4 Interpretation and modi fi cation
4.4.1 Interpretation versus modi fi cation of treaties
The role of ‘subsequent practice’ in the law of treaties is a test case, though we will not test a theory on a set of facts, but whether a particular doctrine fits into a given theoretical framework. Subsequent practice is an ideal candidate for this, because it connects with nearly all of the topics we have discussed above.
Subsequent practice is riddled with uncertainty. Not only is there academic dis- agreement about its precise importance and breadth, but the very placement of this concept at the fringe of debates on interpretation gives rise to higher-level problems. Thus it forms the basis for a discussion of uncertainty at the highest echelons of international legal theory in the next chapters. Not only are we here at a junction between perception and action, we can also see the clash of two different sources of international law and thus it is exemplified here (Section 4.4.3) what will be discussed below (Chapters 5 and 6).
The recently reinvigorated doctrine of subsequent practice claims that the behaviour of states parties to an international treaty after its entry into force and, within international institutions, of international organs created by the constitut- ing instruments, is important from a legal perspective. In its original form (codified in Article 31(3)(b) VCLT) subsequent practice is a factor in the interpretation of the treaty, but there are also claims that subsequent practice can modify treaty law.
The difficulty, say those wishing to see subsequent practice modify treaties, lies in distinguishing interpretation from change. The makers of the treaty in question – the states parties – declare by their acts and statements how they see the treaty.
Where does determination of the meaning end and the change begin, especially as the states parties as Herren der Vertrọge198 can make, unmake and do what they like to a treaty? It can be pointed out, however, that interpretation is not change, even though it may seem so. Interpretation may seem difficult or even impossible to distinguish from change,199 but one can only show that we experience difficulty in ascertaining the borderline, not that it does not exist. That would be confound- ing the ontology of norms with the problems of epistemology.
198 Rudolf Bernhardt, Vửlkerrechtliche und verfassungsrechtliche Aspekte konkludenter Vertrag- sọnderungen, in: Hans-Wolfgang Arndt et al. (eds), Vửlkerrecht und deutsches Recht. Festschrift für Walter Rudolf zum 70. Geburtstag (2001) 15–22 at 16.
199 Rudolf Bernhardt, Interpretation and implied (tacit) modification of treaties. Comments on arts. 27, 28, 29 and 38 of the ILC’s 1966 Draft Articles on the Law of Treaties, 25 Zeitschrift fỹr auslọndisches ửffentliches Recht und Vửlkerrecht (1967) 491–506 at 499; Karl (1983) supra note 125 at 39; Georg Ress, Die Bedeutung der nachfolgenden Praxis für die Vertragsinterpreta- tion nach der Wiener Vertragsrechtskonvention (WVRK), in: Roland Bieber, Georg Ress (eds), Die Dynamik des europọischen Gemeinschaftsrechts. Die Auslegung des europọischen Gemeinschaftsrechts im Lichte nachfolgender Praxis der Mitgliedstaaten und der EG-Organe (1987) 49–79 at 61, 64; Sinclair (1984) supra note 12 at 138; Waldock (1965) supra note 32 at 60 (para 25).
4.4 Interpretation and modification 125
As Wolfram Karl points out in his exhaustive study of the subject,200 the confu- sion has to do with the perception of what the ‘treaty’ is. This, in turn, determines what a change of treaty is. Karl identifies three uses of the term. First, ‘treaty’
could refer to the legal transaction which creates the treaty norms; second, it could mean the norms themselves; third, it could denote the instrument, the document which is the result of the transaction and the expression of the norm.
He therefore argues that ‘change’ could refer to a change of the instrument or text
‘which does not necessarily include a change of the [norms themselves]’,201 or to a change of the norm itself.202
The question is: what is the nature of the norm in international treaty law?
This has already been discussed in Section 4.2.3. If the text is the norm (as was argued there), one could say that only a change in the text would constitute a change of the treaty, while experience shows that the text remains, even if subsequent practice ignores it. The notion of the Wortlautschranke comes in at this point. If an alleged meaning is beyond the frame of possible meanings,203 if it transcends anything the words could ‘legitimately’ mean – being aware that the determination of the frame itself exists, but may be impossible to ascertain (Section 4.2.4) – then the treaty would be changed if it could (legally) be changed without touching the text. It is the approach that is different with interpretation and modification. On the one hand, there is the scholarly search to cognise the meaning,204 while on the other hand there is a decision, an authoritative act, as mentioned in the discussion of ‘authentic interpretation’ (Section 4.2.4).
