4 Interpretation and modi fi cation
5.5.3 Types of con fl ict between lex superior and lex inferior
This seems to be a particularly muddled issue, but it is an immensely important issue for all legal systems and for international law in particular. We must deter- mine under which conditions a norm designated as ‘higher’ actually derogates from a lower norm, otherwise uncertainty will increase to the point that the very existence of a normative order is endangered. Three categories of lex superior–lex inferior conflict situations will be presented and discussed. However, these problems might be systemic and a matter of the very nature of positive normative orders.
We may at least be able to clarify how the structure of international law is made uncertain by analysing this type of norm-conflict.
Three types of conflict will be discussed below:
(1) A substantive norm of a normative order may conflict with a higher norm of the same order. It is a norm-conflict of observability (the first type defined in Section 5.1). A norm of a statute prescribes Op, while some constitutional norm prescribes Oơp.
(2) That norm may conflict with a ‘higher’ norm of the same order expressly derogating from it, which is a conflict of conflicting ideals (the second type in Section 5.1). A constitutional law prescribes ‘the statute Op is repealed herewith’, ơOp.
(3) Despite appearances the most complex case is where a norm does not fulfil all the conditions of the meta-law on law-making (Erzeugungregel), which was not mentioned above. A statute, for example, is not passed with the correct major- ity present and voting. This is not only the most complex case, but potentially also the most destructive.
5.5.3.1 Which is the ‘higher’ norm?
In this case, the problem is defining which norm is ‘higher’ and which is lower.
This problem can be exemplified in the relationship between normative orders:
213 ‘Auf diese Relation zwischen derogierender und derogierbarer Norm kommt es aber im Stufenbau nach der derogatorischen Kraft an.’ ệhlinger (1975) supra note 181 at 23.
214 ệhlinger (1975) supra note 181 at 25–26.
215 Merkl (1923) supra note 77 at 299.
216 ệhlinger (1975) supra note 181 at 26.
5.5.3 Conflict of norms 185
we cannot simply believe a norm’s claims to be superior to another – even if that claim does not amount to an attempt to derogate. A positive moral norm or normative order may claim to be superior to a legal order. A fictional normative order (e.g. natural law) may make the same claim to be anchored in the ‘cosmos of values’, as Verdross put it in 1926,217 and thus to be superior to ‘contingent’ positive regulation. Indeed, a norm an individual creates may claim to be superior to international law. Nobody would, however, honestly argue that in this case personal regulation ‘actually’ is superior to international law.
Because the claim to be observed is inherently equal in all norms, the super- ordination of norms cannot be based on mere claims. A norm can claim to be subordinate to another norm, but this works only because it incorporates that other norm.
However, what other criteria can validly be used to establish superiority? Even within a normative order, a norm may not be directly connected to another – why should there be a hierarchy? We can, however, identify one clear case of hier- archy: a norm depends for its validity on the norm authorising its creation. Only the direct regressus of validity-dependence establishes a hierarchy and it creates a chain of norms leading from the Grundnorm to the lowest norm.218
A norm authorising the creation of norms (norm ‘A’) can be the source of many norms, including a prohibition (‘B’) as well as another authorising norm (‘C’). If a norm created by the authorising norm C (‘D’) conflicts with B as a
‘brother norm’ of the authorising norm C, one might be tempted to conclude that B is higher than D, but this is not so. The lower norm D depends only upon its source-law (directly on C, indirectly also on A) and not upon other, indirectly higher norms like B. ‘A rule is only erroneous with respect to the rules which are directly conditional for its validity.’219 Kelsen saw conflict as impossibility only in this sense:
There cannot exist a conflict between a higher norm and a lower norm, i.e. between a norm determining the creation of another [norm] and that other norm, because the lower norm’s validity derives from the higher norm.220
Therefore, unless a potentially higher norm can be interpreted as part of the meta-norm on norm-creation of the lower norm, it is not lex superior in our sense (Section 5.5.3.3).
