4 Interpretation and modi fi cation
5.5.2 The hierarchy of legal orders
The concept of the hierarchy of legal orders (Stufenbau) is in many ways one of the most important contributions of the Pure Theory of Law’s normativist-positivist approach181 and will briefly be introduced here. The dichotomy of Is and Ought, and, with it, the theory of the Grundnorm forms the foundation of the Pure Theory as a whole (Chapter 7). Hierarchical ordering as a concept is a direct consequence of that core theory182 – not, though, the concrete arrangement of hierarchy that may obtain in concrete normative orders. If the existence of a norm as validity can only be based on another norm183 – if there can be no Ought from Is alone184 – then a connection between norms is established. This hierarchy is the depend- ence of one norm upon another norm for its validity. Since that dependence is one-sided, it makes sense to call the dependent norm a ‘lower’ norm and the norm it depends upon the ‘higher’ norm. The dependence in the sense described above is established by norm-creation. The question ‘Why ought I to obey this statute?’ is answered by referring to the norm that has authorised its creation, for example a constitution. The higher law empowers law-creation; that empower- ment is the reason the resultant law is valid.185
The dynamic character of law makes a norm valid, if and when it was created in a certain fashion determined by another norm. This other norm is the immediate source of validity of the first norm. The relationship between the norm which regulates the creation of another norm and the norm thus created can be visualised as a spatial super-ordination and subordination. . . . The legal order is not a system of coordinate legal norms existing alongside each other, but a hierarchical ordering of various strata of legal norms. Their unity is constituted because a norm which has been created according to the terms of another norm derives its validity from that latter norm, whose creation is, in turn, determined by yet another norm; a regressus ending in the Grundnorm, [whose validity] is presumed.186
181 András Jakab, Probleme der Stufenbaulehre, 91 Archiv für Rechts- und Sozialphilosophie (2005) 333–365 at 333; Theo ệhlinger, Der Stufenbau der Rechtsordnung. Rechtstheoretische und ideologische Aspekte (1975) 9.
182 Behrend (1977) supra note 140 at 61.
183 Kelsen (1960) supra note 2 at 196 (Ch 34 a).
184 Kelsen (1960) supra note 2 at 5 (Ch 4 b).
185 Kelsen (1979) supra note 8 at 82 (Ch 26 I).
186 ‘Da bei dem dynamischen Charakter des Rechts eine Norm darum gilt, weil und sofern sie auf eine bestimmte, das heiòt durch eine andere Norm bestimmte Weise erzeugt wurde, stellt diese den unmittelbaren Geltungsgrund für jene dar. Die Beziehung zwischen der die Erzeugung einer anderen Norm regelnden und der bestimmungsgemọò erzeugten Norm kann in dem rọumlichen Bild der ĩber- und Unterordnung dargestellt werden. . . . Die Rechtsordnung ist nicht ein System von gleichgeordneten, nebeneinanderstehenden Rechtsnormen, sondern ein Stufenbau verschiedener Schichten von Rechtsnormen. Ihre Einheit ist durch den Zusammenhang hergestellt, der sich daraus ergibt, daò die Geltung einer Norm, die gemọò einer anderen Norm erzeugt wurde, auf dieser anderen Norm beruht, deren Erzeugung wieder durch andere bestimmt ist; ein Regreò, der letztlich in der – vorausgesetzten – Grundnorm mỹndet.’
Kelsen (1960) supra note 2 at 228 (Ch 35 a).
