Art and Auto-Interpretation of Treaties

by Duncan Hollis

<br />

First off, I should say how much I admire Richard Gardiner’s book; it has already earned a prominent spot on my bookshelf and I expect it will become a regular reference work for me in any future interpretative exercises.  At the same time, I have lots of questions and comments about the book and the growing density of the VCLT system it purports to describe.  Let me start here with the premise that Richard himself uses to introduce the book–the idea that treaty interpretation is an “art” and that the VCLT now provides us with “rules for practising that art.”  (p. 5).  In doing so, Richard suggests that we can employ the VCLT rules to achieve a “correct” result (p. 6) and, more specifically, that the result should be a “single autonomous interpretation.”  (p. 30).  Accepting the proposition that treaty interpretation is an art, however, leads me to ask the normative question of whether as an art it should strive to some, single common approach for each case, not to mention the descriptive problem of whether states, international organizations, international tribunals, and other subjects of international law have actually done so. 

Simply put, I wonder what the artistic axiom — that beauty lies in the art of the beholder — does for our art of treaty interpretation.   What constitutes a “good” interpretation of a treaty may be as difficult to agree upon as what constitutes good “art.”  Indeed, I see the question of the continuing salience of Prof. McDougal’s work (or the concept of textuality raised by Professor van Damme) as essentially a debate over which treaty interpretation techniques we should celebrate and which we should disapprove.  We might analogize it to debates among various schools of art.  Do we consider photorealism (i.e., textualism) to be better than abstract art (i.e., the New Haven School)?  Or, is the answer somewhere in between a la impressionism (i.e., the VCLT rule)?  

Perhaps Prof. Gardiner’s point is simply that treaty interpretation has evolved to the point where the community of interpretators has agreed upon a single, acceptable technique for our art, namely the VCLT.  It’s become, to use Jan Klabber’s phrase, our battlefield for discourse.  I’d agree with such descriptions to a point–would any international lawyer contest the necessity of consulting the VCLT rule as the start of an interpretative exercise?  But, even assuming that’s true, I wonder about the assertion in at least other three respects. 

First, is it true that the VCLT sets the whole battleground for debate, i.e., is it the only technique or simply the dominant one.  It certainly could be an exclusive method if read broadly.  For example, I for one have never found Article 32 too troublesome on the question of using prepatory work.  I find plenty of space to “confirm” or “determine” whatever meaning a good faith interpretation of the text’s ordinary meaning in context produces.  Now, some might say I go too far, but here I agree with Jan that most lawyers will do this work and marshal the VCLT to confirm it whenever that research favors them (and invoke the “subsidiary” nature of the exercise when it does not).  

Second, I wonder how, if at all, we should conceive of the VCLT and petitions for “special” interpretative rules.  Are they simply variations on the VCLT, or competitors with it?  In particular, can we say the VCLT drives both the ECtHR’s rulings that hint at dynamic treaty interpretation with the WTO’s that don’t, because of special contexts–human rights vs. trade–that lead us to emphasize certain components of the VCLT rule over others.  Might similar reasoning justify broader readings of constituent instruments of international organizations even as it would generate strong objections if applied in a bilateral context?   

Third, even if the VCLT is, so to speak, the governing school of our art, I’m not sure it follows that it requires a single, common, autonomous intepretation in every instance.  On the contrary, the “crucible” approach adopted by the ILC (and Prof. Gardiner) leaves lots of different moving parts in play to produce competing “reasonable” interpretations of a treaty’s text.  Here, I’d cite another idea percolating around at the time the VCLT and the New Haven school gained prominence–Leo Gross’s theory of auto-interpretation.  Gross argued against too quickly disassociating the subject of interpretation (i.e., the treaty) from the interpretator (i.e., states, IO, or international tribunal).  In other words, worry less about the interpretative rule than finding a way to get someone to authoritatively apply it.  For Gross, the horizontal nature of the international legal order frequently hinders the prospects of doing this.  Parties can always agree on one “authentic” treaty interpretation, just as they can delegate to some international organization or tribunal authority to pronounce it.  And certainly, examples like the ECtHR or the ICJ reflect the product of such agreements.  But, in most cases states have not agreed to common or third-party interpretations.  More often, a state retains the right to self-interpret what the treaty means, i.e., an “auto-interpretation.”  And other states and subjects of international law can do the same.  The international legal order is thus more than likely to produce competing, unresolvable, auto-interpretations of a treaty’s text than any single, common interpretation.  The VCLT may have improved things by forcing states to adopt a common framework (and language) for their disagreements, but it alone can’t move subjects of the law beyond that framework to some, actual “correct” interpretation, however, dense the various components of the VCLT rule have become in recent years.   

Indeed, I wonder if Gross didn’t overestimate the value of authoritative interpretations by third parties to produce a single, autonomous meaning.  Witness the U.S. Supreme Court’s disagreements with the ICJ over the Vienna Convention on Consular Relations’s requirements of a remedy for a failure to provide consular notification under Article 36.  Some argue that the ICJ’s interpretation is authoritative, and the United States apparently agreed at least with respect to certain named Mexican nationals even as it found domestic law lacking a method to comply.  But, beyond those named parties, the Supreme Court believed the ICJ got it wrong, and offered its own competing vision of the article’s “correct” interpretation. Who’s right? Isn’t a decision as to which side gave the “right” interpretation essentially a function of who we ask?  (in fact, won’t those we ask answer based mostly on pre-conceived notions of whether international and domestic tribunals should have a vertical, horizontal, or diagonal relationship than the actual interpretations each employed?).  In the end, is there any way to achieve some finality of interpretation, absent an agreed reconciliation negotiated by the United States with Mexico, or the ICJ itself?  Otherwise, as much as I enjoy treaty interpretation, it remains a pretty subjective exercise, with each of us having our tastes and views of what’s good and bad work, even if we use only the VCLT to judge it.

2 Responses

  1. Such are the drawbacks of not having a supreme legal authority in the world.

Trackbacks and Pingbacks

  1. […] What International Law can learn from the Renaissance A very interesting discussion on treaty interpretation is ongoing at Opinio Juris, in particular, over textual vs contextual interpretation and the use of travaux preparatoires of treaties. The discussion has, so far, culminated in an intriguing post by Duncan Hollis, “Art and the Auto-Interpretation of Treaties”. (See […]