• 1 Introduction

  Legal positivists are often accused of being unable to avoid the contradiction that follows from the attempt to formulate a theory of law which tries to adequately preserve and theoretically capture the normativity of the law, yet at the same time takes the law’s content and existence to depend entirely on fact and not on its merit. Our first intuition in relation to such a theory might be that it cannot be very promising. After all, how could anything ever be normative irrespective of its merit? If the law’s content and validity are dependent entirely on fact, how can it ever be or claim to be normative in a more than figurative sense? It takes a second glance to find a promise of coherence in such a positivistic enterprise, and this promise is mostly found in taking positivism to steer a middle way between reductivist realism and natural law theory. Yet on a third glance, to most, everything turns back into nonsense again.

  To Anglo-American readers Kelsen’s Pure Theory is, next to Hart’s positivism, the most famous of these attempted middle ways, a way the attractiveness of which cannot stand a third glance. The various critical attacks levelled against the Pure Theory concur in the claim that even if we tidy up all the versions and formulations of it, even if we bring together all the different doctrines and iron out surface inconsistencies, what we are left with is a theory which on its most favourable reading carries a fundamental contradiction at its core. Even authors, who, like Stanley L. Paulson, can hardly be accused of trying to misunderstand Kelsen, cannot shake off the feeling that Kelsen is ‘running off in two different directions at once.’¹xStanley L. Paulson, ‘The Weak Reading of Authority in Hans Kelsen’s Pure Theory of Law,’ Law and Philosophy 19 (2009): 131-71, 167. Antonio Bulygin famously referred to a fundamental ‘antinomy’ in Kelsen’s work.²xEugenion Bulygin, ‘An Antinomy in Kelsen’s Pure Theory of Law,’ Ratio Juris 3 (1990): 29-45. If we take these to be the last words on the Pure Theory, then it is hard to see how the latter can ever be more than a partially interesting political theory supported by a set of applaudable political and moral convictions and how it could hardly ever be worthy of the serious academic engagement it actually attracts.

  What is barely ever considered is the possibility that the contradiction stems from the very idea that Kelsen tries to steer a middle way. After all, it is hard to see how one could drive a middle way between moralism and reductionism without entering a lazy compromise, without ignoring the contradiction that such a compromise necessarily involves. However, what if the Pure Theory, in contrast to Hart’s positivism, never attempted a middle way? Put differently, what if the Pure Theory attempted a ‘middle way’ between reductionism and moralism only in the way in which Kant drove a ‘middle way’ between rationalism and empiricism, that is, no real middle way at all? After all, the Kantian solution to the great philosophic impasse has not been conciliatory, but radical: what Kant attempted was not to reconcile two opposed world views by creating a syncretistic compromise of both, but to turn everything upside down and to submit the very possibility of us relating to the world to a fundamental reassessment. Kant did not introduce a compromise, but a Revolution der Denkungsart, an intellectual revolution.

  Now, this paper suggests that Kelsen indeed had a fundamentally Kantian solution to the jurisprudential impasse and that when we actually do read Kelsen with Kant, we can see that the heart of the Kelsenian teaching does not lie in simply proposing a new concept of law, but rather in rethinking the possibility of us relating to the law qua possible object of knowledge and in submitting this possible relation to a revolutionary overturn. Kelsen tried to solve the problem of the ontology of the law (‘What is law?’) by translating it into a revolutionary, a ‘transcendental,’ epistemology (‘How can we relate to the law qua possible object of knowledge and how do we have to conceive of a law to which we can actually relate?’). It turns out that our implicit, everyday epistemology of our relation to the law, even though held dear by common sense, actually is incoherent, impossible and ultimately has to be jettisoned.

  Now, even though such a radically Kantian approach to Kelsen is hardly new or revolutionary, not much has been made of it so far and the greatest intellectual effort has been spent on the previously devised, un-Kantian, model of the middle way, even though this model is neither very challenging intellectually nor promises many chances of success.

  The present reading starts with the assumption that the Pure Theory can only hope to be successful if it does not try to steer a middle way between reductionism and moralism, but if it tries to overcome the opposition by making clear that both opponents of the opposition rest on the same ill-conceived convictions about legal validity. Both take it that the law cannot be normative from itself. Here natural law theory and reductionism agree. By correcting this conviction the Pure Theory does not present a middle way between the suppositious opponents, but it presents a true alternative to a spurious alternative; it does not present a third way, but an actual second way.

  Kelsen’s actual solution starts with the demonstration that moralism and reductionism are both half-truths, the strengths of which cannot simply be added in order to get a full truth. Rather, Kelsen demonstrates, quite in line with the Kantian resolution of the antinomy, that both theses are expressions of the same erroneous approach to the law. Both take the law to be derivative of something else.

  In contrast to this, the Pure Theory tries to find a new approach to the understanding of the law, an approach that takes seriously the constitutive functions of the law. It tries to understand the validity of the law as resting in the law itself. As such it is an attempt to find a philosophically satisfactory formulation of absolute positivism.

