not always a good starting point. Indeed, I don’t care what States wanted. If they drafted an ambiguous provision, the interpretation most favorable to the accused must be adopted, even if the travaux préparatoires indicate that the other interpretation was favored. Drafters should do their homework. If they plan to send someone to jail for a considerable period of time based on the Statute, the least they can do is make this crystal clear in the wording of the provisions. On the other hand, in the specific case of chemical weapons, if there is wide public knowledge of States wanting to exclude from the Statute, then it can be relevant in going to show that prosecution specifically for such conduct was not foreseeable.

All in all, given the regular violations of the principle of legality in international criminal case law, my proposal therefore aims, beyond a change in the applicable rules of interpretation, at a change in the state of mind of those applying those rules, be they judges or academics.


  1. I don't disagree with the post in the slightest; I'd be delighted if the VCLT didn't apply, and I agree that even if it does, it cannot be applied without taking into account the specific interpretive rule in the Rome Statute. I know Dapo at least partly agrees; he made a similar point in his essay on sources in the Oxford Companion to International Criminal Justice.

  2. Glad you agree. I've been getting puzzled looks from a number of people since I've put my paper online and tried to explain it. Yes, Dapo does acknowledge the possible tensions between the VCLT and article 22. I just think he doesn't push the idea far enough. 

  3. At least there is a specific problem of penal law, nulla poena sine lege, but that problem is not new, see Germanys reservation to the European Convention on Human Rights. Maybe it is inevitable. What Jordan Paust did suggest was "instant custom", wasnt it? The mysteries of history...