1It is an awkward word. It doesn’t sound good, and nobody’s too sure what it means. Yet it is here to stay. While it may be wise in principle to avoid referring to a term used in such different ways by legal theorists, expunging the word ‘normativity’ from our vocabulary would leave a void in our jurisprudential landscape. It is not just a matter of intellectual history. When Hart declared that he and Kelsen shared a common goal – explaining the normative dimension of law –, he is sometimes (Gardner 2007) taken to have given a solemn expression to something actually unremarkable: one of the features we need to be able to explain is law’s capacity to guide our conduct. Such a minimalist interpretation, however, does not even start to tackle the complexity and scope of the explanatory ambitions traditionally associated with the term ‘normativity’ in legal theory. Beyond its capacity to be used for guidance, law also claims to oblige us, and that claim tends to be central to any account of law’s normative dimension.
2Things typically start getting messy when it comes to explaining the way in which law’s normativity is not equivalent to (or measurable by reference to) its success in imposing obligations. Much of the murkiness currently associated with the concept of normativity can be attributed to a widespread assumption that it needs to (and can) be defined in terms of either guidance – law is normative if it is capable of being used for guidance, if it provides reasons for action – or obligation. Either way is equally problematic. While defining law’s normative dimension in terms of its capacity for being used for guidance would entail a constant fluctuation in its ‘degree’ of normative force (it would be a function of epistemological clarity and circumstances), to confuse normativity and obligation is worse. For it is often when law’s claim to impose obligations is punctually defeated that one gets the best chance of grasping the way in which law’s normative force is enabled on a daily basis by a fabric of socio-cultural practices.
- 1 “[T]he approach discussed in this article does not require the full machinery of a Kantian Transcen (...)
3But to talk of what ‘conditions’ or ‘enables’ law’s normative force is to already take sides: unlike Kelsen, Hart never had the ambition to explain what ‘enables’ or ‘conditions’ law’s normativity. Nor does Bix. Hart’s inscription within a philosophical context dominated by J.L. Austin’s theory of language, aimed at elucidating even the most complex concepts by reference to ‘the things people do with words’, predisposed him to developing what I have called elsewhere (Delacroix 2006) a ‘downstream account’ of legal normativity: its focus is to unveil how law’s normative dimension manifests itself (notably through the ‘critical reflective attitude’), rather than what enables it. That law is normative is a given, for both Hart and Bix.1 So far, so good: different accounts of legal normativity can and do proceed from different starting points, with different explanatory ambitions. The problems start when one dismisses the significance of these differences, or the possibility that fruitful insights may arise from a theory that explicitly questions one’s foundational premises. That there is an unbridgeable gap between is and ought is just one of those assumptions. It is acknowledged as such by Bix (2018: 27): “this view about not deriving ‘ought’ conclusions from ‘is’ statements [...] is generally accepted in modern philosophy”.
4Section 2 of this paper problematizes the above assumption: its hold over contemporary legal theory can indeed be said to be diminishing thanks in large part to better cross-fertilisation between legal theory and a well-established philosophical endeavour to delineate different kinds of naturalism(s). Aside from being in line with a Neo-Kantian tradition that was highly influential at the time, Kelsen’s robust methodological dualism may be said to stem in large part from the widespread belief that the only alternative to ‘Hume’s law’ is a reductive type of naturalism that leaves no room for any kind of ‘objectively true’ ethical statements. On this line of thought, any kind of naturalism entails that we only have ‘is’s’ all the way through, hence there is no ethical objectivity to be had: it is all a matter of personal preferences and subjective endorsements. For Kelsen, the threat inherent in such a relativist stance was ever-present, and it goes a long way toward explaining his stark rebuttal of any attempt to read his theory as ‘a doctrine of recognition’ (Kelsen 1967: 218, n. 83):
the doctrine of the Basic Norm is not a doctrine of recognition as is sometimes erroneously understood. According to the doctrine of recognition positive law is valid only if it is recognised by the individuals subject to it […] The theory of recognition, consciously or unconsciously, presupposes the ideal of individual liberty as self-determination, that is, the norm that the individual ought to do only what he wants to do.