• 2 Absolute Positivism

  If, for now, we identify positivism with the thesis that

  (R₁) In any legal system, whether a given norm is legally valid does not depend on its merits³xGardner’s definition runs: ‘(LP) In any legal system, whether a given norm is legally valid, and hence whether it forms part of the law of that system, depends on its sources, not its merits.’ However, he also claims that ‘“source” is to be read broadly such that any intelligible argument for the validity of a norm counts as source-based if it is not merit-based.’ Gardner takes the two categories source and merit to be ‘jointly exhaustive of the possible conditions of validity of any norm.’ However curious such a claim or stipulation may be, it does allow us to omit as redundant the sources element from the definition of positivism. See John Gardner, ‘Legal Positivism: 5 1/2 Myths,’ The American Journal of Jurisprudence 46 (2001): 199-227.

  or with

  (R₂) The validity and the content of the positive law cannot be derived from moral premises

  then positivism presents itself as a negative or relative position, i.e., as the rejection of certain normative relations of derivation. Positivism, understood in this way, first and foremost tells us what the law is not. We will call this kind of positivism ‘negative,’ or ‘relative positivism.’⁴xFor an instructive typology and historical discussion of positivism and its schools see Brian Bix, ‘Legal Positivism,’ in Blackwell Guide to the Philosophy of Law and Legal Theory, ed. Martin P. Golding & William A. Edmundson (London: Blackwell, 2005), 29.

  However, the question is not only what the law is not. It is also, what the law is. Now, a provisional, working definition of absolute positivism emerges when we try to answer this positive question without abandoning our commitment to relative positivism. What would such an answer look like? All that is left to base the validity and content of the law on would be the law itself. A theory that tries to base the validity of the law on the law itself or, which is the same, on nothing, could be called positive, or absolute positivism:

  (A₁) In any legal system whether a given norm is legally valid depends entirely on the law.

  Or:

  (A₂) The validity and the content of the positive law is based in the law itself, or, which is the same: on nothing.

  All we have done here is to make explicit what is already present in R₁ and R₂. However, since contemporary Anglo-American positivism does not want to upset common sense, it is in its essence relative positivism. It mainly consists in not making explicit the radical thought already implicit in relative positivism. Its main claim thus is that all law is ‘source based.’ However, it does not draw the conclusion or make explicit that by maintaining that the law is ‘source based’ it has already claimed that the law is law-based or, which is the same, that the law is, in an important respect, base-less.

  One way to block insight into the radicality of positivism is to divorce questions about the criteria of membership from questions about the criteria of the authority or bindingness of the law. Such a move is meant to delegate all quirky questions of normativity to a theory different from the sources thesis and to thus take explanatory pressure off the latter. However, it is hard to see how this could solve rather than exacerbate the problem: (1) divorcing the two questions simply adds another pressing question to an already pressing one; (2) by divorcing the question of normativity from the question of membership one has admitted that the normativity of the law cannot be a function of the law qua law, but that it has to come from somewhere else, be it from a comprehensive moral theory (Finnis) or teleological relations to right reasons (Raz); by doing that, however, one has admitted that the real theoretical weight is lifted by that other theory and that the law just plays a certain role in what is fundamentally a moral theory;⁵xSee Philip Soper, ‘Some Natural Confusions About Natural Law,’ Michigan Law Review 90 (1992): 2393-2423. I do not think that Raz’ practical difference thesis, i.e., the claim that in order to be authoritative the law has to purport to make a practical difference by excluding or pre-empting appeal to dependent reasons, which include first-order moral reasons, does get us very far here, since the question whether the law actually does make a practical difference has, in turn, to be decided by moral reasons. See Joseph Raz, The Authority of Law (Oxford: Clarendon Press, 1979), ch. 3. See also Scott Shapiro, ‘The Difference That Rules Make,’ in Analyzing Law: New Essays in Legal Theory, ed. Brian Bix (Oxford: Clarendon Press, 1998), 39 and Jules Coleman, ‘Incorporationism, Conventionality, and the Practical Difference Thesis,’ Legal Theory 4 (1998): 381. this unwittingly turns positivism from a comprehensive legal theory into an accidental element of a moral theory; (3) finally, in however manner the problem of bindingness is ultimately thought to be solved (and I share Kelsen’s view that it cannot be solved in separation from the question of membership) it still leaves the problem of the baselessness of the criteria of membership unsolved.

  In contrast to that, Kelsen’s Pure Theory is at least partially the attempt to make philosophical sense of absolute positivism and to bite the bullet of the challenges of common sense.

  As its title already suggests, the Pure Theory demands purity. This means that it deals with the positive law and the positive law alone. Thus it cannot consider any principles beyond the law and it accordingly has to answer the question of the validity of the law from the law itself. In this attempt to perform the seemingly impossible lies the essence of the Pure Theory and in this attempt it is more programme than completed doctrine.

  Of course, at first sight, the programme to base the validity of the law in the law itself must seem futile. How should the law ever be able to ground and constitute itself? What could this even mean?

  Kelsen answers this question indirectly, via the detour of a sceptical argument about the derivation of validity in general. After all, it is not only absolute positivism that struggles with the derivation of validity. Rather, when trying to account for legal validity, relative positivism faces problems of its own kind, problems which, when attended closely, turn out not to be mere obstacles, but impossibilities of a principled kind.

  Relative positivism faces the following well known, but still fundamental problem: assuming that no legal norm is valid in itself, that no legal norm is valid because of its content, but that it has to be posited in order to be valid, then each legal norm depends for its validity on another legal norm, since being posited means being lawfully posited, which in turn means being posited according to a valid legal norm. Now, if the validity of every legal norm necessarily depends on the validity of another legal norm, then we can demonstrate, in line with the classic pyrrhonic argument, that there can exist no valid legal norms at all, since the resulting chain of derivations of validity, just as any chain of derivations, leads to the following trilemma:⁶xHans Albert, Traktat über kritische Vernunft (Tübingen: J.C.B. Mohr, 1991), 15. we either get (1) an infinite regress, or (2) a logical circle or (3) a dogmatic acceptance of certain truths, i.e., the abandonment of derivation as such.

