Treaty interpretation: the role of party intention and good faith
I very much appreciated reading Eirik’s book. It is certainly a valuable contribution to the literature on treaty interpretation. Using the limited space available in a blog post, I will pick up on an idea introduced by Eirik in Chapter 3 – that there might be something of an interplay between evolutionary interpretation, party intention, and good faith. To use Eirik’s own words, “[e]volutionary interpretation may … be required by good faith”. Myself, I would describe the interplay as follows:
Articles 31-33 confer upon interpreters a discretion. For example, they leave to interpreters to decide whether in the interpretation of a treaty, the ordinary meaning of its terms should be defined based on language conventions that existed at the time of the conclusion of the treaty or conventions that exist at whatever time the treaty is interpreted. This discretion is not absolute, while treaties are always to be interpreted in good faith. To interpret a treaty in good faith is tantamount to interpreting it for a purpose, the purpose being to establish the common intention of its parties. Thus, the principle of good faith sets a limit to any discretion exercised by an interpreter under Articles 31-33. For example, although the choice between the historical and current-day ordinary meaning cannot be justified by direct reference to any rule of international law, in making this choice, the interpreter shall continue his/her search for the intention of the treaty parties.
This description of the law laid down in Articles 31-33 raises two questions. Both seem to be highly pertinent to a constructive debate on evolutionary interpretation. Why should evolutionary interpretation be practiced? And when? Eirik’s book does not engage in detail with any of those questions. I do not criticize him for this. Eirik’s emphasis is on explaining why evolutionary interpretation can be grounded in the idea of party intention, and as a delimitation of the purpose of a book, this is all fine with me.
However, I still find the two questions extremely interesting. Since Eirik and I seem to already have a common approach to many fundamental issues of interpretation, such as the concept of party intention and the role of good faith, I will take the liberty of introducing a few ideas of my own concerning the Why? and When?. Hopefully, we can have a further fruitful discussion about them. Even better, we may end up having a discussion involving more people than just the two of us. (For a more detailed inquiry into the Why and When questions, I would like to refer to a working paper recently posted on SSRN: U. Linderfalk and M. Hilling, “The Use of OECD Commentaries as Interpretative Aids: The Static/Ambulatory Approaches Debate Considered from the Perspective of International Law” (November 19, 2014)).
Why and when should evolutionary interpretation be practised?
In explaining why an evolutionary meaning should be given to a treaty term, international judiciaries often refer to its generic nature. This explanation does not give full justification of the position taken. It begs to be further clarified and refined, especially if we wish to have a constructive discussion of when evolutionary interpretation should be taken. Much of the missing elements can be found in the concept of reference.
Articles 31-33 of the VCLT build upon the idea of a treaty as a vehicle used by treaty parties to get their intention across to readers. To ensure successful communication, parties to a treaty exploit the referring possibilities of the treaty language or languages. Reference is the relationship held between an expression and what the expression stands for in the world at the occasion of its utterance. The referring possibilities of a language turn on the extension to which its conventions help utterers make readers or listeners aware of the existence of an intended relationship between an expression and a referent.
The reference of an expression being primarily a product of the intention of its utterer, an expression used by an utterer for the purpose of reference is a referring expression. Referring expressions can be described as members of different categories: singular referring expressions refer to single entities; general referring expressions refer to groups of entities; and generic referring expressions refer to classes of entities. While the referring possibilities of singular and general referring expressions depend on the language conventions applied at the moment of their utterance, the same does not necessarily apply to generic referring expressions.
To illustrate this point, consider the following extract from Article VI of the Treaty of Territorial Limits between Costa Rica and Nicaragua (see Navigational and Related Rights); consider especially the italicized expression “comercio”:
“La República de Nicaragua tendrá exclusivamente el dominio y sumo imperio sobre las aguas del río San Juan desde su salida del Lago, hasta su desembocadura en el Atlántico ; pero la República de Costa Rica tendrá en dichas aguas los derechos perpetuos de libre navegación, desde la expresada desembocadura hasta tres millas inglesas antes de llegar al Castillo Viejo, con objetos de comercio …”.
[English translation: “The Republic of Nicaragua shall have exclusively the dominion and sovereign jurisdiction over the waters of the San Juan river from its origin in the Lake to its mouth in the Atlantic; but the Republic of Costa Rica shall have the perpetual right of free navigation on the said waters, between the said mouth and the point, three English miles distant from Castillo Viejo, said navigation being for the purposes of commerce …”.]
