4 Interpretation and modi fi cation
5.1 A preliminary definition of conflict of norms
5.2.2 Lex specialis as more e ff ective or re fl ective of
The most common justifications given in international legal writings in favour of preferring the lex specialis – higher effectiveness and stronger ‘state consent’ – both suffer from the same legal theoretical problem:49
[T]the following two reasons for letting a more specific norm prevail over a more general norm can be given: (i) the special norm is the more effective or precise norm, allowing for fewer exceptions . . . and; (ii) because of this, the special norm reflects most closely, precisely and/or strongly the consent or expression of will of the states in question.50 Each of the elements will be dealt with in turn. A norm with a narrower scope of validity is felt to be more effective than one with a more general scope and thus it is said to ‘prevail’:51 it can be doubted whether a narrower norm is necessarily more effective than a wider norm. Effectiveness is a matter for empirical research and
47 Pauwelyn (2003) supra note 4 at 388, 126.
48 Walter (1955) supra note 36 at 90.
49 Jửrg Kammerhofer, Uncertainty in the formal sources of international law: Customary inter- national law and some of its problems, 15 European Journal of International Law (2004) 523–553 at 546.
50 Pauwelyn (2003) supra note 4 at 387.
51 In the same vein: Jenks (1954) supra note 30 at 446; Koskenniemi (2006) supra note 5 at 36 (para 60).
Uncertainty in International Law 152
there may well be constellations of norms where the relationship is different. Even if empirical research showed that special norms were always more effective, this fact could not constitute a reason for the more general norm to lose its validity or be changed by the special norm, i.e. for the lex specialis maxim. Effectiveness is not validity; the effect (or lack thereof) of a prescription in reality cannot have an influence on the existence of that prescription. This is a necessary precondition for the very notion of ‘norms’ – norms have an ideal existence: No Ought from Is, no Is from Ought alone.52 We will return to the necessity of the Is–Ought dichotomy for legal scholarship in Section 7.2, but we can already see here that the conclusion drawn from effectiveness to the loss of validity transcends the dichotomy. Tran- scending the dichotomy means that we can no longer see norms as prescription.
Prescription, however, is the only way the concept of ‘norm’ makes sense.
Using state intentions as argument fails for the same reason. For Pauwelyn, ‘the principle of lex specialis is . . . grounded in the idea that the “most closest, detailed, precise or strongest expression of state consent” . . . ought to prevail’.53 The traditional resolving devices are merely ‘practical methods in the search for the
“current expression of state consent” ’.54 The underlying assumption is that ‘state consent’ is relevant in resolving the dispute. A number of arguments can be advanced against this contention. First, as positive norms of international law both norms are an expression of state consent. The idea that one kind of consent is necessarily better than another would be problematic even in the traditional approaches to international law. If we were to assume state consent as the a priori source and end of all international law, how are we to choose between the differ- ent expressions of consent? From the assumed relevance of consent it is difficult to see how one could deduce a ‘special’ consent to be more valuable than a ‘general’
consent.
Second, the argument from state consent as the ultimate authority merely begs the question, a question that should be asked when dealing with the relative value of the sources of international law. Can state consent be the ultimate source of all international law? It cannot be, because international law’s sovereign is not the states, but international law itself. The focus upon the subjects of law – states only being constituted as subjects by the law in the first place – is an expression of the denial of the duality of Is and Ought. On a legal view, states are authorised to create law only because a meta-law on law-making authorises them. If we assume that states need no authorisation, then a norm (e.g. a treaty) would be created out of fact alone. The reference to the brute fact of state consent must fail, because brute fact neither is nor makes law. If state consent were a source of international law, then the question would assume a different dimension, but then consent would be a norm. If the more special state consent were superior to the more
52 Kelsen (1960) supra note 2 at 5 (Ch 4 b); Kelsen (1979) supra note 8 at 44 (Ch 16).
53 Pauwelyn (2003) supra note 4 at 388.
54 Pauwelyn (2003) supra note 4 at 388; Karl (1983) supra note 38 at 107–108; Koskenniemi (2006) supra note 5 at 37 (para 60).
5.2.2 Conflict of norms 153
general variant, then this also would need to be a norm in order to influence law- making and derogation.
