The temporality of normativity

I The Grundnorm as the foundation for the validity of legal norms

The starting point on which Kelsen’s entire legal theory is based is a strict separation between the domain of facts and norms: a presupposition that the Austrian jurist imports directly from the Kantian philosophical system, where it is understood as a condition of possibility of normativity itself. ¹⁰ The implication Kelsen derives from it had, however, not been formulated in these terms by Kant himself: if facts and norms are categorically irreducible to each other, Kelsen claims, it follows that the foundation for the validity of legal norms can lie only within the domain of norms themselves. From this, in turn, he deduces that norms cannot exist in isolation from each other, which is another way of saying that a norm can exist only as part of a system of norms. ¹¹

As was pointed out by Joseph Raz – one of Kelsen’s most insightful commentators – this constitutes an important contribution to legal philosophy in its own right, because it is sufficient to exclude the whole range of theories that define legal norms with reference to a supposed intrinsic nature, either of the norms themselves (as is the case with natural law doctrines) or of the authority positing them (as is the case with realist theories of sovereignty). ¹² In this respect, however, it is important to point out that Kelsen also draws an important distinction between two types of systems of norms, which serves the purpose of distinguishing between legal orders and other types of normative systems.

For Kelsen, a system of norms can be either ‘nomo-static’ or ‘nomo-dynamic’, depending on the way in which the relations between norms are construed. In a ‘nomo-static’ system, these relations are of a logical nature, in the sense that ‘lower’ norms are deduced from ‘higher’ ones as inferences from presuppositions. For instance, in the domain of morality, it is usually held to be possible to deduce specific obligations from more general moral principles. By contrast, the kinds of normative orders Kelsen refers to as ‘nomo-dynamic’ are characterized by the fact that their norms are created according to procedures defined by ‘higher’ norms within the same system. The relationship between ‘higher’ and ‘lower’ norms therefore has nothing to do with the substantive content of the norms in question, but rather with the authority that poses them: in a ‘nomo-dynamic’ system, ‘higher’ norms authorize specific authorities to create ‘lower’ ones through determined procedures. ¹³

On the basis of this distinction, Kelsen affirms that the distinctive feature of juridical systems of norms is that they are ‘nomo-dynamic’, in the sense that the norms belonging to them have been created through procedures stipulated by ‘higher’ norms within the same system. It is for this reason that Kelsen states that the legal order must be understood as a ‘hierarchy of norms’ that ‘regulates the mechanism for its own production’: because norms can be identified as part of a valid legal order only if they have been created in conformity with the procedures stipulated by ‘higher’ norms belonging to the same order. Within this framework, legal validity accordingly functions as a sort of link, or synapsis, that always points ‘upwards’.

This formulation explains why Kelsen’s legal theory may seem to pose the problem of the ultimate foundation for the validity of legal norms in a particularly acute way. For, if we assume that legal validity is a sort of synapsis that always points ‘upwards’, there emerges the question of establishing what there is at the summit of the system; that is, at the highest point of the ‘hierarchy of norms’. Indeed, at many junctures, Kelsen can be read as attempting to solve this problem with reference to the notion of a Grundnorm, arrived at by following the regressive movement implicit in his conception of legal validity to the point where he claims it must reach the idea of a norm that has not been created in conformity with any previous norm, but whose validity must nonetheless be presupposed as a condition for making sense of the rest of the system. In a section of his General Theory of Law and State entitled ‘The Reason of Validity of a Legal Order: The Basic Norm’, for instance, Kelsen writes:

The derivation of the norms of a legal order from the basic norm of that order is performed by showing that particular norms have been created in accordance with the basic norm. To the question why a certain act of coercion – e.g. the fact that one individual deprives another of his freedom by putting him in jail – is a legal act, the answer is: because it has been prescribed by an individual norm, a judicial decision. To the question why this individual norm is valid as part of a definite legal order, the answer is: because it has been created in conformity with a criminal statute. This statute, finally, receives its validity from the constitution…The validity of this first constitution is the last presupposition, the final postulate upon which the validity of all the norms of our legal order depends. Only upon this presupposition are the declarations of those to whom the constitution confers norm-creating power binding norms…This is, schematically formulated, the basic norm of the legal order. ¹⁴

This response to the problem of the ultimate foundation for the validity of legal norms is, however, subject to a possible misunderstanding. For, if Kelsen is interpreted as implying that the Grundnorm itself – i.e. the presupposition that the authority positing the ‘highest’ positive norm was authorized to do so – is the foundation for the validity of the legal order, there remains the problem of establishing the grounds for that presupposition to begin with.