5The above quote does not sit easily with Bix’s (2018: 32) suggested reading of Kelsen, whereby the presupposed Basic Norm reflects the fact that ‘legal rules and official actions [are] things that people may or may not view in a normative way’. Section 3 of this paper explores the factors that contribute to the tension between the above quote and Bix’s reading (via Raz’s own).
6Whatever else it is, naturalism involves at least one ‘lowest common denominator’ commitment. Its rejection of any dualist metaphysics involves a claim that ‘there is no unbridgeable space between what happens in that [natural] order and any other order in heaven or earth, including the order of our own minds’ (Blackburn 2001: 157): in short, there is no unbridgeable gap between is and ought. On this basis, the challenge any naturalist account of morality must address consists in understanding how the demands and aspirations we characteristically associate with morality may be understood as outgrowths of our animal (rather than noumenal, or god-like etc.) nature: ‘there must be no unmoved mover: no intervention of the divine spark, or gifts from unexplained quarters’ (Blackburn 2001).
- 2 “Of course, the NeoHumean naturalist has not explained real normativity, as Scanlon complains, beca (...)
7Beyond this ‘common denominator’, naturalism comes in many shapes and colours. Because of some versions’ scientistic excesses, ‘the tide of naturalism [which] has been rising since the seventeenth century’ can be perceived as a threat: ‘the regions under threat are some of the most central in human life’ (Price 1997: 247). To protect these ‘regions’ – morality being one of them – an increasing number of contemporary philosophers (from Price to Putnam via McDowell and Blackburn) find themselves ‘on the same [broad or liberal] side of the barricades’ (Redding 2010: 271), even if their respective naturalisms differ in some important ways. Elsewhere (Delacroix 2017) I distinguish between a range of naturalist positions by reference to three conceptual hurdles: defining what counts as ‘natural’, what counts as ‘science’, and defending the possibility of internalism about reasons. Because of naturalism’s supposed incompatibility with the latter position, many naturalists (including Leiter)2 end up falling back upon a subjectivist position, whereby moral propositions refer to the attitudes of people, rather than objective facts independent of human opinion. This need not be so.
- 3 I defend this position in greater detail in Delacroix 2017 and 2019.
- 4 The term reliable indicates a ‘relaxed’ approach (Lenman 2006) to drawing a connection between mora (...)
8A non-reductive naturalism3 is wary of any account that confines ‘the natural’ to that which results from elementary, material forces (as opposed to human forces). In contrast to ‘bald’ versions of naturalism, the explanatory resources at the disposal of non-reductive accounts need not be confined to those inert facts as they are described by Natural Sciences. For it is indeed a challenge to explain how such facts – and the beliefs they give rise to – reliably motivate us to act in certain ways (most of our factual beliefs typically do not). In contrast, a non-reductive naturalism may rely on things like habit and habituation to explain the reliable4 connection between moral judgment and motivation.
- 5 This meta-ethical dilemma is explored in detail in Delacroix 2010.
- 6 In her biography of H.L.A Hart, Nicola Lacey (2004) highlights Hart’s apparent (but unacknowledged) (...)
- 7 As an alternative, Hart could have developed a narrative along Weberian lines. The key challenge wo (...)
- 8 These were known to be influenced by Mackie’s theory. See Wiggins 2005.
- 9 Williams 1985.
- 10 “Why should anyone be disturbed if Williams is right in his claim that there is no independent rati (...)