Traditionally, subsequent practice has been used as a tool in treaty interpretation and as such it is admitted by most writers205 and in numerous dicta of inter- national tribunals.206 We will begin with a discussion of the doctrine’s value as interpretation before looking at the justification for claims that subsequent
200 Karl (1983) supra note 125.
201 ‘mit dem nicht unbedingt auch eine Änderung der Rechtslage einhergeht’; Karl (1983) supra note 125 at 10.
202 Karl (1983) supra note 125 at 9–10.
203 Hexner (1964) supra note 55 at 124.
204 Ress (1987) supra note 199 at 62–63.
205 Amerashinge (1995) supra note 63 at 198; Bernhardt (1963) supra note 4 at 126; Bernhardt (2001) supra note 198 at 17; Salo Engel, Procedures for de facto revision of the Charter, 59 American Society of International Law Proceedings (1965) 108–116 at 114; Fitzmaurice (1952) supra note 9 at 20–21; Fitzmaurice (1958b) supra note 10 at 223; Jacobs (1969) supra note 6 at 327; Karl (1983) supra note 125 at 123–194; Karl (1987) supra note 55 at 84; Kửck (1976) supra note 10 at 42;
McNair (1961) supra note 12 at 424; Pollux (1946) supra note 31 at 78; Ress (1987) supra note 199;
Ress (2002) supra note 55 at 27–30 (MN 27–33); Sinclair (1984) supra note 12 at 136; Skubiszewski (1983) supra note 53 at 896; Waldock (1965) supra note 32 at 59 (para 23).
206 The list is far too large to give at this point. Wolfram Karl provides a comprehensive overview of the relevant jurisprudence (Karl (1983) supra note 125 at 123–194). The Permanent Court of International Justice as early as 1922 intimated that the practice of states parties to a treaty was relevant for its interpretation: ‘[T]he Court might . . . consider the action which has been taken under the Treaty.’ Competence of the International Labour Organisation, Advisory Opinion of 12 August 1922, PCIJ Series B No. 2, 3 (1922) 39.
Uncertainty in International Law 126
practice can change treaties. When reading through the numerous references, in particular those of older date, one is struck by their uniformity. For McNair, for example, ‘the relevant conduct of the contracting parties after the conclusion of the treaty . . . has a high probative value as to the intention of the parties at the time of its conclusion.’207 His statement is typical of older writings, for there subsequent practice is admitted as a means of discovering original party intent.208 These three words have relevance for how one sees the role of subsequent practice in treaty interpretation. Indeed, their use is indicative of the approach one takes to interpretation, which determines what one makes of subsequent practice.
The key to understanding the role of subsequent practice lies in a number of dichotomous pairs of concepts.
The ‘ratione temporis’ distinction perceives the treaty either as static or as dynamic order (Section 4.1.3). On a static approach, the conclusion of a treaty is the key moment for treaty interpretation. The norms created are fixed; Baxter’s
‘photograph’209 is put into the fixing-bath with the meaning of words to remain as they are at that moment. The dynamic approach210 sees the treaty’s conclu- sion as merely one important moment among others. The meaning of treaties changes over time, with the circumstances – and without the parties being involved. The choice of approach ratione temporis determines one’s position with respect to ‘inter-temporal law’: the meaning of words is taken ex tunc or ex nunc, respectively. If one follows Kelsen to define legal interpretation as finding the possible meanings of a norm, the point here would be the different temporal reference points where meanings would be ‘fathomed’. These would either be the possible meanings at the time of a treaty’s conclusion or at the moment of interpretation.
The problem is: why would a treaty text have dynamic properties? At the point of its conclusion and barring formal amendments, the text will remain the same. The question then becomes: what is the text or what does it represent?
Is it the expression of party consent or of some trans-positive norm, in any case something beyond the words? Or is the text the law itself, as argued in Section 4.2.3? It may be impossible to decide. If the ‘true’ treaty lies beyond the text, why do we need a text? The temporal reference point for the possible mean- ings is also unclear; would the possible meaning necessarily be static or even a-temporal? Kelsen himself agreed that his frame theorem allowed for an ex nunc interpretation:
207 McNair (1961) supra note 12 at 424.
208 Treaty of Lausanne (1925) supra note 60 at 24: ‘The facts subsequent to the conclusion of the Treaty of Lausanne can only concern the Court insofar as they are calculated to throw light on the intentions of the Parties at the time of the conclusion of that Treaty’ (emphasis added).
209 Richard R. Baxter, Multilateral treaties as evidence of customary international law, 41 British Yearbook of International Law 1965–66 (1968) 275–300 at 299.