  Option (3) does not even attempt to present a solution to the problem, but is simply an abandonment of the premise, i.e., of the universal need for derivation, or, in our case, the abandonment of the universal positivity of the law. Any legal norm can be justified by claiming that certain norms are valid in and from themselves.

  According to (1), however, no legal norm at all can be valid: an infinite regress means that a final justification cannot be reached and thus no norm is thoroughly justified.

  Finally, option (2) again allows any legal norm to be valid, since every possible legal norm can be derived from a logical circle.

  The trilemma shows us that with logical necessity the premise of positivism, i.e., the requirement of legal norms to be based in other legal norms, leads to the situation in which either no law or any possible law can be valid. Or, put differently: if there is to be valid law, then anything is law and if not anything is to be law, then nothing can be law.

  Now, contemporary Anglo-American positivism is but a formulation of this paradox. It primarily consists in moving about the horns of the trilemma and in trying to combine circularity and dogmatism, i.e., to combine the methods of validating any law with the method of validating no law in such a way that the specific validity of the law can be explained. Contemporary Anglo-American positivism finds its most comprehensive formulation in the so-called sources thesis, which states that the question of whether a legal norm is legally valid depends on the sources of this legal norm and not on its moral merit.⁷x‘(LP) In any legal system, whether a given norm is legally valid, and hence whether it forms part of the law of that system, depends on its sources, not its merits.’ Gardner, ‘Legal Positivism,’ 199-227. With that, however, we have before us less of a theory of law and more of the formulation of a problem, namely the aforementioned trilemma: how can a norm meaningfully derive its validity from a source, which itself is in need to derive its validity from another source, without either leading into an infinite regress, a circularity or giving up the requirement of derivation?

  The situation is quite clear in Hart and has been discussed many times. Hart takes the regress to find its end in the rule of recognition.⁸xHerbert L.A. Hart, The Concept of Law, 2nd ed (Oxford: Oxford University Press, 1994), 94. In order to escape dogmatism, i.e., in order not to link the capacity of the rule of recognition to end the regress to some special intrinsic quality of the rule of recognition, in some element of its content, he says that the rule of recognition is a rule which is accepted by the organs of the given legal community.⁹xIbid., 114. However, thereby he has escaped dogmatism only at the cost of accepting circularity:¹⁰xFor a recent discussion of this circularity and attempts at its solution see Brian Tamanaha, ‘Socio-Legal Positivism and a General Jurisprudence,’ Oxford Journal of Legal Studies 21 (2001): 1-32 and Keith Culver & Mike Guidice, Legality’s Border (Oxford: Oxford University Press, 2010). the rule of recognition depends for its validity on the acceptance by organs, the organ-character of which in turn depends on legal norms, the validity of which can be traced to the rule of recognition.¹¹xOne might be tempted to think that Kelsen and Hart here deal with different questions: whereas Hart’s rule of recognition mainly offers a solution to the problem of membership, i.e., the question of which norms belong to a legal order, Kelsen’s basic norm wants to answer the question of validity which in his view also includes the question of the binding nature of norms. Now, Hart does, of course, also tackle the problem of validity and bindingness. The main difference is that, in contrast to Kelsen, for Hart it does make sense to look at membership in isolation, i.e., from a purely external point of view, whereas for Kelsen it is pointless to think about membership without also considering validity. What matters for the present discussion, however, is that both do actually deal with both questions and the difference can be disregarded at this stage. See Carlos Santiago Nino, ‘Some Confusions Surrounding Kelsen’s Concept of Validity,’ in Normativity and Norms: Critical Perspectives on Kelsenian Themes, ed. Stanley L. Paulson (Oxford: Oxford University Press, 1999), 253-61.

  With such a circle, everything can be justified: I can, for instance, set up the state of Egopolis, a legal system the rule of recognition of which states that everything I declare law, immediately becomes law. In my legislative function I then declare myself to be the only organ of the state of Egopolis and in my function as organ of the state of Egopolis I accept the rule of recognition. Then I posit further legal norms. The validity of these legal norms can be traced to a rule of recognition, which is valid since it is accepted by the organs of the legal order, i.e., by me. Thus everything I declare to be law thereby is valid law.

  In order to escape the arbitrariness of the grounding of validity in such circles, Hart introduced a further condition of validity: it is not enough for the rule of recognition to be accepted by the organs, the norms flowing from it also have to be by and large obeyed by the subjects, i.e., the legal order has to be effective.¹²xHart, The Concept of Law, 117. And precisely this is not the case in the above outlined state of Egopolis. It might thus seem that the problem of circularity and thus the problem of the validity of arbitrary norms is solved.