Uncertainty in International Law 180
It is creation that establishes hierarchy; it establishes the hierarchy of legal con- ditionality (Stufenbau nach der rechtlichen Bedingtheit).187 If, and only if, all conditions imposed by the meta-law on law creation (Rechtserzeugungsregel)188 are met, can the norm created be cognised as a norm of the normative order in question.189 Only then can the norms be ordered in a multitude of spheres between delegating and delegated norms.190 Only then can both norms be seen as belonging to one normative order.191 Because Ought can only come from an Ought, law necessarily orders its own creation.192
Yet because an authorising norm usually authorises the creation of a multitude of norms, and because usually a multitude of norms are created under it, this multitude of norms with a common ‘pedigree’ has been called a Rechtsform,193 what international lawyers would probably describe as ‘norms belonging to the same source’ (Section 6.1). ‘The form legal rules take is determined by the rule that created them; the legal rules that were created according to the same rule of [law-]creation have the same form.’194 Thus in international law, we could call
‘customary international law’ one Rechtsform, because all norms belonging to it were created according to the same rules on custom-creation.
A problem with this deduction appears if we take this conditionality seriously.
In complex modern municipal legal systems the meta-norms on law-creation in its totality may consist of a large part of that legal order, because it is all the condi- tions put together, including norms on who is authorised to create norms having which content observing which procedure.195 This not only creates a very complex network of norms, which may very well lead to epistemological uncertainty. It may also be and is the case that a norm belonging to a lower Rechtsform is a condition for the creation of a norm belonging to a higher Rechtsform. A ‘simple’
statute may contain one of the conditions for the creation of a constitutional law.196 This has the potential to destroy the notion of a uniform and simple
187 Walter (1964) supra note 140 at 60.
188 Walter (1964) supra note 140 at 61.
189 Kelsen (1960) supra note 2 at 239 (Ch 35 e–f).
190 Merkl (1923) supra note 77 at 286–287; Alfred Verdross, Die Verfassung der Vửlker- rechtsgemeinschaft (1926) 43.
191 Merkl (1931a) supra note 140 at 1335–1336.
192 Kelsen (1960) supra note 2 at 73 (Ch 15).
193 Merkl (1931a) supra note 140 at 1311.
194 ‘Der Bestimmungsgrund für die Form der Rechtsvorschriften ist ihre Erzeugungsregel; die gleiche Form haben jene Rechtsvorschriften, die nach der gleichen Erzeugungsregel geschaffen wurden.’ Walter (1964) supra note 140 at 55 (emphasis removed).
195 Walter (1964) supra note 140 at 59–60, 61 (FN 111).
196 This is the case with an oft-discussed provision of the Austrian constitution, where observance of the Law on the Federal Gazette (Bundesgesetz über das Bundesgesetzblatt 2004 (Bundesgesetz- blattgesetz – BGBlG), BGBl I 2003/100) – i.e. publication in the Gazette – is made one of the conditions for law-creation even of constitutional laws by virtue of Article 49 of the 1920/1929 Federal Constitution (Art 49 Abs 1 B-VG 1920 supra note 166). For a discussion of this particular case: Lippold (2000) supra note 116 at 394–398; Walter (1964) supra note 140 at 62.
5.5.2 Conflict of norms 181
hierarchy of Rechtsformen in a hierarchical-pyramidal structure, because the condi- tions transcend the hierarchy of form and partially overturn it.197 Robert Walter diverges from Merkl’s view by denying that Rechtsform can be a criterion for hier- archy.198 One could agree with Rainer Lippold that while it may be highly counter-intuitive to think of a statute as ‘higher’ than the constitution, if one consistently defines ‘form’ as ‘condition of law-creation’, then this statute is indeed part of a higher layer.199
In international law this is not such a problem. While international law neces- sarily has a hierarchy (a constitution in the material sense), it is quite uncertain how and whether a unified hierarchy exists. In international law, the problem is not a highly complex network,200 but the apparent lack of any positive rules on rule- making. We will look at this problem more closely in Chapter 6.