From this perspective, the Grundnorm can appear either as a fact (in the sense that people at the time actually assumed this was the case) or as a norm (in the sense that there would be a norm stipulating that we ‘ought’ to assume that was the case). Either way, however, the notion of the Grundnorm seems incapable of escaping the binary between legal realism and natural law doctrine. For, if it is just a fact, then the theory of the Grundnorm would seem to devolve into a form of legal realism. Alternatively, if we ought to suppose that the authority positing the ‘highest’ positive norm of a given legal order was authorized to do so, then the theory of the Grundnorm would seem to beg the question of the normative grounds for that ‘ought’. In absence of any prior foundation for it, it would seem to have to be understood as some kind of ‘natural’ or ‘absolute’ law.

Indeed, most of the criticisms that have been raised against Kelsen’s theory of the Grundnorm stem precisely from such a reading. Since this is the first confusion I intend to dispel in what follows, it is useful to examine the way in which such criticisms have been formulated, before advancing an alternative reading of the theory of the Grundnorm that is not vulnerable to them.

II Grounds for a ‘transcendental’ reading of the Grundnorm

The main criticisms that have been raised against Kelsen’s theory of the Grundnorm can be classed in three categories. First, it has been advanced that this notion can ultimately be interpreted only as referring to a fact, which would accordingly contaminate the ‘purity’ of Kelsen’s legal theory with an element of legal realism. Already in 1926, for instance, Herman Heller accused Kelsen of having surreptitiously reintroduced facticity in his philosophy of law through the notion of the Grundnorm. ¹⁵ The most notorious advocate of this thesis is, however, Carl Schmitt, who famously affirmed in his Constitutional Theory that Kelsen’s positivism ultimately boils down to a ‘tautology of raw factuality’, since its rejection of any conception of substantive justice implies that the validity of the legal order can only be based on the fact that it is obeyed – i.e. on the fact that it is ‘effective’:

With Kelsen [Schmitt writes] only positive norms are valid, in other words, those that are actually valid. Norms are not valid because they should properly be valid. They are valid, rather, without regard to qualities like reasonableness, justice, etc. but only because they are positive norms. The imperative abruptly ends here, and the normative element breaks down. In its place appears the tautology of raw factuality: a norm is valid, when it is valid, because it is valid. That is positivism. ¹⁶

The criticism of the notion of the Grundnorm raised by H. L. A. Hart in his book on The Concept of Law relies on a similar reasoning. The only difference is that for him the claim that the juridical order is ultimately based on a fact does not constitute a problem, since it is the core of the theory he puts forward in his own book. This is why Hart asserted that the Grundnorm is a ‘needless reduplication’ of the idea that the validity of the legal order ultimately rests on a fact:

If a constitution specifying the various sources of law is a living reality [he writes] in the sense that the courts and officials of the system actually identify the law in accordance with the criteria it provides, then the constitution is accepted and actually exists. It seems a needless reduplication to suggest that there is a further rule to the effect that the constitution (or those who ‘laid it down’) are to be obeyed. ¹⁷

The second category of objections that have been raised against Kelsen’s theory of the Grundnorm claims that this notion ultimately falls back into a disavowed natural law doctrine. In his article entitled ‘Validity and the Conflict between Legal Positivism and Natural Law’, for instance, Alf Ross affirms that since the Grundnorm is explicitly not construed as a positive norm, it can ultimately only be understood as a moral one, prescribing obedience to any legal order that is effectively in place. ¹⁸ Similarly, in his book on Hans Kelsen’s Pure Theory of Law: Legality and Legitimacy, Lars Vinx interprets Kelsen’s basic norm as implying a ‘teleological’ vision of the rule of law that instantiates ‘a minimum legal morality’ in Lon Fuller’s sense, one that respects the rational freedom of legal subjects. ¹⁹