  Unfortunately, Hart solved the problem of circularity by introducing a condition which is not itself legally justified and which thus can only apply dogmatically: from within the myriad of rules of recognition which can be legitimized by a circular argument, this shall be valid, which is actually obeyed by the subjects. Or, put differently: for each legal order that is actually obeyed, a circular argument can be found which legitimizes the rule of recognition. Thus as a circular argument the rule of recognition simply provides legitimacy to any given effective legal order, or, to be precise: it provides the semblance of legitimacy. In truth the normativity shall not be conditioned by the rule of recognition itself, but by the effectivity of the legal order. The century old question, how normativity can be derived from effectivity, how validity can be derived from obedience, is not solved but obscured by conjuring up the circular argument.¹³xFor a discussion of this issue see Andrei Marmor, ‘Legal Conventionalism,’ Legal Theory 4 (1998): 509 and Gerald Postema, ‘Coordination and Convention at the Foundations of Law,’ Oxford Journal of Legal Studies 11 (1982): 165. For an attempt to find a solution via efficacy see Gerald Postema, ‘Conformity, Custom, and Congruence: Rethinking the Efficacy of Law,’ in The Legacy of H.L.A. Hart: Legal, Political, and Moral Philosophy, ed. Matthew H. Kramer, Claire Grant, Ben Colburn, & Antony Hatzistavrou (Oxford: Oxford University Press, 2008), 46.

  Hart does not solve the problem of the specific validity of the positive law, but he veils it by switching between circularity and dogmatism in order to escape an infinite regress.

  Theories that add moral elements, like Dworkin’s interpretivism or Finnis’ natural law theory, at least in the respect discussed here, do not fare much better. Rather, the problem of justification of validity of the law is partly obfuscated, partly exacerbated by the addition of moral norms, which are themselves in want of justification.

  Now, the trilemma argument does not simply state that it is problematic or difficult to find validity in chains of derivation, but it demonstrates that it is logically impossible to derive validity at all. Note the strengths of the claim: derivation of validity is logically impossible. This is the only true absolute result that relative positivism can provide.

  Now, the curious twist of absolute positivism and the Pure Theory is to base its own constitution of validity on this logical impossibility. It is, so to speak, only in the transit through absolute scepticism about validity that the necessity of absolute positivism becomes clear. Whereas modern Anglo-American positivism despairs over the logical acuteness of this scepticism, the Pure Theory, insofar as it is an attempt to find a formulation of absolute positivism, takes the strictness of this sceptical insight as its origin. The truth of relative positivism lies in the insight that there can be no validity when one hopes to derive validity, when one wants to find validity in an external relation of one norm to another. To overcome relative positivism and enter absolute positivism all one has to do is to give up this belief that the validity of one norm can be found in an external relation it has to another norm.

  The argument runs parallel to the overcoming of Agrippa’s trilemma in epistemological debates about the impossibility of justification in general. In that context the sceptical argument claims that no belief can be justified since any belief can be justified only by another justified belief, which again leads us into the known trilemma. Now, the classical Kantian way out of this trilemmatic structure has been to rethink our understanding of what a relation is:

  ‘What is negated in all three arguments of the trilemma of justification is the nativeness and independence of the relation. To the ontological bias of our realist common sense the relation is but the empty and variable space between the actual things, between the substances, between the empirically given entities. In contrast to this thing-ontological dogmatism, scepticism correctly draws from all this the conclusion, that our cognition of the actual existing things is variable and fundamentally empty.’¹⁴xKurt Walter Zeidler, Grundriß der Transzendentalen Logik (Cuxhaven: Traude Junghans Verlag, 1997), 149.

  As long as we stick with our realist common sense, i.e., the idea that what is actual are the things and that the relation between these things is only contingent, secondary and external to them, then we should not be surprised that our relation to these things also turns out to be contingent and ultimately external to these things. It is thus our realist ontological commitments tacitly implicit in our common sense, which lead to the epistemological impasse of the impossibility of knowing anything. Consequently, it is in the name of having a promising epistemology, i.e., in the name of being able to know anything at all, that we have to criticize, rethink and ultimately change our ontological commitments. Ultimately, this means that we have to take the relation to be the actually existing entity and the ‘thing’ to be the secondary entity. This is what the Kantian reformulation of the object from being a ‘thing in itself’ to being a ‘phenomenon’ or, which is the same, a ‘thing for us’ actually means.¹⁵xSee Henry Allison, Kant’s Transcendental Idealism. An Interpretation and Defence (New Haven: Yale University Press, 1983), 29.

  Now, in facing the trilemma jurisprudence has to demand from juristic common sense to go through a similar critique and ultimately overthrow of its implicit ontological commitments: it demands a reversal of the conviction that the relation is secondary and external to the relata to the conviction that the relation is fundamental and primary.

  In the legal context this means the following:¹⁶xFor some reasons why the transcendental argument looks slightly different in the legal context than it does in the context of theoretical philosophy, see Christoph Kletzer, ‘Kelsen, Sander, and the Gegenstandsproblem of Legal Science,’ German Law Journal 12 (2011): 785-810. relative positivism takes the law to be a sum, an aggregate of legal norms, the relation of which is external to these legal norms. Its fundamental thesis, the sources thesis, states that a norm derives its validity from another positive norm. However, derived it must be! Absolute positivism, in contrast, starts with the insight that the impossibility of establishing legal validity by means of derivation stems from the separation of the law into a legal norm, on the one hand, and the grounding of validity, on the other. Absolute Positivism, in contrast, accepts that the law is not a sum of legal norms which are then, successively, somehow related to each other, but rather that the positive law itself is legal relation. The law is not a collection of norms, but the relation of these norms, i.e., it is the creation, justification and application of norms. To say that the positive law actually is legal relation is simply to say that the law is legal process.