But there is that other Stufenbau in the Pure Theory, the hierarchy of derogatory force (Stufenbau nach der derogatorischen Kraft),201 also developed by Merkl, but not explicitly distinguished by Kelsen.202 This hierarchy is highly relevant to our topic, since norm-conflicts can only be solved by derogation. We need to know when norms can validly derogate from each other. In other words, we need to know which norms have that sort of derogatory force. Expecting clear answers from the Pure Theory’s protagonists, one may be disappointed to read time and again that they approach the matter from the other angle. ‘A legal norm which has derogat- ing force vis-à-vis another legal norm, while the latter . . . has no such derogating force therefore holds a higher rank.’203 Adolf Merkl is interested in portraying hierarchies, not in derogating force as such, which in this case is merely the criterion for the ordering of norms.
This is because the first Stufenbau is a necessary element of all normative orders.
Every normative order has at least two layers of norms. It has at least the positive norm created and the presumed (quasi-fictional) Grundnorm. If ‘A’ had never issued a norm and would now do so, this norm, e.g. ‘All humans ought to wear red hats’, can only be cognised as a norm if the Grundnorm: ‘Follow A’s orders’ is presupposed. Thus, there are at least two levels in any normative order, although there can be more. However, for the reasons developed in Sections 5.2 and 5.4, the hierarchy of derogatory force is not a necessary element of every normative order. Derogation never is a logical operation and needs to be stipulated by
197 ệhlinger (1975) supra note 181 at 16–17.
198 Walter (1964) supra note 140 at 62.
199 Lippold (2000) supra note 116 at 399.
200 ệhlinger (1975) supra note 181 at 17.
201 Walter (1964) supra note 140 at 55.
202 Behrend (1977) supra note 140 at 42.
203 ‘Ein Rechtssatz, der gegenỹber einem anderen Rechtssatz derogierende Kraft hat, wọhrend dieser . . . ihm gegenỹber keine derogierende Kraft hat, ist aus diesem Grunde von hửherem Range.’ Merkl (1931a) supra note 140 at 1340; Walter (1964) supra note 140 at 54; Robert Walter, Der Stufenbau nach der derogatorischen Kraft im ửsterreichischen Recht. Zum 75. Geburtstag von Adolf Julius Merkl, 20 ệsterreichische Juristen-Zeitung (1965) 169–174 at 169.
Uncertainty in International Law 182
positive norms; derogability is a feature of positive regulation.204 Hence it depends on the positive norms within a concrete normative order whether such a hierarchy of derogation exists and what it looks like.205 However, we must not fall into the trap of begging the question thus: if a hierarchy is created by derogation, the higher norm will derogate the lower norm, thus any derogating norm is higher and any norm that claims to derogate derogates. In this way one would only have proven that the hierarchy of derogation comes from derogation, while the dero- gating force comes from hierarchy.206 This sounds like a tautology, but we must have a closer look at this, since this argument has been used frequently.
In particular, we must answer the question whether the hierarchy of legal conditionality can be the basis for the hierarchy of derogation. Is norm-creation the argumentative basis that identifies the lex superior in the lex superior maxim? (It needs to be said, however, that the question of identifying the lex superior is a different question to that of whether the lex superior maxim truly obtains by neces- sity.) Prima facie this option seems attractive, even natural. Should not the norm that has created another norm also determine when it should end? Why should not a constitutional provision simply destroy a statute? Erich Vranes argues that
‘[i]f there is hierarchically higher and lower law in a legal order, the lex inferior has in principle to yield to the lex superior, if the structure of the legal order is not to be led ad absurdum.’207
In contrast, Merkl holds that the two hierarchies may diverge significantly. For him, a norm being called ‘higher’ than another does not automatically mean that it is higher in every respect.208 A good example of the divergence of the two hierarchies may be the relationship that obtains between a statute and the judg- ment of a constitutional court derogating from that statute.209 At best, the author- isation for the constitutional court to derogate from statutes will be found in the same constitution as the parliament’s authorisation to create statutes. Thus, at best, they have the same rank in the hierarchy of validity. In the hierarchy of derogation, however, the judgment claims to derogate from the statute and thus there is at least the possibility that it is higher than the statute. It may not be the case that equal origin means equal rank, for solely from the fact that one source (a constitution, for example) provides a multitude of authorising norms for a multi- tude of Rechtsformen, e.g. statutes and administrative orders, one cannot conclude that they are of equal rank.210