This line of argument was developed further by Graham Hughes, on the basis of the observation that whenever Kelsen attempts to spell out the actual content of the Grundnorm, he suggests that it stipulates that we ought to treat the constitution of the legal order that is effectively in place as legally valid. For instance, commenting on a passage in Pure Theory of Law in which Kelsen writes that ‘the reason for the validity is the presupposed basic norm…according to which one ought to comply with an actually established, by and large effective, constitution’ (1933: 212) Hughes writes that:

In this passage, the notion of effectiveness does not appear simply as a condition for the validity of the legal order; rather, it has become part of the prescriptive content of the basic norm that one ought to comply with an effective constitution…In this sense, Kelsen appears to be propounding a general moral maxim to the effect that we ought to comply with effective coercive social orders. ²⁰

From this point of view, the fact that Kelsen refuses to supply the Grundnorm with any more substantive content appears even more problematic, since it suggests that his theory ultimately boils down to a ‘blind justification’ of the juridical order that is effectively in place. This may be the reason why even theorists who are not inclined to regard a fall back into natural law as problematic in itself, have judged such a ‘Kelsenian’ version of it completely insufficient. ²¹

Finally, a third category of objections raised against Kelsen’s theory of the Grundnorm claims that the concept is incoherent, because it oscillates between legal realism and natural law doctrine without succeeding in providing a viable alternative to either of them. In an article on ‘The Basic Norm and Democracy in Hans Kelsen’s Legal and Political Theory’, for instance, Andreas Kalyvas writes that: ‘The Grundnorm is the black box of Kelsen’s legal, constitutional and political theory…As such, it resides in an undefined zone, a veritable “no man’s land” between the formal and the material, validity and facticity, the objective and the subjective.’ ²²

In another article, published shortly after the first English translation of Kelsen’s Pure Theory of Law, Julius Stone advanced essentially the same objection in the form of three questions to which he claims Kelsen’s theory is constitutively incapable of providing an answer. In the light of the material covered above, these questions can be read as a way of summarizing all the most important objections that have been raised against Kelsen’s conception of the ultimate foundation for the validity of legal norms. The first asks whether the Grundnorm is to be interpreted as ‘internal’ or ‘external’ to the legal order itself; that is, as a ‘positive’ norm that grounds the validity of the system, or as a ‘hypothetical’ presupposition that is added retroactively by the jurist. The second concerns its content and asks whether the Grundnorm is to be understood as a purely ‘formal’ presupposition of legal theory, or whether it implies a ‘substantive’ theory of justice, on which its validity would be based. Finally, Stone suggests that it is unclear what Kelsen means when he asserts that the existence of a Grundnorm must necessarily be ‘presupposed’: is that meant to refer to a matter of fact or to a norm implicit in the natural order of things? ²³

Despite the evident proliferation of such objections, it is my contention that almost all of them can be met by clarifying that, within the framework of Kelsen’s system, the Grundnorm is not meant to function as the foundation for the validity of the legal order, but rather as a ‘transcendental’ condition of its possibility. This point has been made very clearly by Stanley Paulson in several of the texts where he has sought to bring out the analogies between the conceptual status of the Grundnorm in Kelsen’s theory of law and the idea of a ‘transcendental category’ in Kant’s critical philosophy. ²⁴

In a passage of his General Theory of Law and State introduced very soon after the one quoted above, for instance, Kelsen writes that:

By formulating the basic norm, we do not introduce into the science of law any new method. We simply make explicit what all jurists, mostly unconsciously, assume when they consider positive law as a system of valid norms and not as a complex of facts, and at the same time repudiate any natural law from which positive law would receive its validity…The basic norm is the answer to the question: how – and that means under what conditions – are all these juristic statements concerning legal norms, duties, rights and so on, possible? ²⁵

Even more explicitly, in an article entitled ‘Natural Law Doctrine and Legal Positivism’, included as an appendix to his General Theory of Law and State, Kelsen writes that:

We have already encountered above the idea of a basic norm, which from the point of view of legal positivism constitutes the ultimate assumption and hypothetical basis of any positive legal order…Just as the transcendental logical principles of cognitions (in the sense of Kant) are not empirical laws, but merely conditions of all experience, the basic norm itself is not a positive legal norm, because it has not been made, but is simply presupposed as the condition of all positive legal norms. And as one cannot know the empirical world from the transcendental logical principles, but merely by means of them, so positive law cannot be derived from the basic norm, but can merely be understood by means of it. ²⁶

These passages ought to make it absolutely clear that Kelsen sought to provide a purely ‘constructivist’ theory of law, according to which legal norms are not material entities existing in the world alongside other facts, but rather products of a specific way of looking at the world, governed by certain presuppositions. Drawing the Kantian analogy further, we could therefore say that for Kelsen legal norms are not ‘things in themselves’, but rather specific kinds of ‘phenomena’, whose objective existence depends on a certain number of cognitive assumptions governing the way in which they are perceived. This implies that, for Kelsen, the law is in the final analysis a product of legal cognition, just like for Kant all objects of perception are the result of certain constitutive features of the way in which the human mind perceives the world.

