The growing public interest in investment treaties and investor-State dispute settlement has prompted an increasing number of States to open to public view aspects of investment treaty negotiations. During the negotiation of the Transatlantic Trade and Investment Partnership (‘TTIP’), for example, both the European Union and the United States sought to ‘maximise’ transparency in the negotiations, recognising that this would be ‘indispensable’ to securing understanding and acceptance of the resulting agreement by domestic constituencies.

While the more transparent negotiation of investment treaties is intended to assuage concerns about the legitimacy of investment treaties and investment treaty arbitration, this practice has broader implications. Through transparency efforts, a growing volume of negotiating materials are now publicly available. The impacts of such materials beyond the negotiation phase is an open question. In these circumstances, how arbitral tribunals use such materials during treaty interpretation attains crucial significance.

Materials produced during the negotiation of investment treaties, commonly called travaux préparatoires (‘travaux’), are given formal significance as a ‘supplementary means’ of treaty interpretation under Article 32 of the Vienna Convention on the Law of Treaties (‘VCLT’). This provision provides:

Supplementary means of interpretation

Recourse may be had to supplementary means of interpretation, including the preparatory work of the treaty and the circumstances of its conclusion, in order to confirm the meaning resulting from the application of article 31, or to determine the meaning when the interpretation according to article 31:

(a) leaves the meaning ambiguous or obscure; or

(b) leads to a result which is manifestly absurd or unreasonable.

Article 32 reflects that materials like the preparatory work to a treaty might potentially reveal the shared intention of the treaty parties, and thus indicate the scope of treaty party consent.

Investment tribunals face three particular practical difficulties in navigating the use of travaux to interpret investment treaties. First, they encounter practical difficulties associated with ascertaining the existence of travaux and regulating its production in the arbitral proceedings. Second, tribunals may grapple with difficulties when determining the inferences that may be drawn based on travaux. Third, uses of travaux by arbitral tribunals may influence (State) perceptions of the legitimacy of arbitral outcomes.

In this blog post we discuss these issues by looking at different approaches to defining travaux, the challenges that might be caused by using travaux in the interpretation of investment treaties, and a potential response.


Defining Travaux: The VCLT’s Concept of ‘Preparatory Work’

The VCLT does not define the concept of ‘supplementary means’ by way of an exclusive list. Instead, it refers to two forms of supplementary means – ‘the preparatory work of the treaty and the circumstances of its conclusion’ – by way of illustration. These two concepts are themselves left undefined.

Investment treaty tribunals generally understand travaux at least to encompass agreed negotiating texts that have been shared between the parties. At the outer reaches, however, tribunals have also held the term to encompass documents not shared between the parties (for example, one State’s record of the negotiations), particularly when there is a dearth of travaux-type material. Tribunals have also accepted into evidence witness statements from persons present at the negotiations as a form of ‘supplementary means’ of interpretation under Article 32 of the VCLT. However, such a broad approach to defining ‘travaux’ and ‘supplementary means’ of interpretation holds risks.


Challenges Associated With the Use of Travaux

Article 32 materials – and travaux in particular – are particularly likely to be incomplete, imbalanced and/or, inaccurate. For older treaties, negotiated before the turn to transparency, such materials may also be non-existent or otherwise unavailable. Reference to such materials might also skew the interpretive exercise, for example by introducing self-serving or biased materials. References to travaux and associated materials by treaty interpreters might therefore entail risks. The practical utility of travaux to arbitral tribunals varies significantly in light of their ‘authenticity, completeness and availability’.1)Anthony Aust, Modern Treaty Law and Practice (3rd edn, Cambridge University Press 2013) 218. Travaux may show how the negotiators arrived at the final text, but may also contain misunderstandings and points that the negotiators ultimately discarded. Crucial decisions adopted may not appear in the record because they have been resolved in private, off-the-record meetings. And so on.

A further key practical difficulty associated with the use of travaux in investment treaty arbitration is that of access. Typically only one party to the dispute (the host State) will have ready access to travaux and like materials. As such, where a tribunal acknowledges the utility of travaux to the interpretation of the investment treaty before it, the tribunal needs to ensure that the investor-claimant’s status as a non-party to the investment treaty does not disadvantage it in the presentation of its case. As a rule, investment tribunals deem travaux to be discoverable. For example, many NAFTA tribunals have taken facilitative approaches during the document discovery process to assist claimants to access any travaux held by the respondent State. A minority of tribunals have adopted a more restrictive approach to the discovery of travaux.