5.2.3 Lex specialis as positive norm
Is the lex specialis maxim a norm of positive international law? One of the least controversial applications of the lex specialis maxim is the treaty-based inter se abrogation from a norm of customary international law.55 The Court observed in North Sea Continental Shelf:
Without attempting to enter into, still less pronounce upon any question of jus cogens, it is well understood that, in practice, rules of international law can, by agreement, be derogated from in particular cases, or as between particular parties . . .56
The connection to the lex specialis maxim was made in Nicaragua:
In general, treaty rules being lex specialis, it would not be appropriate that a State should bring a claim based on a customary-law rule if it has by treaty already provided means for settlement of such a claim.57
This is customary international law’s ius dispositivum character; a feature known from municipal legal systems, where certain civil codes will provide default regula- tion if the parties do not provide different rules in a contract. The VCLT also includes several such ius dispositivum provisions, formulated along the lines of
‘unless the treaty otherwise provides’ (e.g. Article 20(5) VCLT). Hence, the alleged ius dispositivum character of customary international law turns out to be an expres- sion of the claim that the lex specialis maxim is a norm of customary international law. A norm is never dispositive by default and this feature needs to be included in positive regulation. In contrast, all norms claim observance without respect for the inter se ‘arrangements’ by default. In the original Roman law sense they are all by definition ius cogens. This argument will be further substantiated below (e.g. Section 5.5.1). Derogation (even if only by an inter se agreement) needs a basis in norms, for only norms can influence the validity of norms. Both the North Sea Continental Shelf and the Nicaragua judgments also contain the alternative view:
customary international law and international treaty law exist side by side and do not derogate from each other.58 Assuming that the maxim is a norm of customary international law, several problems arise.
55 Koskenniemi (2006) supra note 5 at 39–40 (paras 66–67), 23–24 (paras 52–54); Pauwelyn (2003) supra note 4 at 155–157, 212–236, 391–392; Verdross and Simma (1984) supra note 3 at 414.
56 North Sea Continental Shelf (Federal Republic of Germany v. Denmark; Federal Republic of Germany v. Netherlands), Judgment, ICJ Reports (1969) 3 at 42 (para 73).
57 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 137 (para 274).
58 North Sea Continental Shelf (1969) supra note 56 at 38–39 (paras 61–65) Nicaragua (1986) supra note 57 at 93–96 (paras 174–179).
Uncertainty in International Law 154
(1) It is doubtful that customary law is capable of ‘referring’ to other norms at all (Section 3.2.5).59 Because customary law is based on behavioural regularities (customs), customary law can only have such content which can be reflected as behavioural pattern; these patterns are required to form state practice. This ‘real- world’ behaviour, e.g. the passage of a ship through straits, or the signing of a piece of paper cannot refer to the ideal or normative content of such action. The specific ideal significance is not part of the behavioural pattern, hence is not part of state practice and thus cannot form part of the content of a customary norm.
A norm with the content ‘earlier norms are repealed, if . . .’ cannot emerge, because the state practice that would be necessary for its formation – a ‘behaviour’
that shows the loss of validity of a norm – cannot exist.
(2) Even if we discount the doubts regarding the capability of customary regu- lation and assume the lex specialis maxim to be a norm of customary international law, the problems persist. Such a move does not solve the problem that we are once again confronted with a conflict of norms. One norm claims observance and another norm of the same type, normative order and even ‘rank’, claims to derogate from it or from any norm with certain characteristics. The relationship of norm to human behaviour is indirect due to the Is–Ought divide: norms can only postulate a claim. The norm–norm relationship, however, is direct; either a norm which claims to derogate does or it does not; there is no prescription–reality disconnect.60 A widely accepted example from traditional international legal doc- trine has a ius cogens norm simply end the validity of a contrary treaty (cf. Section 5.5.1). On the other hand, if a human were to create a derogating norm that claimed to derogate from a judgment issued against himself, nobody would argue that the judgment ‘actually’ loses its validity.61 Norms can influence the validity of other norms; mere behaviour cannot. The prohibition against murder is not dependent upon whether murders actually occur.
The question thus becomes which constellations cause norms to derogate and which fail in doing so. This is the key question and the reason why a definition of the notion of ‘norm conflict’ is not really important at all. Derogation of norms is not dependent upon conflict and conflict does not determine whether one norm derogates from another. This problem will not be discussed just yet; we will men- tion here that the hierarchy of norms and the unity of normative orders are key notions in this determination, because norms both lose and also derive their validity from other norms (Section 5.5).