Once this is made clear, most of the above objections against the theory of the Grundnorm dissolve because it emerges that the Grundnorm need not be interpreted either as a fact or a norm, but is rather posited as one of the necessary presuppositions of legal cognition. Its conceptual status is therefore ‘external’ to the domain of law, but not in the sense in which facts or natural laws are assumed to be external to it. Rather, the Grundnorm is external to the legal order because it is a presupposition jurists must make in order to constitute the object of legal perception to begin with. That is why it is perhaps better to say – relying again on the Kantian analogy – that the Grundnorm occupies a ‘transcendental’ position with respect to the legal order, as a necessary condition for its cognition.

By the same token, however, this interpretation disqualifies the Grundnorm from the possibility of being the foundation for the validity of the legal order as such. For, if the Grundnorm is the condition of possibility of legal cognition, it cannot also be an object of that cognition. Thus, from the perspective of legal cognition, the ultimate foundation for the validity of legal norms cannot be the Grundnorm. This may suggest that the ‘transcendental’ reading of Kelsen’s legal theory simply displaces the problem of the ultimate foundation for legal validity, without solving it. For, even from the perspective of legal cognition, one may still ask what is the foundation for the validity of the ‘first’ or ‘highest’ positive norm within the legal order.

Indeed, as I pointed out above, this is a question which Kelsen himself seems to invite his readers to ask, since when he introduces the notion of the Grundnorm he often does so by following the regressive movement implicit in his conception of legal validity up to a point where he claims legal validity must simply be presupposed. Even after having clarified that the Grundnorm ought not really be situated at the ‘summit’ of the system of norms, but rather as a ‘transcendental’ presupposition of its cognition, there may therefore appear to remain the problem of explaining what is situated at the summit of the hierarchy of norms, from within the perspective opened up by the ‘transcendental’ reading of the Grundnorm.

What I would like to suggest in what follows is that Kelsen’s texts contain the resources for solving this puzzle, even though Kelsen himself did not really clarify this point explicitly, and indeed often presented his theory in a misleading way, that actually makes it impossible to solve. ²⁷

III Time as a necessary element of legal validity

The central intuition on which the specific reading of Kelsen I would like to propose is based is that the question concerning the ultimate foundation for legal validity implicitly depends on a theologico-political conception of the legal order as an entity created ex nihilo, through a gesture analogous to that through which God is supposed to have created the universe. For, it is only from the perspective of that ‘original moment’ that the question of foundations seems problematic. This is especially the case within the logic of Kelsen’s legal theory, since it posits that legal validity depends on norms being created in conformity with procedures stipulated by prior norms. From the perspective of a theologico-political conception of the legal order, this makes the problem of the ultimate foundation for legal validity insoluble, because it effectively implies that what Kelsen would be attempting to construct is a political theology without God.

Unsurprisingly, such a reading reproduces all the problems associated with the ‘paradox of constitutionalism’. For, it either leads to an endless regress or to the necessity of positing an ‘absolute’ that breaks with the logic of system, and must therefore be understood as either fact or norm. Making this clear, however, also indicates the way for a possible overcoming of the problem, because it suggests that the condition for doing so is to abandon the theologico-political conception of the legal order arbitrarily assumed to begin with. My contention is that Kelsen’s writings contain the resources for doing that, in the logical implications of the notion of ‘nomo-dynamicity’. As I already pointed out above, this notion refers to a conception of the legal order as a system that ‘regulates the mechanism for its own production’. ²⁸ This implies on the one hand that Kelsen understands legal norms as creative principles, in the sense that they define the procedures through which further norms can be created; and on the other hand that the foundation for their validity always lies in a previous norm, which authorizes their creation.