217 Verdross (1926) supra note 190 at 31: ‘im Kosmos der Werte verankert’.
218 Schilling (1994) supra note 40 at 402.
219 ‘[D]ie Fehlerhaftigkeit einer Rechtsvorschrift besteht immer nur Hinblick auf die direkt bedin- gende Regelung.’ Lippold (2000) supra note 116 at 390 (emphasis removed).
220 ‘Zwischen einer Norm hửheren Stufe und einer Norm niederen Stufe, das heiòt zwischen einer Norm, die die Erzeugung einer anderen bestimmt, und dieser anderen Norm kann kein Konflikt bestehen, da die Norm der niederen Stufe in der Norm der hửheren Stufe ihren Geltungsgrund hat.’ Kelsen (1960) supra note 2 at 212 (Ch 34 e); Kelsen (1979) supra note 8 at 207 (Ch 59 I e).
Uncertainty in International Law 186
5.5.3.2 A power to derogate?
If a potentially higher norm not only conflicts with, but claims to derogate from, the lower norm – if, in our case, B claims ‘norm D is herewith repealed’ – the situation becomes more complicated. We cannot simply take derogating norms at their word and admit derogation of norms vis-à-vis any norms that claim to do so.
In addition to the problems of defining a higher norm outside the realm of the Erzeugungsregel we are still left here with our uncertainty as to the actual derogatory force of derogating norms. As mentioned above, to claim that higher norms dero- gate from lower norms because they are higher is a tautology. The matter seems to be different if the meta-norm on norm-creation itself (in our case: C) were to stipulate the conditions for derogation as well. ‘Only the source of validity can also be the source of loss of validity of a norm.’221 However, conditions for law- destruction are precisely not conditions for law-creation; therefore, any derogation- conditions are not creation-conditions (Erzeugungsregeln). Therefore, C in this case is not in a different position from B as other ‘higher’ norms claiming to derogate.
5.5.3.3 The paradox of truly higher law
What about the case where a norm does not conform to the conditions for its creation? The easy answer is that such a norm simply does not exist. No deroga- tion is necessary, because no norm was created in the first place:
Legal science has to acknowledge as law each act pretending to be law, seemingly having legal quality, . . . if it is in fact determined by delegating higher law; and it has to deny [the act’s] claim to validity . . . if it is not determined by the condition [in higher] law.222 There no need to invoke powers of derogation nor to worry about the lex superior maxim, because an automatic relationship obtains between these two norms – or so it seems. The Pure Theory of Law, however, has problematised norm-conflicts which occur when erroneous acts are created in positive legal orders. Its responses are the concepts of the Error Calculus (Fehlerkalkül) and the Tacit Alternative Clause (Alternativermọchtigung) to reconcile the unique nature of norms with the reality of positive enactment. The Fehlerkalkül theory is a crucial, often misunder- stood and complex part of the Pure Theory of Law.223 While a full discussion of
221 ‘[N]ur die Quelle der Geltung kann zugleich auch Grund der Nichtgeltung einer Norm sein.’
Merkl (1923) supra note 77 at 256 (emphasis removed); Thienel (1988) supra note 42 at 26–27.
222 ‘Die Rechtswissenschaft nun hat jeden solchen mit dem Prọtexte der Rechtsnatur auftretenden, mit dem Scheine von Rechtsqualitọt begabten Akt als Recht anzuerkennen, . . . wofern er sich durch das delegierende hửhere Recht tatsọchlich determiniert erweist; und sie hat dem Akte seinen Geltungsanspruch abzuerkennen, . . . wenn er die Determination durch das bedingende Recht vermissen lọòt.’ Merkl (1923) supra note 77 at 286–287 (emphasis removed).
223 This discussion of the Fehlerkalkül theory draws heavily on Christoph Kletzer’s paper on the subject: Christoph Kletzer, Kelsen’s development of the Fehlerkalkül-Theory, 18 Ratio Juris (2005) 46–63.