210 Karl (1983) supra note 125; Karl (1987) supra note 55; Ress (1987) supra note 199; Ress (2002) supra note 55.
4.4.1 Interpretation and modification 127
That the law is open to more than one interpretation is certainly detrimental to legal security; but it has the advantage of making the law adaptable to changing circumstances, without the requirement of formal alteration.211
It is conceivable within Kelsenian theory that either the ‘interpretative [frame itself ] may (and probably will) shift in the process of time in conformity with changing circumstances’212 or that some possible meanings become impossible while other meanings become possible. A shift of the choice of one meaning to another by an authoritative organ would also bring dynamism, whether the one or the other is considered a correct meaning or not.213 The difficult question is whether the frame of possible meanings can shift over time. It could be said that it cannot – international treaty law is the text which remains; therefore, its meanings remain. The correct temporal reference point is ex tunc; anything else constitutes a change.214
The ‘ratione personae’ distinction takes either a subjective, party-oriented, or an objective, independent, view of what the meaning of the treaty is. What goal does the use of subsequent practice have? Is it directed at proving intent or expression (Section 4.1.1)? A subjective approach sees the parties in a position of exclusive control over the existence as well as the interpretation of the treaty. They truly are its masters and state will and sovereignty are to be the guiding principles.215 Consequently, subjectivists tend to see the meaning expressed in party intentions;
the proper aim of treaty interpretation is finding these intentions. Objectivists, on the other hand, tend to see a treaty as a given ideal, transcending its makers’
wishes. They wish to discover either the meaning of the text or of the aims and goals as something ‘objectively given’.
Why would a treaty need to be subjected to its creators; why would it need to be viewed in a subjectivist manner? States may make and unmake treaties, but they may do so only because there is a legal order authorising them to create law. In principle, states are not in a privileged position vis-à-vis the law of treaty- creation (meta-treaty law), no more than a national judge is as against his
211 Kelsen (1950) supra note 52 at xiv–xv.
212 Bernhardt (1963) supra note 4 at 132; Hexner (1964) supra note 55 at 123; Engel (1965) supra note 205 at 109.
213 Contra: Karl (1983) supra note 125 at 38.
214 Bos (1980) supra note 3 at 152; Fitzmaurice (1958b) supra note 10 at 212; Georg Schwarzenberger, Myths and realities of treaty interpretation. Articles 27–29 of the Vienna Draft Convention on the Law of Treaties, 22 Current Legal Problems (1969) 205–227 at 213; contra: Ress (2002) supra note 55 at 23 (MN 19). Kelsen himself seems to argue obiter that the meaning of a norm can change over time (Kelsen (1979) supra note 47 at 151 (Ch 50)). However, as he makes clear in the accompanying endnote, only a theory which asserts ‘that the meaning of an interpreted norm is constant as long as it is not changed explicitly in appropriate action by the normgiving authority’ ( Jerzy Wróblewski, Semantic basis of the theory of legal interpretation, 6 (N.S.) Logique et Analyse (1963) 397–416 at 415), only such a ‘static’ theory can be proposed from a positivistic point of view; the ‘dynamic’ theory is a fiction. Kelsen (1979) supra note 47 at 303 (N 128).
215 Karl (1983) supra note 125 at 25, 148.
Uncertainty in International Law 128
country’s code of penal procedure. States may have a much greater leeway in creating norms than the judge, but both create valid norms only because their respective legal order authorises them to. There is no natural or inherent power of states to create international treaty law. The only true sovereign in international law is international law itself; no state is legibus solutus vis-à-vis international law. Treaties ought to be observed whatever the actions of their makers – unless the legal conditions for a change of law are fulfilled. On the other hand, it is the act of will of the states parties that creates positive treaty law. As argued in Section 4.2.2, however, only the fact, not the content, of that will is relevant: a treaty’s content is determined by the language of the treaty.
These two dichotomous choices make four differing combinations of approaches to treaty interpretation possible. Because this section is concerned with the role of subsequent practice in interpretation, we will test them against the doctrine of subsequent practice to see what role it would be able to play in each.