  To claim that the law is legal process means that the problem of the validity of a legal norm is not a philosophical or jurisprudential problem, but a legal problem to be solved not by legal theory but by the law itself.¹⁷xThis is not Raz’s point, that legal propositions are internal to a legal discourse and do not make sense outside of such a discourse or Hart’s point that questions about what ‘legally’ ought to happen only makes sense relative to a discourse constituted by the rule of recognition. It is not about a discourse at all. It is a claim about the actual legal process, i.e., about what legislators, judges, legal officials, and, ultimately, legal subjects do and how an interpretation of these actions is schematised by legal rules. For absolute positivism validity is not to be determined by legal theory, but by positive law itself.¹⁸xNino’s helpful distinction between ‘rules’ and ‘judgements of validity’ leading to two different chains of derivation takes the latter ‘judgements of validity’ to be ‘typically formulated by jurists.’ With ‘jurists’ he presumably has academics in mind. The authentic and thus truly relevant ‘judgments of validity,’ however, do not happen in an academic but in a legal context in the legal process itself, i.e., as judgements (in Nino’s sense) done by parliamentarians (judging that the constitution is valid), by judges (judging statutes and constitutions to be valid), by executive forces (judging court orders, statutes and constitutions to be valid), and by legal subjects (judging executive commands, court orders, statutes and constitutions to be valid). See Nino, ‘Some Confusions,’ 257. One is tempted to adduce the unlikely support of Dworkin here, who claimed that jurisprudence was the general part of adjudication, i.e., that the task which we believe is to be fulfilled by legal philosophy, is actually always already fulfilled by the law itself.¹⁹x‘Jurisprudence is the general part of adjudication, silent prologue to any decision in law,’ in Law’s Empire, Ronald Dworkin (Cambridge: Harvard University Press, 1986), 90.

  Absolute positivism thus solves the problem of the validity of law by declaring that this problem has already been solved by the law itself.²⁰xIn the context of Bruno Celano’s theory of validity as disquotation this means that we have to locate the activity of disquotation in the positive law itself (Bruno Celano, ‘Validity As Disquotation,’ Analisi e diritto (1999): 35-77). Here, with ‘disquotation’ the following is meant: just as in a minimalist theory of truth the trivial criterion of truth can be given as ‘“p” is true iff p,’ or ‘“Snow is white” is true iff snow is white,’ allowing to disquote sentences, to use a sentence which was formerly only mentioned, to make p out of ‘p,’ in Celano’s disquotational theory of validity a similar relation applies: ‘the norm “q” is valid iff q,’ or ‘The norm “Children ought to obey their parents” is valid iff children ought to obey their parents.’ Now, Celano thinks that the disquotational statements are statements of meta-jurisprudence, and that ‘validity’ is a device in the vocabulary of ‘a substantive ethical theory purporting to specify the conditions under which the law, or particular legal norms, ought to be obeyed’ (235), i.e., it belongs to the vocabulary of an inquiry into what the law ought to be and ‘not to the vocabulary of a scientific description of positive law as it actually is. This latter is, however, precisely the use to which validity as disquotation has been put, in the Pure Theory of Law, by Kelsen himself. In the Pure Theory of Law, a conceptual, necessary relation holds between positive law, on the one hand, and validity (validity as disquotation) on the other hand: positive legal norms are, as such, binding’ (236). In Kelsen’s work the concept of disquotation, conversely, rightly features as a technique not of meta-ethics or legal science but of the positive law itself. Of course, for Kelsen positive legal norms ‘as such’ can never be binding since positive legal norms ‘as such,’ i.e., irrespective of their relation to other legal norms, separated from the legal process, are not positive legal norms. In contrast to what Celano writes, Kelsen’s criterion of binding force is this: ‘“p” has binding force, if another norm q as a scheme of interpretation allows interpreting it as having binding force.’ Or: ‘p’ has binding force if another norm q disquotes ‘p.’

  In identifying validity as a problem to be solved by legal theory, relative positivism implicitly claims a final competence in relation to the determination of legal validity. It is unaware that in doing that it actually arrogates a legal competence, a competence, however, which is at odds with its own doctrine of the positivity of the law. In relative positivism, positivism thus recoils back at itself: the argument rightfully directed at natural law theory, i.e., that one must not confuse philosophical content with legal content and that the moral philosopher does not have competence to posit laws, also cuts against relative positivism.

  Relative positivism is thus only a semi-positivism, a half-hearted doctrine that is always exposed to a tu quoque claim.

  Absolute positivism, in contrast, makes the precarious attempt not to overstep the limits of its competences when talking about the positive law. After all, that there are competences which are conferred by positive law and which can be overstepped, this is the fundamental positivistic insight. Absolute positivism thus does not try to establish the validity of the law, as the validity of the law cannot be established without giving up the fundamental positivistic convictions. Rather, it lets the law be, it lets it be valid, it lets the mode of validity immanent in the law be itself. It thus defers all competence to the law itself.

  This deferral of the problem of validity of the law to the law itself finds its legal expression in the basic norm.²¹x‘The basic norm is a judgment of validity,’ yet at the same time ‘this fundamental judgement of validity, according to Kelsen, is itself a norm.’ Nino, ‘Some Confusions,’ 258. The basic norm marks the borderline between law and legal theory. As we have seen above, the positive law, if it wants to be valid as positive law, must not depend in its validity on conditions given to it by legal theory, but has to posit the conditions of validity for itself. Now, it does precisely that in presupposing (German ‘voraus-setzen’ or pre-positing) the conditions of validity in the basic norm.²²x‘Auch das Voraussetzen ist ein Setzen, aber ein Setzen, das das Setzen zugleich als aufgehoben setzt. [Presupposing, too, is a positing. It is a positing, however, which posits its own positing as cancelled.]’ Dieter Henrich, Hegel im Kontext (Frankfurt/Main: Suhrkamp, 2010), 120.