204 Behrend (1977) supra note 140 at 36–38.
205 Robert Walter portrayed this type of hierarchy for the Austrian legal order in 1965: Walter (1965) supra note 203.
206 Schilling (1994) supra note 40 at 400.
207 ‘Soweit es in der Rechtsordnung hửherrangiges und niederrangiges Recht gibt, muss die lex inferior im Grundsatz der lex superior weichen, wenn nicht die Struktur der Rechtsordnung ad absurdum geführt werden soll.’ Vranes (2005) supra note 116 at 397–398.
208 Merkl (1931a) supra note 140 at 1342.
209 Behrend (1977) supra note 140 at 39.
210 Merkl (1931a) supra note 140 at 1340–1341.
5.5.2 Conflict of norms 183
Here we have the first moves towards a collapse of the two Stufenbauten into one unified hierarchy, or at least the recipe for a close connection. In Der Aufbau der Rechtsordnung Robert Walter argues that while positive law determines what derogatory force a norm has, it does not say so explicitly. Therefore, we have to deduce the amount of derogatory force from other rules, in particular from the norms on law-creation. These are important, because the importance (Bedeutung) of the resulting norms within the legal system is determined by how complicated the procedure of law-creation is shaped. If, he argues, the creation of norm X is tied to conditions a–c, while the creation of norm Y is tied to conditions a–e, one has to deduce from this that Y as the law created by a more complex pro- cedure cannot be derogated from by X as a norm created under simpler conditions.211
While this may work in the relationship between two modes of norm-creation within a domestic legislature, such a differentiation seems less likely in inter- national law. Is customary international law more complex to create than inter- national treaty law? Without attempting to answer this question we can say that these two sources are not comparable. Neither constitutes a more ‘complex’
mechanism, because they are different. While treaties are forms of contract, the other is a customary process. The most effective counter-argument against Walter’s distinction is that the criterion used (the complexity of norms) is at heart an empirical distinction and within the Pure Theory this would mean the intro- duction of an extraneous element into norms which cannot change the ontology of ideals. The particular method with which norms are created itself cannot determine whether a norm can derogate from another norm.
Walter continues that it has to be assumed that norms having the same conditions for law-creation, as stipulated by the validity-hierarchy (Rechtsform), also have the same derogatory force, because if that were not so, a differentiation of Rechtsformen according to their derogatory force would be impossible. If it is so assumed, the norms on law-creation (Erzeugungsregel) determine not only form, but also derogatory force. Walter holds that assumption to be true, for ‘derogatory force’ is nothing but a specific form of competence necessarily granted by the Erzeugungsregel.212
At this point the gauntlet is picked up by Theo ệhlinger, who counters that the last argument is not a necessary conclusion from the Pure Theory’s conception of hierarchy. The Erzeugungsregel only determines that the norm has a derogating function – it only determines the derogating norm. The norms on norm-creation do not tell us what it can do to the norm whose validity it purports to end. They do not tell us whether the derogating norm is in a position vis-à-vis the norm purportedly derogated from that allows it to actually derogate. ‘But this relationship between derogating and derogable norm is what matters for the hierarchy of
211 Walter (1964) supra note 140 at 59.
212 Walter (1964) supra note 140 at 59–60; Walter (1965) supra note 203 at 170; cf. Schilling (1994) supra note 40 at 401.
Uncertainty in International Law 184
derogation.’213 ệhlinger believes that he has detected un-Merklian thoughts in Walter’s argument. If there is a connection between the conditions of law- creation and derogation, it is only a matter of positive law being pragmatically made in this form, not a logical necessity.214 Merkl himself strongly doubted such a connection.215 ệhlinger’s admission that Walter’s thesis does admit a divergence between the two hierarchies216 may not be helpful in determining the source of the justification for derogation.