5.5.3 Conflict of norms 187
the minutiae is neither possible nor necessary here, we will briefly analyse the impact of the two concepts on the conflict of norms in international law.
What happens if a court judgment delivers a ‘wrong’ verdict? What happens if the Security Council finds a ‘threat to the peace’ under Article 39 UN Charter when there is no such threat? A student of the Kelsenian theory of the hierarchy of norms up to this point might answer that these two decisions are void, they are a nullity. This statement is in principle correct, yet there are problems with this view:
[E]ven the slightest violation of a condition for the validity of, for instance, a judgement would result in the immediate nullity of the judgement. . . . Thus, only two kinds of judgements could be said to exist, neither of which would allow for appeal: (a) perfect judgements, against which appeal is per definition impossible, and (b) non-judgements, which cannot have legal effect and against which appeal is thus useless.224
Any positive regulation of appeals procedures would be useless. Since these pro- cedures do not exist with respect to the Security Council, this would not matter anyway. Even if that did not matter (and ontologically it does not), there is another problem. In such cases we are indeed faced with an act of will whose sense is a norm and hence we do have the pretence of norm-creation. Yet on the other hand that act did not fulfil the conditions laid down by the rules on law-making.
The Pure Theory of Law developed the two theories in order to reconcile two sets of dichotomies. First, there is the impossibility of ‘law contrary to law’225 which clashes with the partial acceptance of it in various positive legal orders. Second, we have the presence of a positive act of will which clashes with the partial non- fulfilment of the meta-law on law-making.
(1) Adolf Merkl invented the Error Calculus in order to reconcile this first dichotomy. It is a feature of positive law that allows us to cognise acts as law, despite a modicum of errors in their creation. Positive law, he believes, sometimes distinguishes between voidness (absolute Nichtigkeit) and voidability (Vernichtbarkeit);
this is reflected in a two-stage process of law-creation. The conditions are bifur- cated into maximum and minimum conditions for law-creation. An act that does not fulfil the minimum conditions is not a norm in the first place; an act that fulfils all minimum conditions, but does not fulfil at least one of the maximum condi- tions is law, but faulty law.226 The positive regulation of the bifurcation is usually accomplished by stipulating grounds of appeal. If the act of law-creation does not fulfil the conditions a–c, no norm results; if a norm fulfils conditions a–c, but not d–f, it is open to appeal and can be voided there.227
Yet there are three problematic features of Merkl’s theory. First, it does not answer the theoretical question raised in the second dichotomy of positive acts of
224 Kletzer (2005) supra note 223 at 47.
225 Kelsen (1960) supra note 2 at 271 (Ch 35 j α).
226 Lippold (2000) supra note 116 at 324.
227 Kletzer (2005) supra note 223 at 48.
Uncertainty in International Law 188
will not fulfilling the conditions. This is why Kelsen invented the Tacit Alternative Clause, discussed infra. Second, the Error Calculus depends upon positive regula- tion for its existence. If, as in most of international law, including the Charter, there is no distinction between minimum and maximum conditions, all condi- tions, even the least ‘important’ ones, are by default minimum conditions. All conditions for law-creation have to be fulfilled for law-creation to work.