  The law presupposes the basic norm. Now, since Kelsen sometimes writes that we can presuppose the basic norm but do not have to presuppose the basic norm,²³xSee Hans Kelsen, Reine Rechtslehre, 2^nd ed (Wien: Verlag Österreich, 2000), 224. one might be under the impression that the presupposition of the basic norm is a psychological act.

  However, the presupposition of the basic norm is not a psychological act but a logical relation. It is not the case that one first has to presuppose the basic norm in order to then say something about the law. Rather one has always already presupposed the basic norm by, for instance, questioning the validity of a putative legal norm, as questioning the validity of a legal norm means asking if amongst the other valid legal norms there is one that authorizes the creation of the norm in question. In asking this question, one has already presupposed the validity of other norms and thus the validity of the basic norm. One cannot ask a legal question, one cannot say or think anything legal without already having presupposed the basic norm. Presupposing the basic norm thus is not a psychological accident but is by logical necessity implied in speaking about the law.²⁴xSee Uta Bindreiter, ‘Presupposing the Basic Norm,’ Ratio Juris 14 (2001): 143-75, 168: ‘It seems that the presupposition of the basic norm was intended, by Kelsen, to evoke a specifically “legal” use of language. Since the linguistic form of norm-formulations (used to issue norms) and of norm-statements (used to “describe” norms) may well be exactly the same, the difference between norm-formulation and norm-statement must lie in the illocutionary force of the respective utterances.’ The law presupposes its own validity. It is thus the law and not we that presupposes the basic norm and presupposing the basic norm the positive law posits itself.

  One can, of course, avoid presupposing the basic norm. However, one can only do so by not speaking or thinking about the law as law. Thus the presupposition of the basic norm is more of an academic and less of a legal problem.

  The presupposition of the basic norm brings us into the internal space of the law. In the presupposition of the basic norm the law tells us what is law and what are legally relevant facts. Viewed from this inside there is no reality outside of the law that is independent of the law yet relevant to it.

  Thus philosophy has nothing to tell the law which the law would not know from itself. If there is to be law then it is by necessity. Put differently: viewed from the point of view of the law the law exists necessarily and the validity of the law leaves open no meaningful questions to be answered by legal philosophy. Viewed from the point of view of philosophy, however, there can be no positive law and no legal validity. Relative positivism is but the law viewed from the point of view of philosophy. The irresolvable paradox it faces is that, for it, there can be no valid law at all.

  So it is the positive law itself that has fulfilled and completed a task which philosophy has set itself and which it necessarily failed at completing, since a law that is in need of a constitution through philosophy cannot be positive law. Relative positivism is a philosophical doctrine which thinks it is a doctrine about the law, but which actually is a doctrine about its own, i.e., philosophy’s, supreme competence, a doctrine, however, which conflicts with its own positivistic commitments. As a theory of the supreme competence of philosophy and reason, relative positivism is actually closer to a natural law theory than it might think. Whereas relative positivism takes reason and its mouthpiece philosophy to be the final arbiter of all formal questions relating to the law, natural law theory takes reason and its mouthpiece philosophy to be the final arbiter of all questions relating to the law. In relation to natural law, relative positivism has given up some competences over the positive law. Absolute positivism surrenders all competences over the positive law.

  Absolute positivism is thus aware that it is a philosophical doctrine about philosophy, about the limits of philosophy. As such a philosophical doctrine about the limits and incompetence of philosophy, absolute positivism to a certain extent has to be both a reflexive and also an anti-philosophical doctrine.²⁵xThe reflexive nature of positivism comes to light as an inner dividedness of positivism. Positivism was always torn between a philosophical and a decidedly anti-philosophical strand. A throughout philosophical positivism is a more recent phenomenon and positivistic tradition is rife with strong anti-philosophical tendencies. There is, for instance, the common opinion that the philosophical reflection of the law relates to the positive law as institutes to pandects, i.e., as textbook introduction to the real law. Philosophy can help write textbooks for novices. However, it cannot help us master the complexity of the positive law itself. See Friedrich Carl von Savigny, Vom Beruf unserer Zeit für Gesetzgebung und Rechtswissenschaft (Hildesheim: Georg Olms, 1967), 115.

• 3 Conclusion

  A true positivism thus has to be able to bear these anti-philosophical tendencies within it. Positivism cannot be itself without at the same time being a doctrine about the end of philosophy, about the outpour of philosophy into the world. Insofar as the loss of God has meant that philosophy was promoted from the maidservant (ancilla) to the vicar of theology, we can say that only an absolute positivism can free itself from a mode of thinking that was formed by the paradigms of theology. For the problem with the loss of God is not only that we lose the firm foundations of the divine rules and the secular rules based on these divine rules. Rather the problem is that the loss of God has changed the paradigm of the validity of rules itself. Without God rules hold and apply differently, they are valid in a different manner and sense. And whereas relative positivism privily awaits the revelation of a final validity, awaits and lacks the absolute, absolute positivism has learned the full lesson of the absolute immanence of validity: the world does not lack legitimacy. Neither does the law.²⁶xSee also, Hans Blumenberg, Die Legitimität der Neuzeit (Frankfurt/Main: Suhrkamp, 1999).

  This paper tried to read the Pure Theory as an attempt to formulate an absolutely positivistic theory of law. It tried to move Kelsen away from the middle way and to understand his theory as a philosophical grappling with the radical self-constitution of legal validity.