9For the purpose of this paper, however, the salient question is: what factors, if any, contributed to Kelsen’s endorsing the robust methodological dualism he is famous for – this is the focus of section 3 –, and how does this compare to Hart’s own meta-ethical dilemma? I have argued elsewhere5 that the latter dilemma just did not allow Hart to develop what could have been a (non-reductive) naturalist account of legal normativity, whereby there would be no ‘gap’ between social facts on the one hand and social rules on the other. There are reasons to believe that Hart did have some affinities for such an account, which are in part corroborated by his reliance on both Weber’s Economy and Society6 and Wittgenstein’s Philosophical Investigations. Hart could, for instance, have chosen to expand upon the latter’s reference to ‘custom’ as a way of explaining how the causal processes constitutive of habit acquire the significance they do when they evolve into rule-following practices. On this account, there is no ‘gap’ to bridge, only a narrative7 that takes us from human beings with needs and desires (‘the sphere of causes’) to internalized standards of right and wrong. That Hart chose to remain agnostic instead, limiting his account to a rather thin notion of ‘acceptance’, can be explained by various factors, including, most interestingly, his late but candid acknowledgment of a life-long dilemma between his sceptical inclinations8 and a longing for some external, rational foundation for ethics. In his review of Bernard Williams’ Ethics and the Limits of Philosophy,9 this longing is expressed negatively, as the fear that, were we to look for the ground of the moral ‘I must’ in ‘what is most deeply inside us’, it would not prove ‘enough’ to sustain a sense of moral necessity.10
10Hart’s acknowledgment is particularly interesting because in some ways it echoes the concerns cited by Habermas to explain why he has chosen to retain the quasi-transcendentalist assumptions underlying his theory, despite his pragmatist inclinations:
I once asked Habermas in a public forum what was the most difficult aspect of his philosophy to defend. He didn’t hesitate to answer: quasi-transcendentalism. And when I asked why he thought that he had to defend it [...] his answer was straightforward: the Holocaust (Aboulafia 2002: 4).
11While Kelsen first developed the dualist methodology underlying his theory (under the influence of a growing Neo-Kantian school of thought) well before the Holocaust, the political reality he was confronted with is likely to have played some role in his growing mis-trust of human judgment (and consequent determination to isolate the legal sphere from the messy and increasingly threatening reality of politics). This aspect of Kelsen’s theory is discussed in the next section, not only because it calls into question Bix’s proposed ‘subjective endorsement’ interpretation, but more importantly, because it illustrates a methodological issue, which is often at the root of misunderstandings between the Continental and Anglo-American traditions (section 4).
- 11 “For the norms of natural law, like those of morality, are deduced from a Basic Norm that by virtue (...)
- 12 Kelsen explicitly addresses this state of social disturbance and lack of self-confidence in Kelsen (...)
12Kelsen's project not only aims to distinguish the object of legal cognition from the objects of other sciences, such as sociology or ethics. It also requires this object to be ‘autonomous’, explainable by reference to normative considerations alone, without having to resort to either moral or factual considerations. As for the moral considerations, Kelsen’s opposition to natural law theories manifests itself through his dismissing any attempt to explain law's normativity by reference to some pre-existing natural laws. Kelsen indeed rejects any attempt to locate the source of legal normativity in a concept of the good that would precede human action (be it given by God or incarnated in Nature or Reason11). This rejection proceeds from an effort to resist a growing temptation to restore social confidence12 by providing a substantial justification of the legal order.
- 13 “On one reading [...], Sein and Sollen mark two points of view, the explicative and the normative; (...)
13As for factual considerations, Kelsen’s sharp distinction between the order of value and the order of reality, between Sollen and Sein, allows no passage from one order to the other. On this view, the worlds of facticity and normativity are not only differentiable: they are totally independent, corresponding as they do to two different spheres of knowledge.13 Any endeavour to derive legal norms from social practices misunderstands the different epistemological status characterising legal norms and social facts.
- 14 Kervégan (1995: 238) describes Kelsen as “a characteristic representative of normativism given his (...)