As is evident from the immediate context of utterance, “comercio” is a generic referring expression. It refers to a class defined by some assumed institutional practice, probably the usage of the Spanish language by Spanish speaking people in general. By very definition, this means that (in contrast to any singular and general referring expression) the existence of the referent of “comercio” cannot be tied to any particular moment in time. Costa Rica and Nicaragua acted on either of the following two assumptions:
(1) If in 1858, Spanish speaking people refer to comercio as the purchase and sale of physical goods, but not as services of any kind, then they will continue using the term similarly throughout the entire life span of the Treaty.
(2) If in 1858, Spanish speaking people refer to comercio as the purchase and sale of physical goods, but not as services of any kind, then at some point during the life span of the Treaty they might start using the term differently.
Depending on whether Costa Rica and Nicaragua acted on the first or the second assumption, obviously, the referent of “comercio” will be defined by the usage of this term among Spanish speaking people either in 1858, or at each and every moment of reading of the Treaty.
Now, the problem is that categorizing referring expressions may not always be a simple task. This is for two reasons. First, language conventions allow users to use many terms to refer both either singularly or generally and generically. ‘Vessels’, for instance, may be used to refer both to a group of two or more vessels, and to a generically defined class of vessels. Secondly, as explained earlier, among generically referring expressions, a distinction has to be made between those whose referents remain stable over time, and those whose referents may possibly change or develop. (Let’s just call them static and dynamic generic referring expressions, respectively.)
Consequently, in order to be able to identify the referent of a generic referring expression such as, for example, “comercio” in the 1858 Treaty between Costa Rica and Nicaragua, interpreters would have to understand two things. First, they would have to understand that a class of activities, and not any single activity, is being referred to. Secondly, readers would have to grasp the assumed institutional practice, by which the class is to be defined. This observation is key to the justification of any proposition arguing that evolutionary interpretation should be practiced.
To illustrate, consider the finding of the ICJ in Navigational and Related Rights: the ordinary meaning of “comercio” is to be defined based on Spanish language conventions that are applied at the time of interpretation. This finding entails the proposition that “comercio” is a dynamic generic referring expression. Parts of the necessary justification of this proposition can be found already in the immediate context of the expression. The place occupied by “comercio” in Article VI clearly shows that it does not refer to any particular activity, but to a class of activity. The interesting question is how the ICJ may go about justifying the proposition that this class is to be defined by the Spanish language practiced in the 21st century, rather than the language practiced in 1858.
Let’s approach this issue generally. Obviously, when interpreters confront a generic referring expression, and decide that it is either static or dynamic, they would have to justify this choice. As I would suggest, justification can be given by resort to a series of general factors. Such factors include the following:
- The purpose of the treaty or single treaty provision interpreted
For example, the mere purpose of a boundary agreement – fixing a boundary – would seem reason to assume that treaty parties intended generic referring expressions to be defined by the relevant institutional practices existing at the time of conclusion, rather than at whatever time the treaty is interpreted. (See e.g. Guinea – Guinea-Bissau Maritime Delimitation.) The interpretation of “comercio” as a dynamic generic referring expression suggests that Article VI has a different purpose than to put an end to all territorial differences – perhaps to promote trade relations for the mutual benefits of the two nations. (See Navigational and Related Rights.)
- The intended life-span of the treaty
If treaty parties assumed that a treaty would remain in force for a very long time, this is reason to assume that they intended generic referring expressions to be defined by the relevant institutional practices existing at whatever time the treaty is interpreted, rather than at the time of its conclusion; and, possibly, vice versa. (See La Bretagne Arbitration.)
- The understanding of other temporal references in the treaty
For example, tax treaties typically contain repeated reference to domestic legislation. The common understanding is that such expressions refer to the laws in force at whatever time the treaty is interpreted. This observation helps establish the referents of other expressions in a tax treaty. If tax treaty parties conceived of a generic referring expression such as “the law of the Contracting State” as dynamic in nature, then why should we assume that they conceived differently of other generic referring expressions?
- Possible third party effects of the treaty
Democratic theory suggests that if a law has onerous consequences for a person (NN), to confer legitimacy on this law, it would have to be introduced by a legislative assembly, the composition of which NN was herself able to potentially affect (such as for instance by taking part in free public elections). In so far as concerns tax treaties, for example, at least those concluded between states known for otherwise honouring democratic ideals, their third party effect consequently works to inform interpretation. To establish the meaning of a generic referring expression it would seem to make sense to ask whether the respective parliaments of the two of more parties, when incorporating the treaty in domestic law, could possibly acquaint itself with the relevant institutional practice or practices, such as for instance the language of international taxation. A concern over the democratic interest of the tax-payer is reason to assume that parties intended generic referring expressions to be defined by tax treaty language conventions existing at the time of conclusion of a tax treaty, rather than at whatever time the treaty is interpreted.