What are rules of interpretation?

4 Interpretation and modi fi cation

4.1 Treaty interpretation – the conventional debate

4.1.2 What are rules of interpretation?

Another classical debate, which has all but stopped since the signing of the Vienna Convention and the persistent invocation of Articles 31–33 VCLT by the International Court of Justice,28 is the precise status of the ‘rules’ of treaty interpretation. The question is primarily whether any of these ‘rules’, ‘principles’

or ‘maxims’ are norms of international law. Despite the confusion sown by imply- ing that there is a choice between ‘obligatory’ rules and other rules29 – obligatory norm (or rule) being an analytical term30 – it is possible to portray and discuss the writings on the matter with some degree of clarity.

Before the adoption of the Vienna Convention there seemed to be unwillingness, especially among scholars from the common law tradition, to see interpretation as guided by rules. Arnold McNair typifies this approach: ‘The many maxims and phrases which have crystallised out and abound in the text-books and elsewhere are merely prima facie guides to the intention of the parties.’31 However, it is difficult

25 Lauterpacht (1949) supra note 8 at 82.

26 Tammelo (1967) supra note 21 at 5–6.

27 Lauterpacht (1949) supra note 8 at 75.

28 Santiago Torres Bernárdez, Interpretation of treaties by the International Court of Justice following the adoption of the 1969 Vienna Convention on the Law of Treaties, in: Gerhard Hafner et al.

(eds), Liber amicorum Professor Ignaz Seidl-Hohenveldern – in honour of his 80th birthday (1998) 721–748.

29 McDougal (1967) supra note 21 at 992.

30 For Immanuel Kant, analytical terms are such where the subject contains the predicate. Immanuel Kant, Kritik der reinen Vernunft (1781, 1787) A 6–7, B 10. There cannot be a non-obligatory norm, because ‘norm’ itself is defined as obligatory. Hence, the term ‘non-obligatory norm’ is a contradictio in adiecto.

31 McNair (1961) supra note 12 at 366 (emphasis added). Also denying rule status, but for widely differing reasons, e.g.: Bos (1980) supra note 3 at 21–22; Kửck (1976) supra note 10 at 80; Pollux, The interpretation of the Charter, 23 British Year Book of International Law (1946) 54–82 at 66 (There are some indications that the author of that article, using the pseudonym ‘Pollux’, is Edvard Hambro); Tammelo (1967) supra note 21 at 48.

Uncertainty in International Law 92

to ascertain what precise position some hold, because the differentiation between phenomena differs for each scholar. The ILC’s last Rapporteur on the law of treaties, Humphrey Waldock, for example, distinguishes between ‘principles and maxims’ and ‘methods’ of treaty interpretation. Phrases such as ut res magis valeat quam pereat or argumentum a contrario are held to belong to the former group: ‘[t]hey are for the most part principles of logic and good sense valuable only as guides to assist in appreciating the meaning’.32 The textual, subjective and teleological approaches are examples of methods. Waldock does not assert ‘that there is no obligatory rule in regard to methods of interpretation’,33 but he does not claim the opposite either. Some statements imply that there are some legal rules of interpretation: ‘Accordingly, the choice before the Commission is believed to be either . . . or to seek to isolate and to codify the comparatively few rules which appear to constitute the strictly legal basis of the interpretation of treaties.’34 There is some difficulty in extracting a clear position from a text containing multiple qualifiers.

Let us assume for a moment that the rules of interpretation are norms of international law. A norm of international law must have been created in accord- ance with its meta-law on law-creation – it must belong to one of the ‘sources’ of international law. Therefore, a ‘stock-taking of the sources of international law’35 is required. Taking Rudolf Bernhardt’s analysis as a model, we will look at the sources enumerated in Article 38(1)(a)–(c) of the Statute of the International Court of Justice (ICJ Statute) in turn to see whether rules of interpretation are and can be treaty or customary international law or ‘general principles of law’.36

(1) There is treaty law on the matter in force today. The Vienna Convention on the Law of Treaties has three articles devoted to treaty interpretation, with the title of Article 31 expressly proclaiming the general rule of interpretation. Within its confines it is law. However, these confines are rather narrow: first, treaty law is only valid as treaty law inter partes;37 second, the Vienna Convention applies only to treaties concluded after 27 January 1980;38 third, its provisions apply only to parties to the Vienna Convention. This puts the UN Charter outside its direct applicability. Nevertheless, there are international treaty law norms concerning treaty interpretation in force today.