One can argue at ths point that it is only the content of certain norms in purporting to regulate conflict that gives them special powers to end the validity of another norm. Such an argument is not likely to succeed either, because the specific function of norms lies in their form (the Ought, the claim to be observed), not in their content. Thus form equals existence equals validity and norms
59 See also Thienel (1988) supra note 42 at 24–25.
60 Kelsen (1979) supra note 8 at 86 (Ch 27 III), 168 (Ch 57 IV).
61 Merkl (1918b) supra note 42 at 1125.
5.2.3 Conflict of norms 155
differing by content would still share the same existence as validity. Their validities might clash, but that clash is not decided by their content. Indeed, this may toll the death knell for the very idea of a conflict resolution technique based on speciality.
(3) We are also faced with the problem of the inter-source relationships when a customary international lex specialis maxim claims to derogate from an inter- national treaty norm. An example might be where the lex generalis is itself a treaty, as is the case in the relationship between the International Covenant on Civil and Political Rights 1966 and the European Convention on Human Rights 1950. The problem is similar to the one described above, but with the complicating factor of the inter-relationship of two different formal sources of international law. While it seems unwise to bow to majority opinion and accept that there is no hierarchy between the sources of international law,62 it is also difficult to accept a subordin- ation.63 Treaties cannot be derived from custom (pace Kelsen)64 because customary international law cannot contain the authorisation to create international treaty law (pacta sunt servanda). Custom could theoretically be derived from treaty, but that would only establish subordination to one specific treaty. Also, the only candidate for such an authorisation, Article 38(1)(b) of the ICJ Statute, by its own words does not found the validity of a formal source of international law. It merely states what the applicable law before the Court is. The ‘default solution’ is that the two sources are not normatively connected and are in a similar position as if a single individual purports to derogate from a criminal judgment. They belong to two unconnected normative orders, hence any derogation seems a priori excluded (Sections 6.2.1 and 6.3.2).65
(4) What if the lex specialis maxim is a norm of international treaty law or a ‘general principle of law recognised by civilised nations’ (Article 38(1)(c) ICJ Statute)? Both would result in the same inter-source problematique and the same questions of its relationship to norms of the same source would arise. In addition, each treaty is its own normative island;66 the principle of pacta tertiis nec nocent nec
62 E.g: Michael Akehurst, The hierarchy of the sources of international law, 47 British Year Book of International Law 1974–1975 (1977) 273–285 at 275; Karl (1983) supra note 38 at 86;
Koskenniemi (2006) supra note 5 at 47 (para 85); Pauwelyn (2003) supra note 4 at 94; Verdross and Simma (1984) supra note 3 at 322.
63 Kammerhofer (2004b) supra note 49 at 548–550.
64 Hans Kelsen, Principles of international law (1952) 314.
65 The unconnectedness means that there is no influence upon the other’s norms’ validity, because the claims to be observed are a priori equal.
66 Hans Kelsen, Reichsgesetz und Landesgesetz nach ửsterreichischer Verfassung, 32 Archiv des ửffentlichen Rechts (1914) 202–245, 390–438 at 209: ‘Like the corporeal world of Is, also the intellectual world of Ought has its own law of impenetrability. When I argue that a norm of one authority is repealed by a norm of another authority, I merely negate the first authority [as authority] and put the second authority in its place’ ‘Wie die kửrperliche Welt des Seins hat auch die geistige Welt des Sollens ihr Gesetz der Undurchdringlichkeit. Wenn ich die Norm der einen Autoritọt durch die Norm einer anderen Autoritọt fỹr aufgehoben ansehe, bedeutet das einfach, daò ich die erstere Autoritọt negiere und die zweite an ihre Stelle setze.’