The reason why taking this idea seriously provides the means for abandoning a theologico-political conception of the legal order as an entity created ex nihilo through a sovereign gesture analogous to that through which God is supposed to have created the universe is that it implies that legal validity must be understood as an essentially temporal category, in the sense that legally valid norms must always necessarily have a past (since they have to be created in conformity with procedures stipulated by prior norms) and also open the horizon for a possible future (in the sense that they define the procedures for the creation of other norms). From this it follows that the problem of the beginning – or the ‘origin’ – of legal validity cannot emerge from the perspective of legal cognition, because the beginning is strictly speaking not situated within time, but rather outside it or above it.

Indeed, this is something that the theology of creation makes very clear: the act through which God is supposed to have created the universe is not situated conceptually in time, so it makes no sense to ask what happened before that. Time as duration is something that belongs only to the world as already created, thus it is only from the perspective of creation that the idea of a beginning makes sense. ²⁹ Analogously, it is only from the perspective of a theologico-political conception of the legal order as an entity created ex nihilo through a sovereign gesture analogous to that through which God is supposed to have created the universe that the problem of the beginning – or the ‘origin’ – of legal validity can emerge. This is not, however, the perspective that results from taking the notion of ‘nomo-dynamicity’ seriously. For, if we assume that the distinctive aspect of legally valid norms is that they are ‘nomo-dynamic’, it follows that they must always have a past and a future, and therefore that the idea of a beginning or an ‘origin’ cannot belong to the domain of legal validity.

In other words, what I am claiming here is that Kelsen’s conception of legal validity implies that its beginning or ‘origin’ is strictly speaking not a possible object of legal cognition. Thus, the attempt to regress back to an ‘original moment’ in which the legal order would have been created through a sovereign gesture analogous to that through which God is supposed to have created the universe cannot but run into a category that is entirely foreign to political theology, but well known to phenomenology: that of an ‘always-already’. ³⁰ To be sure, this does not imply that from an ‘external’ – i.e. sociological or historical – point of view there cannot exist moments of rupture in which certain norms or constitutions, which are subsequently treated as valid, are posited without obeying the procedures stipulated by any prior norms. However, from such an ‘external’ point of view, what becomes incomprehensible is the idea of legal validity itself, because history and sociology can only treat norms as facts, i.e. not as legally binding entities. Thus, even from this point of view, the attempt to regress back to the beginning or ‘origin’ of legal validity proves to be incoherent.

To the extent that the problem of the ultimate foundation of legal validity emerges only from the perspective of such an ‘original moment’, this reveals that it is actually a false problem. More precisely, what the above analysis reveals is that the problem of the ultimate foundation for the validity of legal norms is the result of an error, or at least a confusion, consisting in the conflation of two categorically different perspectives: the ‘sociological’ or ‘historical’ perspective – from the point of view of which the notion of a beginning makes sense, but that of legal validity is incomprehensible – and the perspective of legal cognition, from the point of view of which legal validity is something distinct, but that of a beginning is strictly speaking impossible. Thus, the answer to the question of what lies at the ‘summit’ of the hierarchy of norms in Kelsen’s legal theory is that the question itself is ill-conceived because it is a bit like asking what was there before time. Since ‘before’ is a temporal category, it cannot possibly apply to time itself. For there to be time there must always already be a past. Analogously, for there to be norms there must always already be norms. Thus, it makes no sense to ask what was there before the first legally valid norm was posited.

By extension, these are also the grounds for a possible ‘Kelsenian’ solution to the so-called ‘paradox of constitutionalism’. Such a solution consists in pointing out that the supposed paradox only emerges because it is assumed that the constitution must have been created ex nihilo, through a sovereign gesture analogous to that through which God is supposed to have created the universe. That assumption is, however, inconsistent with the premises from which it is possible to interpret the constitution as a valid legal norm to begin with. Thus, from a ‘Kelsenian’ perspective, there can be no ‘paradox of constitutionalism’, because the attempt to regress back to the ultimate foundation for the validity of the constitution must either run into the ‘always-already’ of legal validity or dissolve its own object by adopting a sociological or historical perspective on the legal order.