Unjustified restrictions on the discoverability of travaux risk falling afoul of principles of equality and fairness that underpin investor-State arbitration. This has interesting parallels to early debates about recourse to travaux in a series of early cases before the Permanent Court of International Justice. The Court hesitated in using travaux where one of the disputing parties before it had not been a participant in the negotiations for the treaty at issue. Of course, this may in time change as more negotiating documents are placed online as part of a general push towards greater transparency of investment treaty negotiations. For most investment treaties, however, discovery rules remain an important means by which tribunals can broaden or narrow the potential role that travaux might play in the interpretative process.


Using ‘Supplementary Means’ to Interpret Investment Treaties: A Way Out of the Maze

These challenges in using travaux also explain why the utility of travaux to treaty interpretation is often limited in practice. As is well known, the appropriate role for ‘supplementary’ materials was a matter of considerable controversy during the drafting of the VCLT. It prompted the International Law Commission (ILC) to relegate these materials to a ‘supplementary’ role in the interpretive scheme of the VCLT.

Thus far, investment treaty tribunals have not articulated a general test to assess the relevance of materials under Article 32, instead preferring to confront these issues of relevance on a case-by-case basis. We consider that two approaches could be taken to determine the ambit of the concept of ‘preparatory work’ and/or ‘supplementary means’ under Article 32 of the VCLT.

First, it might be said that particular materials fall outside the scope of these concepts in toto, and thus outside the means of interpretation envisaged by Article 32. Such delineation might occur by reference to the types of materials capable of disclosing the parties’ collective intentions. The advantage of such an approach is that it results in a clear delineation of materials, indicating clearly which materials can play any role in the interpretive process. The disadvantage is that the approach is very binary and risks excluding potentially relevant materials from the interpretive process.

A second, better, option for regulating the use of travaux is to adopt a sliding scale of relevance by reference to the features of a given material. This approach reflects that the reference in Article 32 of the VCLT to ‘preparatory work’ is broad and in any case illustrative only. Tribunals might conclude that a wide range of materials fall within the concept of ‘preparatory work’ or – at least – the broader concept of ‘supplementary means’, but that their relevance in any given case will depend upon their particular features.

Under this second approach, interpreters would not exclude the material in entirety on the grounds of it not constituting ‘preparatory work’ or a ‘supplementary means’ of interpretation. Instead, having determined that the materials under Article 32 are relatively expansive, interpreters could nonetheless recognise differences between the types of materials that might be referred to under Article 32 by adopting a sliding scale of relevance. According to such a scale, an interpreter might, for example, give more weight to materials that are capable of manifesting the joint intent of the parties. While a unilateral internal document might therefore not be excluded ex ante it might nonetheless be given lesser weight than, for instance, a joint report on the negotiations signed by all treaty parties. Such an approach would encompass as relevant all material created during a treaty negotiation, but would confer most weight on materials demonstrating the common intention of the treaty parties.

One implication of our proposed approach is that the growing availability of travaux may render references to unilateral materials less important or even superfluous. This is particularly important for modern investment treaty negotiations, where the issue of State ratification is increasingly subject to widespread domestic appraisal and debate. In these circumstances, the adoption of a cautious approach to the use of unilateral ratification materials in treaty interpretation best balances the potentially one‑sided nature of such materials with their potential utility.

In a forthcoming article, we will propose a four-pronged test to ascertain whether a given material constitutes ‘preparatory work’ within the meaning of Article 32 of the VCLT. These criteria are linked closely to the text of Article 32 and its purpose, which envisages the use of supplementary means of interpretation to uncover the treaty parties’ intentions. The second approach is a better fit for the intent behind Article 32 of the VCLT. It may also assist tribunals to balance the rights and interests of parties to arbitral proceedings when determining issues associated with the access of both parties to travaux.



To ensure their continued legitimacy, investment treaty tribunals must utilise all means at their disposal to respect and uphold the bargains struck in investment treaties. Travaux have the potential to offer insights into the intentions underlying treaty provisions that are notoriously open-ended. The test we propose allows the treaty interpreter to harness the benefits of reference to travaux whilst supporting them to avoid its attendant challenges. The utilisation of the proposed sliding scale of relevance to determine the weight to be given to materials that might be referred to as travaux is important to both host States and investors, and to the systemic legitimacy (and independence) of the system more generally.


To see our full series of posts celebrating the 50th jubilee anniversary of the Vienna Convention on the Law of Treaties, click here

The ideas in this blog post are further elaborated in The Vienna Convention on the Law of Treaties in Investor-State Disputes: History, Evolution, and Future, edited by Esmé Shirlow & Kiran Nasir Gore (Kluwer, 2022). 


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Kluwer Arbitration


1 Anthony Aust, Modern Treaty Law and Practice (3rd edn, Cambridge University Press 2013) 218.
This page as PDF

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.