Uncertainty in International Law 156
prosunt merely expresses that each treaty is based immediately on the Grundnorm
‘pacta sunt servanda’ and its relationship to other treaties is different from that between norms of the same treaty.67 However, this necessary connection obtains only in the relationship of treaty to treaty. Norms may very well have more subjects (its scope ratione personae) than law-makers, e.g. domestic statutes. Inter- national treaties’ contractual nature and their being part of international law does not change that theoretical possibility – it depends on positive law. Despite the interjection of ‘sovereignty!’ and despite a naturalistic preconception of the nature of the contract, the third party rule is not an absolute necessity and needs to be laid down in positive law, i.e. we need the meta-law on inter- national treaty law-creation to specify that their product may only have a certain personal scope.68
If the lex specialis maxim is a ‘general principle’,69 one additionally has to con- tend with grave theoretical doubts as to the very possibility of this source as positive international law. How can a comparison of scientific abstractions from diverse legal systems in any shape be willed as part of international law? Verdross has shown that general principles of law are essentially a natural law-construction and are not positive law at all.70 Their ‘positivisation’ in Article 38(1)(c) ICJ Statute is relevant only for the Court, adds Kelsen: ‘[The general principles of law] are norms, which become international law applicable before the International Court of Justice, [only] because Article 38(1)(c) authorises the International Court of Justice to apply them.’71
5.3 Lex posterior legi priori derogat
The key question of this section goes beyond the topic of norm-conflict as trad- itionally conceived. Can norms change over time? The very possibility of change in normative orders (at least over time) is at stake here and the lex posterior maxim is the most popular72 answer to that question. While it seems obvious how the problem of change is a problem of norm-conflict, what is being argued is the reverse, namely that change in law presupposes the lex posterior maxim. ‘As with any law, [international law] may change over time. . . . As a result, any later norm can, in principle, overrule an earlier one (lex posterior derogat legi priori).’73 This
67 See Section 5.3.2.
68 For a similar argument: Robert Kolb, The formal source of ius cogens in public international law, 53 Zeitschrift fur ửffentliches Recht (1998) 69–104 at 82–83.
69 Czaplin´ski and Danilenkow (1990) supra note 10 at 21.
70 Alfred Verdross, Vửlkerrecht (3rd ed. 1955) 11–12, 22–23.
71 ‘Es [handelt] sich [bei den allgemeinen Rechtsgrundsọtzen] um Normen . . ., die dadurch von dem Internationalen Gerichtshof anzuwendendes Vửlkerrecht werden, daò Art. 38 Đ 1c den Internationalen Gerichtshof ermọchtigt, sie anzuwenden.’ Kelsen (1979) supra note 8 at 99 (Ch 28).
72 Czaplin´ski and Danilenkow (1990) supra note 10 at 19.
73 Pauwelyn (2003) supra note 4 at 14.
5.3 Conflict of norms 157
confusion of the concepts of change, derogation and of the role of maxims74 leads to the apocryphal status of the lex posterior maxim in international law, and to the ‘argument by necessity’ for its existence, in legal theory. This section will accordingly focus on the argument that the lex posterior maxim is a rule of logic and that its existence is necessary by virtue of the very existence of ‘prescriptive sentences’.75
A typical argument in favour of the lex posterior maxim might look something like this, with ‘A’ taking the role of proponent, while ‘B’ plays the sceptic: A: Law can change. B: Why should it? A: Because the law-makers can create successive norms N1 at t1, N2 at t2. B: But why would the law change just because there is a multiplicity of norms which are successively created? A: Because the later modifies the former. B: Why should the later norm necessarily be ‘better’ than the former?
A: Because the two norms contradict each other, the former becomes ‘false’ in accordance with the principle of the unity of the legal order and is deleted due to principle of excluded contradiction. B: Norms are not true or false statements.
Both norms were created and none was not created. What makes the original norm-creation invalid? A: Because the current will of the law-maker trumps his prior will. B: How so? A: Because the current will derogates from the older will, we can infer from the law-maker’s different and later norm-creation that he wills a derogation of prior conflicting norms. B: It is begging the question to say that a later norm derogates a former norm because it is later. In order to do so we have to presuppose derogating power to later norms and where should that presupposition come from?
As in Section 5.2 above, there are three avenues of justification:
(a) lex posterior as a rule of logic;
(b) lex posterior as an expression of effectiveness or intentions (which will not be treated again);
(c) lex posterior as a positive norm.
In this section we will focus on the first justification, because of its prevalence in scholarly discussion.
5.3.1 Lex posterior as a rule of logic
The claim that derogation and change – whether by way of the ‘primacy’ of the later norm or not – is connected to formal logic is highly complicated. Inter- national legal doctrine does not explain this in detail and in legal theory only
74 Karl (1983) supra note 38 at 59.
75 Amedeo G. Conte, Hans Kelsen’s deontics, in: Stanley L. Paulson, Bonnie Litschewski Paulson (eds), Normativity and norms. Critical perspectives on Kelsenian themes (1998) 331–341 at 331.
From this terminology alone one can see why the reasoning by logic is unable to prove the point at issue and why such approaches transcend the Is–Ought dichotomy.
Uncertainty in International Law 158