Although this is not the principal reasoning Kelsen himself appears to employ in most of the instances in which he addresses the question of the ultimate foundation for legal validity, it is possible to find some textual support for it in several of his writings. First of all, it is worth recalling that, in many of his discussions of the concept of ‘nomo-dynamicity’, Kelsen explicitly affirms that the legal order must be understood as a ‘perpetual process’ whose distinctiveness lies in the specific logic of its internal development. For instance, in a crucial passage of his Pure Theory of Law, he writes that:

The doctrine of the hierarchy of the legal order comprehends the law in motion: i.e. in its perpetually renewed process of self-regeneration. It is a dynamic theory of law, as opposed to a static theory which attempts to comprehend the law without consideration of its creation, only as a created order. ³¹

What appears particularly relevant in this passage, from the point of view of the interpretation I am attempting to put forwards, is the link established between the notion of ‘nomo-dynamicity’ and the conception of the legal order as a ‘perpetually renewed process of self-regeneration’. For, that clearly points towards the idea of a temporal horizon that has no end, from the point of view either of its past or its future. The logical implication is that, strictly speaking, legal validity – in the transcendentally derived sense – cannot have a beginning, because by definition the beginning does not have a past.

Another significant passage from the point of view of the interpretation I am proposing occurs in the section of his General Theory of Law and State entitled ‘Time as a Constitutive Element of the State’, in which Kelsen explicitly reaffirms the centrality of the category of time in his conception of the state, and therefore law, by stating that:

It is a characteristic of the traditional theory of the state that it considers space – the territory – but not time as an ‘element’ of the state. A state exists, however, not only in space but also in time, and if we regard territory as an element of the state, then we have to regard the period of its existence as an element of the state too. ³²

This contrast between an exclusively ‘spatial’ conception of the state and one including the ‘temporal’ dimension clearly corresponds to the opposition between ‘static’ and ‘dynamic’ conceptions of legal validity. However, what this passage adds, or at least clarifies, is that the temporality of the state (and therefore, in Kelsen’s conception, of legal validity) cannot be reduced to a succession of single instances taking place independently of each other. For him, the state – and therefore law – occurs within the temporal arc of a historical ‘period’. This includes the idea of duration, and therefore that of a connection between past, present and future. For, it is only within such a temporal framework that the specific kind of recursive movement that for him is characteristic of law can take place. From a ‘static’ – or purely ‘spatial’ – point of view, the movement whereby the law ‘regulates the mechanism of its own production’ is simply incomprehensible.

Finally, a further confirmation of the fact that Kelsen’s theory of law must be inscribed within the framework of a temporal horizon marked by duration – and therefore continuity – can also be found in his writings on democracy (which, as several commentators have already pointed out, constitute an essential complement to his juridical writings ³³ ). In particular, what is worth pointing out here is that the central argument Kelsen provides to justify the majority principle crucially depends on the assumption that every act of juridical creation is in reality a revision of a previously established legal order.

The core of this argument is the idea that the majority principle ‘maximizes’ the number of individuals who can be considered free within a given social order, in the sense of having consented through their vote to the constraints imposed upon them by law. ³⁴ This, however, is not the case from the standpoint of a hypothetical ‘original moment’ in which the legal order would be created ex nihilo. For, from that perspective, a principle of unanimity would certainly do better than majority rule in maximizing the number of citizens who could be considered free within the framework of the legal order. Conversely, if we suppose that all legislative acts are in reality revisions of a previously existing legal order, Kelsen’s argument does appear valid, because the principle of unanimity would imply that in theory even a single individual could prevent all the others from changing the legal order effectively in place. Thus, it could produce a situation in which more individuals oppose rather than support the legal order that is effectively in place.

From this it follows that, in formulating his argument for majority rule, Kelsen must necessarily be taking for granted that the legal order is to be understood as a ‘perpetually renewed process of self-regeneration’. As I already pointed out above, from this point of view, the idea of an ‘original moment’ in which the juridical order would have been created ex nihilo is simply incomprehensible. Indeed, this becomes explicit when Kelsen writes that:

The original creation of the social order or of the government is not part of our social experience. The individual is usually born into an already established social order, in the creation of which he did not participate. Thus…only the alteration, the development, of the social order is practically in question. And from this perspective, the principle of an absolute, not a qualified, majority represents the relatively greatest approximation to the idea of freedom. ³⁵

