In confrontations in particular disputes and especially in cases referred for third-party decision, it is common for one side to seek to overcome the prima facie "ordinary meaning" of a text by recourse to various supplementary means of interpretation. Because international decision processes are nonjury, the common law's elaborate code regulating admissibility of evidence has no analogue in international law; international courts and tribunals tend to allow the introduction of almost any material adduced by the parties but only occasionally to rely on it in their decision. Of late, however, there seems to be an increasing tendency for international investment tribunals to admit and rely on travaux preparatoires in their own treaty interpretations. The ultimate ex cathedra endorsement of this trend by an especially distinguished group of jurists, whose imprimatur all but ensures its installation as a rule of international law, appears in a recent decision of an ad hoc committee operating under Article 52 of the ICSID Convention. (1) In Malaysian Historical Salvors v. Government of Malaysia, which was decided on April 16, 2009, the ad hoc committee said, "In any event, courts and tribunals interpreting treaties regularly review the travaux preparatoires whenever they are brought to their attention; it is mythological to pretend that they do so only when they first conclude that the term requiring interpretation is ambiguous or obscure." (2)
We do not contest the accuracy of the ad hoc committee's summary of trends in what we may call, for convenience, the "Salvors' doctrine," that is, the automatic admissibility and review of any travaux preparatoires adduced by one or the other of the parties, without requiring a prior assertion, let alone decision, confirming the obscurity of the text or an ineluctably absurd reading resulting from the application of the method prescribed in Article 31 of the vienna Convention on the Law of Treaties. (3) Nor do we question that travaux are being consulted only for the purpose of confirming the reasonable interpretation that the application of Article 31 has produced. (4) There is, indeed, considerable evidence of court and tribunal practice supporting the conclusion that the carefully bounded and contingent role for travaux preparatoires prescribed in the Convention is increasingly honored in the breach in international investment awards and decisions, the most prolific source of jurisprudence in contemporary international law. Once the Article 32 material has been admitted, a tribunal finds itself under some pressure to address it in its decision, lest it be criticized for not having dealt with every matter before it. (5) Each subsequent application serves to reinforce the per se relevance and applicability of preparatory material, successively reaffirming the Salvors' doctrine.
We also do not suggest that travaux preparatoires mayor should always be ignored; that, too, would be inconsistent with the Vienna Convention, though we will argue that the belief that answers to textual obscurity or clear indicators of a proper interpretation can be found in travaux is, itself, rather fanciful. We do submit, however, (1) that Article 31 with its conditions for application is the default position of international law and policy and continues to be a cogent method for determining the shared normative universe of the parties to an agreement; and (2) that, no matter how general jurisprudence may have evolved with respect to the use of travaux, in treaties for the benefit of third parties (6)--the classic pacta in favorem tertii--additional compelling reasons support urging that resort to travaux preparatoires should hew faithfully to the very limited contingencies of Vienna Convention Article 32 and not to the Salvors' doctrine. Wholly apart from what the Convention requires, we suggest that reasons of economy, fairness, equality of arms, and the stability of expectation, the last a principal objective of investment law, compel such fidelity with respect to international investment treaties.
However much care parties may take to express their commitments with precision and to anticipate the various factual scenarios to which those commitments will relate, the predictability of the application of those commitments depends upon two things: first, upon a commonly accepted canon of interpretation and, second, upon the faithful application of that canon by those called upon to construe the commitments in question, whether they be the parties in the course of performance or decision makers resolving a dispute about that performance. Just as agreements facilitate cooperative behavior by stabilizing expectations about reciprocal rights and duties, so rules of interpretation of those agreements are designed to ensure that those stabilized expectations are respected. The foundational principle, pacta sunt servanda, remains unachievable if it is not accompanied by and implemented through agreed modes of interpretation.
International law's canon for interpreting international agreements is codified in the Vienna Convention on the Law of Treaties. Its provisions have become something of a clause de style in international judgments and arbitral awards: whether routinely and briefly referred to or solemnly reproduced verbatim, they are not always systematically applied. But a failure to apply the rules of interpretation properly may distort the resulting elucidation of the agreement made by the parties and do them an injustice by retroactively changing the legal regime under which they had arranged and managed their affairs.
The Vienna Convention has two major provisions on interpretation. The first, Article 31, bears the title or chapeau "General rule of interpretation"; the second, Article 32, bears the title or chapeau "Supplementary means of interpretation." It is clear from the respective chapeaus and the mandatory character of the word "rule" in Article 31...