Third, it can be argued that all true conditions are minimum conditions any- way and that Merkl’s Error Calculus is not truly a differentiation between condi- tions for law-creation. If successful creation of a norm – however ‘faulty’ the resultant norm may seem – no longer requires some of the original conditions stipulated by the meta-law (e.g. conditions d–f of the original a–f), then these are not conditions for the creation of law! On a consistently normativist reading of the Error Calculus doctrine, it is no more than the creation of a possibility of derogating from a certain category of norm (those fulfilling a–c, but not d–f). This possibility is perhaps enshrined in a different norm altogether and the ‘faulty norms’ (fulfilling a–c, but not d–f) are properly valid, but are destroyed after they are created. Derogation, however, is not conditional upon a ‘faulty’ norm; deroga- tion can be conditioned on any set of facts, e.g. that the norm in question is a lex prior. Hence, because in our case the norm in question is validly created, it is also a member of the legal order. Rudolf Thienel takes the example of the derogation of a norm by virtue of an automatic condition for law-destruction enshrined in the norm’s source of validity, e.g. a ‘sunset clause’, and distinguishes it from the case where another norm claims to derogate.228 Once a norm is valid, however, its derogation by whatever norm is no longer part of the condition for law-creation. As a matter of ideology, the voiding of a statute by a constitutional court may seem to be the court’s cognition of the unconstitutionality of that statute, but legally it is a derogation by the court using a provision defining certain conditions for the derogation of statutes by the court.
(2) The second dichotomy is tackled by Kelsen’s own Tacit Alternative Clause (Alternativermọchtigung), a doctrine that has proven rather difficult to interpret. It was created because the Error Calculus is unable to hinder a ‘faulty’ decision from being perpetuated to the highest judicial instance of a legal order which is no longer subject to appeal. The rulings of such organs as the International Court of Justice or the Supreme Court of the United Kingdom are not final because that quality is ‘accorded by the positive law itself’.229 Rather, the opposite is the case.
Since appeals procedures are created by positive regulation, all norm-creation is by default final unless an appeals procedure is expressly enacted.
As Christoph Kletzer points out, there are only three possible solutions to an
‘unlawful’ final decision. (a) The norms are valid, but annullable – which would make the decision anything but final. (b) They are void ab initio, a solution which he criticises for being unascertainable without the law specifying someone to
228 Thienel (1988) supra note 42 at 26–27.
229 Kletzer (2005) supra note 223 at 51.
5.5.3 Conflict of norms 189
ascertain – which would not make the decision final either. (c) The only option left is that they are fully valid and not subject to appeal. In this case ‘we have to find legal rules in which the validity of these “unlawful” final decisions is grounded’.230 Kelsen’s train of thought is that finality means more than non-annulability.
Finality means that all organs authorised to create norms are not bound only by the meta-law on law-creation. They are bound by the concrete Erzeugungsregel, but also by a completely different norm:
That the legal order confers the force of law to a judgement of a court of last instance means that not only is a general norm valid that predetermines the content of the judgment, but also a general norm according to which the court may for itself deter- mine the content of the individual norm to be created. The two norms form a unit, because the court of last instance is authorised to create either an individual legal norm whose content is predetermined by the general norm, or an individual norm whose content is not so predetermined, but is to be determined by the court of last instance for itself.231
This solution is counter-intuitive, but seems to be the only way to reconcile the finality and the possibility of error. The problem is, however, that if the alternative clause is part of the positive law, it is not the Tacit Alternative Clause, but part of the positive conditions for law-creation. A tacit clause, on the other hand, is not positive law. Only positive norms can be members of a positive normative order;
fictitious norms cannot (Section 6.2.2.2). To impute a tacit meta-norm is to intro- duce a fictional norm into an otherwise positive normative order. This is a still- born child of aberrance from the ideal of purity of method in the Vienna School’s programme.
Thus even in the most obvious case of a norm-conflict being resolved ex ante, even in this easy case we have to face a problem that becomes a paradox. On the one hand, the duality of Is and Ought – the very idea of norms – demands of us to found the validity of norms only on norms. Norm-creation is such a founding exercise, for in creation the validity (existence) of norms is established. The opposite must there- fore also obtain: without a basis in norms, alleged norms cannot be norms – it cannot be otherwise, for basing a norm on fact alone would transcend the duality of Is and Ought and therefore make it impossible for us to cognise a ‘something’ as norm. The normativity of normative systems demands a strict foundation in norms.