In the light of this passage, the interpretation of Kelsen’s legal theory I am proposing here can be read as a way of tracing one of the fundamental intuitions of Kelsen’s theory of democracy back into his general theory of law. For, there is no reason to believe that what Kelsen is here asserting with respect to democracy – i.e. that ‘only the alteration, the development, of the social order is practically in question’ – should not apply to other kinds of legal orders as well. If this is accepted, the logical implication must be a revision of the status normally assigned to the Grundnorm within the framework of Kelsen’s legal theory. For, if from the point of view of legal cognition there can be no ‘original moment’, then the foundation for the validity of legal norms cannot be assumed to reside in a ‘basic norm’ situated conceptually above and beyond the legal order itself. The Grundnorm must rather be interpreted ‘transcendentally’, as defining the conditions under which legal validity makes sense. And, from that perspective, ‘nomo-dynamicity’ must appear as an essential component of legal validity as well.

This interpretation of Kelsen’s legal philosophy also draws on a number of prior readings of his work, while at the same time complementing them and challenging some aspects I still find problematic. Before concluding, it may therefore be worthwhile to examine briefly some of these prior readings, not only to provide further confirmation for the central hypothesis of this article, but also to clearly mark out its originality.

First, in an article entitled ‘Critique of Dualism: Hans Kelsen and the Twentieth Century Revolution in International Law’, ³⁶ Hauke Brunkhorst has suggested that the truly ‘revolutionary’ aspect of Kelsen’s legal theory is its critique of the ‘metaphysical dualism’ between the law and the state (but also, more deeply, between appearance and reality) that characterized all prior legal philosophy. In place of this dualism, for Brunkhorst, Kelsen posited a ‘continuum between different spheres’, which leaves room for reciprocal interference. This leads Brunkhorst to suggest that Kelsen was able to overcome the problem of the ultimate foundation for legal validity by ‘transforming the dualism of legislative will and executive performance, of political generation and professional application of legal norms…into a continuum of concretization that never ends but goes on and on in a circle’. ³⁷

On this basis, echoing a point that has already been advanced by Jochen von Bernstorff to the same effect, ³⁸ Brunkhorst claims that:

From the very beginning Kelsen’s critique of dualism goes beyond the conceptual scope of neo-Kantianism and signifies a turn similar to and simultaneous with the pragmatic turn in philosophy that was performed by thinkers as different as John Dewey, the young Martin Heidegger, the young Max Horkheimer, or the young Lukacs, and later Quine, Austin, Kuhn, Rorty, Gadamer, Davidson, Putnam, Brandom, Habermas, Bernstein and others. ³⁹

This reading has in common with mine the idea that Kelsen must be understood as a truly post-foundational thinker, in the context of a broader philosophical revolution that took place over the course of the 20th century (a point I will return to in more detail in the conclusion of this article). There remains, however, an important conceptual difference between Brunkhorst’s and Bernstorff’s respective readings of Kelsen and mine: whereas both Brunkhorst and Bernstorff explicitly attempt to carry Kelsen’s legal philosophy beyond the neo-Kantian conceptual framework in which the Austrian jurist had himself explicitly inscribed it, my reading remains predicated on a strictly ‘transcendental’ interpretation of the conceptual status of the Grundnorm. This is significant because it implies that, under my interpretation, there remains a fundamental element of ‘dualism’ at the heart of Kelsen’s theory of law: the dualism between facts and norms, which is conceptually analogous to the distinction between ‘things in themselves’ and ‘phenomena’ in Kant’s critical philosophy.

As I have sought to point out above, such a ‘transcendental’ reading of the Grundnorm is essential to overcome the various objections that have been raised against it by the commentators I examined, because if it is not interpreted as a ‘transcendental’ condition of possibility for legal cognition, there remains the problem of establishing what is the conceptual status of the Grundnorm, which in turn raises the spectre of ‘realism’ (if it is interpreted as a fact) or natural law doctrine (if it is interpreted as a norm). This is reflected in the fact that in his article Brunkhorst is ultimately forced to recommend ‘cutting off’ the notion of the Grundnorm entirely, in order to arrive at a purely ‘hermeneutic’ conception of the legal order as an ‘auto-poietic’ system of norms. ⁴⁰ What Brunkhorst does not seem to realize, however, is that this leaves open the question of establishing what lies ‘at the start’ of the hermeneutic process of ‘concretization’ whereby, under his interpretation, the law is said to ‘constantly re-interpret itself’.