  However, since the Pure Theory does not only attempt to deal with legal validity, but also tries to present a programme of legal scientific work and since absolute positivism with its anti-philosophical tendency is an unpopular position, the proposed reading is of course precarious. It is, however, both promising and much needed, since it defends the Pure Theory against the charge of inconsistency and tries to formulate a coherent idea of the Pure Theory.

Noten

• 1 Stanley L. Paulson, ‘The Weak Reading of Authority in Hans Kelsen’s Pure Theory of Law,’ Law and Philosophy 19 (2009): 131-71, 167.

• 2 Eugenion Bulygin, ‘An Antinomy in Kelsen’s Pure Theory of Law,’ Ratio Juris 3 (1990): 29-45.

• 3 Gardner’s definition runs: ‘(LP) In any legal system, whether a given norm is legally valid, and hence whether it forms part of the law of that system, depends on its sources, not its merits.’ However, he also claims that ‘“source” is to be read broadly such that any intelligible argument for the validity of a norm counts as source-based if it is not merit-based.’ Gardner takes the two categories source and merit to be ‘jointly exhaustive of the possible conditions of validity of any norm.’ However curious such a claim or stipulation may be, it does allow us to omit as redundant the sources element from the definition of positivism. See John Gardner, ‘Legal Positivism: 5 1/2 Myths,’ The American Journal of Jurisprudence 46 (2001): 199-227.

• 4 For an instructive typology and historical discussion of positivism and its schools see Brian Bix, ‘Legal Positivism,’ in Blackwell Guide to the Philosophy of Law and Legal Theory, ed. Martin P. Golding & William A. Edmundson (London: Blackwell, 2005), 29.

• 5 See Philip Soper, ‘Some Natural Confusions About Natural Law,’ Michigan Law Review 90 (1992): 2393-2423. I do not think that Raz’ practical difference thesis, i.e., the claim that in order to be authoritative the law has to purport to make a practical difference by excluding or pre-empting appeal to dependent reasons, which include first-order moral reasons, does get us very far here, since the question whether the law actually does make a practical difference has, in turn, to be decided by moral reasons. See Joseph Raz, The Authority of Law (Oxford: Clarendon Press, 1979), ch. 3. See also Scott Shapiro, ‘The Difference That Rules Make,’ in Analyzing Law: New Essays in Legal Theory, ed. Brian Bix (Oxford: Clarendon Press, 1998), 39 and Jules Coleman, ‘Incorporationism, Conventionality, and the Practical Difference Thesis,’ Legal Theory 4 (1998): 381.

• 6 Hans Albert, Traktat über kritische Vernunft (Tübingen: J.C.B. Mohr, 1991), 15.

• 7 ‘(LP) In any legal system, whether a given norm is legally valid, and hence whether it forms part of the law of that system, depends on its sources, not its merits.’ Gardner, ‘Legal Positivism,’ 199-227.

• 8 Herbert L.A. Hart, The Concept of Law, 2nd ed (Oxford: Oxford University Press, 1994), 94.

• 9 Ibid., 114.

• 10 For a recent discussion of this circularity and attempts at its solution see Brian Tamanaha, ‘Socio-Legal Positivism and a General Jurisprudence,’ Oxford Journal of Legal Studies 21 (2001): 1-32 and Keith Culver & Mike Guidice, Legality’s Border (Oxford: Oxford University Press, 2010).

• 11 One might be tempted to think that Kelsen and Hart here deal with different questions: whereas Hart’s rule of recognition mainly offers a solution to the problem of membership, i.e., the question of which norms belong to a legal order, Kelsen’s basic norm wants to answer the question of validity which in his view also includes the question of the binding nature of norms. Now, Hart does, of course, also tackle the problem of validity and bindingness. The main difference is that, in contrast to Kelsen, for Hart it does make sense to look at membership in isolation, i.e., from a purely external point of view, whereas for Kelsen it is pointless to think about membership without also considering validity. What matters for the present discussion, however, is that both do actually deal with both questions and the difference can be disregarded at this stage. See Carlos Santiago Nino, ‘Some Confusions Surrounding Kelsen’s Concept of Validity,’ in Normativity and Norms: Critical Perspectives on Kelsenian Themes, ed. Stanley L. Paulson (Oxford: Oxford University Press, 1999), 253-61.

• 12 Hart, The Concept of Law, 117.

• 13 For a discussion of this issue see Andrei Marmor, ‘Legal Conventionalism,’ Legal Theory 4 (1998): 509 and Gerald Postema, ‘Coordination and Convention at the Foundations of Law,’ Oxford Journal of Legal Studies 11 (1982): 165. For an attempt to find a solution via efficacy see Gerald Postema, ‘Conformity, Custom, and Congruence: Rethinking the Efficacy of Law,’ in The Legacy of H.L.A. Hart: Legal, Political, and Moral Philosophy, ed. Matthew H. Kramer, Claire Grant, Ben Colburn, & Antony Hatzistavrou (Oxford: Oxford University Press, 2008), 46.

• 14 Kurt Walter Zeidler, Grundriß der Transzendentalen Logik (Cuxhaven: Traude Junghans Verlag, 1997), 149.

• 15 See Henry Allison, Kant’s Transcendental Idealism. An Interpretation and Defence (New Haven: Yale University Press, 1983), 29.