The claim by Heribert Kửck of the irrelevancy of conventional norms thus may mean that he seeks to superimpose upon treaty norms another ideal; that of

32 C.H.M. Waldock, Third report on the law of treaties [A.CN.4/167, A.CN.4/167/Add.1–3], 16 Yearbook of the International Law Commission 1964 (1965) Volume II, 1–65 at 54 (para 6).

33 Waldock (1965) supra note 32 at 54 (para 7).

34 Waldock (1965) supra note 32 at 54 (para 8) (emphasis added).

35 ‘Bestandsaufnahme der Vửlkerrechtsquellen’; Bernhardt (1963) supra note 4 at 28.

36 This is without prejudice to the question of which sources of international law there are and what role Article 38 ICJ Statute has to play (Chapter 6).

37 For the ICJ’s claim that the VCLT’s norms on treaty interpretation are customary international law see Torres (1998) supra note 28 and infra.

38 Article 4 VCLT states that it is not retroactive, applicable only to treaties concluded after its own entry into force.

4.1.2 Interpretation and modification 93

(general) hermeneutics. He asks whether these norms can adequately portray cognition:39

The rules of interpretation laid down in the VCLT are ‘hermeneutic rules’. . . . Because one cannot lay down a canon of hermeneutic interpretation as norms, there is no possibility of judging the correctness and completeness of these ‘rules’ in abstracto. Only in connection with a concrete case can it be proven whether the ‘rules of interpretation’ of the VCLT are adequately describing the interpretative and cognitive act in this case.40

If, therefore, in a concrete case the rules of the Vienna Convention were found not to be effective in adequately describing the interpretative act, their value as norms would be null – voided by the claimed superior hermeneutic ideal of how textual cognition actually does work; positive law is declared void by reason of theoretical predispositions.41 Such a strong claim transcends the categorical duality of Is and Ought and puts in place a fictional norm that does not belong to the legal system ‘international law’.

Kửck is right in drawing our attention to an important distinction. One can understand the function of these rules or maxims either as norms imposed on how a text ought to be understood, or as ‘laws of nature’, as a description of the process of understanding, abstracted from those cases where the text was correctly understood. Whereas the latter are a mere guide (‘If you follow these rules, you will get the correct meaning’), the former are norms properly speaking (claiming to be observed) whatever method may work best. This challenge, while it impinges upon the question at hand is a matter of general hermeneutics, a science devoted to describing cognition of texts, and will be discussed in Section 4.3.3.

Codifications such as the Vienna Convention have a galvanising effect.

Academic uncertainty diminishes, the norms become a dogma transcending the scope of the codification instrument, critique is stifled – all because of the psycho- logical impact of a written text. Strictly speaking, however, Articles 31–33 VCLT are legal norms of interpretation42 – they make their content obligatory, whatever the merit of the method(s) they contain. In a sense, therefore, the rules of ‘interpret-

39 Kửck (1976) supra note 10 at 70, 80. See infra.

40 ‘Die insoweit in der WVK gegebenen Interpretationsregeln sind hermeneutische “Regeln”. . . . Da sich kein hermeneutischer Interpretationskanon normieren lọòt, fehlt es auch an der Mửglichkeit, in abstracto ỹber die Richtigkeit und Vollstọndigkeit dieser “Regeln” zu urteilen. Nur im Zusammenhang mit einem konkreten Fall kann sich erweisen, ob die “Interpretationsregeln”

der WVK jeweils den Auslegungs- und Verstehensvorgang adọquat beschreiben.’ Kửck (1976) supra note 10 at 91 (second emphasis added).

41 Such a move is akin to the argument that the law that does not fulfil certain moral criteria is non- law; for a critique of such arguments cf.: Adolf Julius Merkl, Zum Interpretationsproblem, 42 Grỹnhutsche Zeitschrift fỹr das Privat- und ệffentliche Recht der Gegenwart (1916) 535–556, reprinted in: Hans Klecatsky, René Marcic, Herbert Schambeck (eds), Die Wiener rechtstheo- retische Schule. Ausgewọhlte Schriften von Hans Kelsen, Adolf Julius Merkl, Alfred Verdross (1968) 1059–1077 at 1076.