For, if the legal order is not understood as a product of legal cognition (and therefore as belonging to a domain that is categorically distinct from that of facts) but simply as a ‘hermeneutic process’ existing on a continuum with other social spheres, there remains the problem of accounting for its ‘origin’, since interpretation may be open-ended with respect to its goal (i.e. with respect to the temporal dimension of the future) but still supposes a starting point in the past. From this perspective, my reading of Kelsen can be understood as a way of reinscribing Brunkhorst’s (and Bernstorff’s) insight concerning the ‘auto-poietic’ structure of Kelsen’s legal theory within the framework of a neo-Kantian (and therefore irreducibly dualist) interpretation of the conceptual status of the Grundnorm. My contention is that this offers a more persuasive response to the problem of the ultimate foundation for legal validity, than Brunkhorst’s or Bernstorff’s.

The other important source of inspiration for the reading I am proposing is Michel Troper, whose work on Kelsen in turn draws significantly from Norberto Bobbio’s, arguing that since the Grundnorm is posed as an ‘epistemological’ condition for legal cognition, it cannot be interpreted as the ‘ontological’ foundation for the validity of the legal order as such. ⁴¹ On this basis, Troper has suggested an alternative theory of the foundation for the validity of the legal order, according to which such a foundation should not be looked for in an ‘original’ moment or norm, but rather in the logic through which legal norms are applied. In particular, by relying on an analysis of the category of ‘legal interpretation’ Troper has suggested that legal norms derive their ‘meaning’ – and therefore in the final analysis also their ‘validity’ – from the constraints that are imposed on the way in which judges can interpret them. ⁴²

As has been pointed out by Pierre Brunet in an article comparing Kelsen’s and Troper’s respective legal theories, this ultimately leads to a conception of the legal order as ‘both static and dynamic’, because legal validity ceases to point merely ‘upwards’ to the prior norms stipulating the procedures for creating a given norm, and also begins to indicate ‘downwards’ towards the range of possible interpretations to which specific norms are amenable. ⁴³ The interpretation of Kelsen’s theory I have been seeking to put forwards brings Troper’s intuition in the opposite direction. Instead of introducing an element of ‘staticity’ in Kelsen’s theory of law, it radicalizes the idea of ‘nomo-dynamicity’ by applying it to the ‘transcendental’ conditions of legal cognition, thereby arriving at the claim that the very idea of an ultimate foundation for legal validity is the result of a misconception.

Conclusion

In order to begin spelling out some of the broader implications of this interpretation of Kelsen’s legal theory, by way of conclusion, I will return to the philosophical issue mentioned at the start of this article, concerning the search for an ultimate foundation of normativity in general, and the apparent need to posit a notion of the ‘absolute’ to solve this problem.

The above analysis has revealed that this apparent need is in reality the result of an essentially theologico-political conception of normativity, according to which the latter stems from a sovereign gesture analogous to that through which God is supposed to have created the universe ex nihilo. From this perspective, the problem of the foundation of normativity is indeed insoluble without positing a notion of the ‘absolute’, because that would effectively amount to a political theology without God.

However, construing normativity as something essentially temporal provides a means for overcoming this conundrum, because it implies that the idea of an ‘original moment’, from the point of view of which the question of the ultimate foundation is posed, is necessarily exterior to the domain within which the notion of normativity makes sense. Thus, construing normativity as something essentially temporal opens the way for a truly post-foundational account of its conditions of possibility.

Through further analysis, this could perhaps be used to demonstrate the incoherence of the very notion of political theology, as defined by Schmitt. For, if ‘secularization’ is understood as the entry into the domain of ‘secular’ time – that is, the time of the saeculum, which is marked by duration ⁴⁴ – it follows that the idea of a God implicit in that of ‘political theology’ cannot be a part of it, because God (and therefore the ‘absolute’) is situated by definition outside the saeculum.

In turn, this might lead to a different reading of the notion of modernity itself. For, if modernity is to be understood as a Neuzeit – that is, as a new temporality, marked precisely by its ‘secularization’ ⁴⁵ – it follows that it must necessarily also be understood as a domain from which the reference to an ‘absolute’ or ‘transcendent’ source of normativity is excluded. Inverting the title of another famous book on this topic, we could therefore say that it is not true that ‘We have never been modern’. ⁴⁶ Rather, modernity is something that can exist only in the temporality of an ‘always-already’.