• 16 For some reasons why the transcendental argument looks slightly different in the legal context than it does in the context of theoretical philosophy, see Christoph Kletzer, ‘Kelsen, Sander, and the Gegenstandsproblem of Legal Science,’ German Law Journal 12 (2011): 785-810.

• 17 This is not Raz’s point, that legal propositions are internal to a legal discourse and do not make sense outside of such a discourse or Hart’s point that questions about what ‘legally’ ought to happen only makes sense relative to a discourse constituted by the rule of recognition. It is not about a discourse at all. It is a claim about the actual legal process, i.e., about what legislators, judges, legal officials, and, ultimately, legal subjects do and how an interpretation of these actions is schematised by legal rules.

• 18 Nino’s helpful distinction between ‘rules’ and ‘judgements of validity’ leading to two different chains of derivation takes the latter ‘judgements of validity’ to be ‘typically formulated by jurists.’ With ‘jurists’ he presumably has academics in mind. The authentic and thus truly relevant ‘judgments of validity,’ however, do not happen in an academic but in a legal context in the legal process itself, i.e., as judgements (in Nino’s sense) done by parliamentarians (judging that the constitution is valid), by judges (judging statutes and constitutions to be valid), by executive forces (judging court orders, statutes and constitutions to be valid), and by legal subjects (judging executive commands, court orders, statutes and constitutions to be valid). See Nino, ‘Some Confusions,’ 257.

• 19 ‘Jurisprudence is the general part of adjudication, silent prologue to any decision in law,’ in Law’s Empire, Ronald Dworkin (Cambridge: Harvard University Press, 1986), 90.

• 20 In the context of Bruno Celano’s theory of validity as disquotation this means that we have to locate the activity of disquotation in the positive law itself (Bruno Celano, ‘Validity As Disquotation,’ Analisi e diritto (1999): 35-77). Here, with ‘disquotation’ the following is meant: just as in a minimalist theory of truth the trivial criterion of truth can be given as ‘“p” is true iff p,’ or ‘“Snow is white” is true iff snow is white,’ allowing to disquote sentences, to use a sentence which was formerly only mentioned, to make p out of ‘p,’ in Celano’s disquotational theory of validity a similar relation applies: ‘the norm “q” is valid iff q,’ or ‘The norm “Children ought to obey their parents” is valid iff children ought to obey their parents.’ Now, Celano thinks that the disquotational statements are statements of meta-jurisprudence, and that ‘validity’ is a device in the vocabulary of ‘a substantive ethical theory purporting to specify the conditions under which the law, or particular legal norms, ought to be obeyed’ (235), i.e., it belongs to the vocabulary of an inquiry into what the law ought to be and ‘not to the vocabulary of a scientific description of positive law as it actually is. This latter is, however, precisely the use to which validity as disquotation has been put, in the Pure Theory of Law, by Kelsen himself. In the Pure Theory of Law, a conceptual, necessary relation holds between positive law, on the one hand, and validity (validity as disquotation) on the other hand: positive legal norms are, as such, binding’ (236). In Kelsen’s work the concept of disquotation, conversely, rightly features as a technique not of meta-ethics or legal science but of the positive law itself. Of course, for Kelsen positive legal norms ‘as such’ can never be binding since positive legal norms ‘as such,’ i.e., irrespective of their relation to other legal norms, separated from the legal process, are not positive legal norms. In contrast to what Celano writes, Kelsen’s criterion of binding force is this: ‘“p” has binding force, if another norm q as a scheme of interpretation allows interpreting it as having binding force.’ Or: ‘p’ has binding force if another norm q disquotes ‘p.’

• 21 ‘The basic norm is a judgment of validity,’ yet at the same time ‘this fundamental judgement of validity, according to Kelsen, is itself a norm.’ Nino, ‘Some Confusions,’ 258.

• 22 ‘Auch das Voraussetzen ist ein Setzen, aber ein Setzen, das das Setzen zugleich als aufgehoben setzt. [Presupposing, too, is a positing. It is a positing, however, which posits its own positing as cancelled.]’ Dieter Henrich, Hegel im Kontext (Frankfurt/Main: Suhrkamp, 2010), 120.

• 23 See Hans Kelsen, Reine Rechtslehre, 2^nd ed (Wien: Verlag Österreich, 2000), 224.

• 24 See Uta Bindreiter, ‘Presupposing the Basic Norm,’ Ratio Juris 14 (2001): 143-75, 168: ‘It seems that the presupposition of the basic norm was intended, by Kelsen, to evoke a specifically “legal” use of language. Since the linguistic form of norm-formulations (used to issue norms) and of norm-statements (used to “describe” norms) may well be exactly the same, the difference between norm-formulation and norm-statement must lie in the illocutionary force of the respective utterances.’

• 25 The reflexive nature of positivism comes to light as an inner dividedness of positivism. Positivism was always torn between a philosophical and a decidedly anti-philosophical strand. A throughout philosophical positivism is a more recent phenomenon and positivistic tradition is rife with strong anti-philosophical tendencies. There is, for instance, the common opinion that the philosophical reflection of the law relates to the positive law as institutes to pandects, i.e., as textbook introduction to the real law. Philosophy can help write textbooks for novices. However, it cannot help us master the complexity of the positive law itself. See Friedrich Carl von Savigny, Vom Beruf unserer Zeit für Gesetzgebung und Rechtswissenschaft (Hildesheim: Georg Olms, 1967), 115.

• 26 See also, Hans Blumenberg, Die Legitimität der Neuzeit (Frankfurt/Main: Suhrkamp, 1999).