230 Kletzer (2005) supra note 223 at 52.
231 ‘Die Tatsache, daò die Rechtsordnung einer letztinstanzlichen Gerichtsentscheidung Rech- tskraft verleiht, bedeutet, daò nicht nur eine generelle Norm in Geltung steht, die den Inhalt der gerichtlichen Entscheidung vorausbestimmt, sondern auch eine generelle Norm, derzufolge das Gericht den Inhalt der von ihm zu erzeugenden individuellen Norm selbst bestimmen kann.
Diese beiden Normen bilden eine Einheit; so zwar, daò das letztinstanzliche Gericht erm- ọchtigt ist, entweder eine individuelle Rechtsnorm, deren Inhalt durch die generelle . . . Norm vorausbestimmt ist, oder eine individuelle Rechtsnorm zu erzeugen, deren Inhalt nicht so voraus- bestimmt ist, sondern durch das letztinstanzliche Gericht selbst zu bestimmen ist.’ Kelsen (1960) supra note 2 at 273 (Ch 35 j α).
Uncertainty in International Law 190
On the other hand, only an authorised organ is authorised to decide – and its decision is a decision, not cognition of law. If and when the Security Council says there is a ‘threat to the peace’ under Article 39 UN Charter, there is a threat to the peace because it (and no one else) is authorised by law to make that decision.232 The Council, courts, and law-makers in general are authorised to decide – and it is their actual decision that counts. There is no logical deduction of a norm from a higher norm, as Kelsen shows in the second part of his critique of the role of logic in normativist legal scholarship,233 the first part being directed against the prin- ciple of excluded contradiction (Section 5.3.1). A criminal court’s judgment is not a logical deduction from the Penal Code, for the act of will is a necessary condi- tion for the creation of a positive norm. The authorising norm does not ‘contain’
the norms that can potentially be created under it. Only a real act of will creates them. Therefore, positive law-making takes on a quasi-autonomous form – the creation is constitutive, whether or not it conforms to the conditions. The positivity of norms gives positive acts of will creative powers.
This is the paradox of positive normative orders: a truly superior norm endangers positivity, while a truly positive norm endangers the unity of the nor- mative order. It is not a paradox of the Vienna School, but of the very nature of positive normative orders. The Pure Theory only brings to light what other theor- ies manage to hide behind pragmatism. With his dialectical completion between the traditional positivists’ emphasis on empirical creation of law, on the one hand, and the naturalists’ emphasis on the normative on the other hand – with his Copernican revolution in legal thought – Hans Kelsen created a viable normativist-positivist theory of norms. Yet this creation has its weak points, for as we face the paradox we find that Kelsen can solve it only by introducing an
‘impure’ element – the Tacit Alternative Clause. A pure theory of law as norms must be consistent – sans peur et sans reproche – wherever the cognition of what is there may lead us. In this spirit, it is hoped to present a more consistent (if not more ‘user-friendly’) approach here.
(3) The take on the paradox adopted here is that any act of will purporting to create a norm does indeed do so, whether or not it fulfils the conditions for norm- creation of one given normative order. If it fulfils these conditions, it is a norm belonging to that normative order. If it does not, it is a norm not belonging to this order, but belongs to its own normative order, consisting of it and a Grundnorm necessary to perceive it as norm. The Grundnorm, as a supposition234 that is neces- sary to be able to conceive of norms (Section 7.1.2), serves as an ‘as if’-element.
Presupposing a basic norm means that the ‘faulty norm’ can be seen as norm. If, for example, the creation of a statute under the municipal law of a country specifies conditions a–f and the act in question fulfils a–c only, it is a norm as long as there is an act of will to that effect; yet it is not a norm of that particular municipal system.
232 Kelsen (1960) supra note 2 at 274 (Ch 35 j α).
233 Kelsen (1979) supra note 8 at 179–203 (Ch 58).
234 Kelsen (1979) supra note 8 at 206–207 (Ch 59 I d).
5.5.3 Conflict of norms 191