(1) The static/subjective combination was traditional doctrine’s approach and as such static (‘original’), subjective (‘party’) and directed towards discovering mean- ings attributed rather than expressed (‘intent’). For this view, the parties’ subsequent practice gives an indication as to what they understood the treaty to mean when they concluded it.216 (2) The dynamic/subjective choice reflects the parties’
current views. While it is still the treaty-makers whose intentions are reflected in their practice or in the practice of organs they have created, shifts in their under- standing of the terms become relevant. This approach is justified as (current) consensus among states parties,217 which appears before the Court as a variant of what can be called the ‘mental economy’218 argument – the Court declines to discuss an issue, because the parties to the dispute before it agree on a notion
216 Bernhardt (1963) supra note 4 at 131 (referring to prior international jurisprudence); Fitzmaurice (1952) supra note 9 at 20; Fitzmaurice (1958b) supra note 10 at 212; Haraszti (1973) supra note 8 at 143; McNair (1961) supra note 12 at 424; Pollux (1946) supra note 31 at 78; probably also Waldock (1965) supra note 32 at 59 (para 24). Also, the Permanent Court’s and the present Court’s jurisprudence seems to support this approach, e.g.: Treaty of Lausanne (1925) supra note 60 at 24;
Jurisdiction of the Courts of Danzig, Advisory Opinion of 3 March 1928, PCIJ Series B No. 15 (1928) 18; Case Concerning the Payment of Various Serbian Loans Issued in France, Judgment of 12 July 1929, PCIJ Series A No. 20 (1929) 38; Case Concerning the Payment in Gold of the Brazilian Federal Loans Issued in France, Judgment of 12 July 1929, PCIJ Series A No. 20 (1929) 119; Corfu Channel (United Kingdom v. Albania), Merits, Judgment of 9 April 1949, ICJ Reports (1949) 4 at 25; Constitution of the Maritime Safety Committee of the Inter-Governmental Maritime Consultative Organization, Advisory Opinion of 8 June 1960, ICJ Reports (1960) 150 at 167; Case concerning the Temple of Preah Vihear (Cambodia v.
Thailand), Merits, Judgment of 15 June 1962, ICJ Reports (1962) 6 at 33, 35. For a detailed analysis Karl (1983) supra note 125 at 127–135.
217 Karl (1983) supra note 125 at 144–156; Ress (1987) supra note 199 at 57; Ress (2002) supra note 55 at 27 (MN 27). This argument will return as a construct to justify change in Section 4.4.2.
218 Jửrg Kammerhofer, Oil’s Well That Ends Well? Critical comments on the merits judgment in the Oil Platforms case, 17 Leiden Journal of International Law (2004) 695–718 at 707–708.
4.4.1 Interpretation and modification 129
anyway.219 Also, estoppel can be argued as being capable of determining the interpretation by subsequent practice as a matter of a state’s current opinions.
(3) On the dynamic/objective combination subsequent practice becomes impor- tant because it is said to put the treaty in touch with reality (Vertragswirklichkeit).
For this approach it seems imperative that the norm must not diverge too much from reality and must remain effective to remain law. But law, even international law, is something categorically different from reality, an ideal idea specifying an Ought, no matter what actually happens. For Wolfram Karl, an objective approach is always dynamic; he does not believe that an objectively given norm would not sway with the vagaries of treaty application. (4) The static/objective approach seems to coincide well with the Pure Theory’s ideas. On this approach, a treaty norm is a valid norm and thus beyond change through subjective percep- tion. Being an objective ‘thereness’, the treaty text is ossified at the moment of its conclusion or formal amendment. Thus a treaty equals Baxter’s photograph: it is fixed for eternity until it is burnt, torn or taken anew.
The relevance of subsequent practice in a static/objective approach is this:
interpretation is about finding the meaning(s) of a treaty, even when we use subsequent practice. It, in turn, can only be a method to help discover the meaning. Practice can at best be evidence of the meaning.220 The connection between practice and meaning is coincidendental, not necessary. Wolfram Karl’s warning against using subsequent practice merely as a supplementary means of interpretation, because that would further weaken practice’s importance,221 could be seen as a petitio principii: the importance of subsequent practice for interpretation has yet to be proven. To see an approach as useless because it would minimise practice’s value presupposes what has to be proven. It must be established that subsequent practice is capable of cognising the norms of international treaty law correctly. It is doubtful that this is the case, because the norms already are the text and practice cannot add to or remove from the text. However, practice could be argued to be capable of determining the possible meanings of the norm.
However, like other factors of interpretation it has no necessary or automatic claim to epistemic value, much less of discovering the ‘normative truth’. On the contrary, practice as application of norms by its subjects can be observance just as likely as it can be breach of a given norm. Who is to say that the subjects’ actions, even if they are the law-makers, will always, sometimes or at any time reflect the prescription? Take the private behaviour of parliamentarians in their function as
219 Access to, or Anchorage in, the Port of Danzig, of Polish War Vessels, Advisory Opinion of 11 December 1931, PCIJ Series A/B No. 43 (1931) 140. The Court does not make this a rule, however.
In Nicaragua it was satisfied that both parties were in agreement as to the law on the use of force, but decided to ascertain the law for itself: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States of America), Merits, Judgment of 27 June 1986, ICJ Reports (1986) 14 at 99 (para 188).
220 Fitzmaurice (1952) supra note 9 at 21; Fitzmaurice (1958b) supra note 10 at 224; McNair (1961) supra note 12 at 424.
221 Karl (1983) supra note 125 at 126, 138.
Uncertainty in International Law 130