42 Bos (1980) supra note 3 at 21.

Uncertainty in International Law 94

ation’ are not interpretation as cognition of law, but are an addition to the set of norms to be interpreted:

‘New legal rules are often disguised as rules of interpretation.’ These new legal rules are precisely these statutory rules of interpretation. They can exclude a meaning otherwise possible; they can declare a merely possible meaning to be the prescribed meaning.43 Statutory rules of interpretation (like Articles 31–33 VCLT) – because they are norms themselves – modify the law by superimposing themselves upon all norms to which they are applicable. A treaty provision subject to the Vienna Convention has to be read ‘through the lens’ of Articles 31–33 and thus these rules change the treaty provision. Conversely, norms of interpretation need to be interpreted as well. Article 31 is interpreted together with the substantive treaty norms to which it is applied. To ‘stress their essential unhelpfulness’44 cannot hurt the provisions of the Vienna Convention; norms do not depend for their validity on their

‘usefulness’.

(2) The second possibility is that rules of interpretation are customary inter- national law norms.45 Two problems may arise in this context. The ‘practice’

considered by scholarship is not state practice, but mostly the pronouncements of international tribunals. If the ICJ’s practice is capable of forming the behavioural regularity that is the objective element of customary law, the pronouncements are statements of how the tribunal concerned thinks the law is shaped. Admitting such ‘opinio iuris’ may be difficult as a matter of positive law: states are the makers of customary international law, and it is their will expressed in the subjective element that counts. To argue that the repetitive judicial pronouncements make customary international law, ‘unless one denies a law-making function to judicial practice in flagrant contradiction to actual developments’,46 is to commingle law and fact and to assume in the last consequence that whatever happens is law. If courts make pronouncements, do they become law merely by being made and without a superior norm making these pronouncements the building block of new law?

No; that would mean that whatever happens ought to happen.

The second problem, much more formidable, is that customary international law is dependent upon customs, accumulated factual behaviour as a pattern in reality (Section 3.2.5). Rules of interpretation cannot be grounded in a factual pattern, because they refer to a non-factual cognitive process. State practice can- not be accumulated from an interpretation of ‘ordniary meaning’, because – as

43 ‘ “In der Form von Auslegungsregeln verbergen sich sehr oft neue Rechtsvorschriften.” Diese neuen Rechtsvorschriften sind eben die gesetzlichen Auslegungsregeln. Einen – sonst mửglichen – Sinn des Gesetzes sind sie in der Lage auszuschlieòen, eine nur mửgliche Bedeutung zur gebotenen zu erheben.’ Merkl (1916) supra note 41 at 1076.

44 Lauterpacht (1949) supra note 8 at 51; McNair (1961) supra note 12 at 366.

45 Bernhardt (1963) supra note 4 at 29; Haraszti (1973) supra note 8 at 212.

46 ‘es sei denn, man bestreitet der Gerichtspraxis in offensichtlichem Widerspruch zu der tatsọchlichen Entwicklung die rechtsbildende Kraft’; Bernhardt (1963) supra note 4 at 30 (emphasis added).

4.1.2 Interpretation and modification 95

pointed out in the previous chapter – the factually observable events in the process of interpretation are not interpretation itself and we can only describe these extraneous events. In order to be able to utilise the facts as practice, we would have to employ an interpretive scheme for reality. This we cannot do, since customary law-making as primitive method allows only the usage of the facts and cannot pierce the factual veil. Just as a customary international law norm cannot be a meta-law of law-creation, because the law-creation is not real behaviour, so an ideal, customary law cannot regulate interpretation, because it too is divorced from observable reality.

(3) One could also discuss the validity of rules of interpretation as ‘general principles of law’. Surely, the ideal source from which to draw rules of interpret- ation would be a comparison of municipal legal orders, which frequently contain statutory rules of interpretation? The general question is whether the fact that the municipal law of many states contains norms with similar content is enough to create a norm of positive international law. What is missing to make it positive is an act of will. While the municipal norms all have their own act of will,47 this does not encompass the will to create international law. Comparative legal scholarship is not international law-making, despite Article 38(1)(c) of the Statute. In this particular case, the rules applicable to international law do not, by their nature, appear in the municipal setting, where the rules are far too diverse anyway.48

So, if, apart from the rules laid down in Articles 31–33 VCLT, ‘rules’ of inter- pretation are not norms after all, what are they? The most promising avenue is that they are a description of the cognitive-hermeneutic processes when humans read texts. We might be able to distil, to induce, a manual of right cognition, where following the rules would ensure that the text be understood, while otherwise the text would not be cognised correctly. But we will discuss this further below, in Section 4.3.3.