14Now, the combined weight of the above methodological commitments mean that legal normativity is necessarily cut off from any extraneous explicative resource.14 One of Kelsen’s very early works, i.e. Das Problem der Souveränität (1920) reveals a striking awareness of the difficulties inherent in this ambition to conceptualise the legal sphere in a ‘closed’ and self-sustaining way:
- 15 In his subsequent writings, Kelsen's way of coming to terms with this enigmatic sentence amounts to (...)
to want to determine the choice of juristic starting point juristically would be tantamount to standing on one's shoulders, [and] would be equivalent to Munchausen's attempt to pull himself out of the swamp by his own pigtail (Kelsen 1960: 96).15
15So, one may ask: if he was aware of the considerable difficulties (which I document in detail elsewhere (Delacroix 2006)) underlying this ambition to build a ‘methodologically self-standing’ account of what he calls ‘the specific lawfulness of the law’, why did Kelsen end up radicalising – rather than merely following – the lead of his fellow fin de siècle Neo-Kantians (who sought to extend Kant's transcendental inquiry to fields outside the realm of natural science)? Kelsen (1998: 170) is indeed critical of both Laband’s “utter failure in his effort to separate the depiction of the positive law from politics” and Jellinek’s ‘two-sides’ theory of law (one side being examined from a legal point of view while the other is apprehended sociologically). Kelsen (1949b: 444-445) actually goes as far as to claim that Kant’s rejection of the application of the transcendental method to the fields of law and ethics is a sign of weakness:
A complete emancipation from metaphysics was probably impossible for a personality still as deeply rooted in Christianity as Kant's. This is most evident in his practical philosophy [...] At this point, Kant abandoned his method of transcendental logic [...] So it happens that Kant, whose philosophy of transcendental logic was pre-eminently destined to provide the groundwork for a positivistic legal and political doctrine, stayed, as a legal philosopher, in the rut of the natural-law doctrine.
- 16 “Just as laws of nature link a certain material fact as cause with another as effect, so positive l (...)
- 17 “[T]he Pure Theory is well aware that one cannot prove the existence of the law as one proves the e (...)
- 18 Indeed, to provide an answer to the sceptic, Kelsen would have had to resort to the progressive ver (...)
16Kelsen thus leverages the transcendental method to build a theory according to which the category of imputation conditions our knowledge of legal propositions, just as the category of causation conditions our knowledge of natural phenomenon.16 Where Kelsen departs from Kant, however, is in relying exclusively on the regressive version of Kant’s transcendental argument: Kelsen indeed has to start from the fact that one has cognition of legal norms, to then assert that this cognition is possible only if the category of normative imputation is presupposed. He then concludes that cognition of legal norms necessarily presupposes the category of imputation. Kelsen does acknowledge the fact that the sceptic is unlikely to assent to the first premise17 – the fact of normative cognition (given that these knowledge claims are precisely the target of the sceptic) – but his theory was never meant as an answer to such sceptical positions, including theoretical anarchism.18
17This leads us to a central point: Bix builds upon Kelsen’s acknowledgment that an anarchist need not presuppose the Basic Norm to develop a ‘subjective endorsement’ interpretation of Kelsen’s understanding of legal normativity. There are many ways such an interpretation makes sense. Not only does it tie in nicely with Raz’s infamous concept of ‘detached point of view’, it may also be the only interpretation that ‘works’ if one is not prepared to question the “view [according to which] not deriving ‘ought’ conclusions from ‘is’ statements [...] is generally accepted in modern philosophy” (Bix 2018: 27) – a view Kelsen would certainly endorse. Yet there’s a catch.
18Kelsen’s stance as to whether or not the anarchist can be seen as presupposing the Basic Norm, changes drastically over the years: in ‘Value Judgments in the Science of Law’, Kelsen (1942: 329) states that the anarchist “will decline to speak of ‘lawful’ and ‘unlawful’ behaviour, of ‘legal duties’, or ‘delicts’. He will understand social behaviour merely as a process whereby one forces the other to behave in conformity with his wishes or interests [...] He will, in short, refuse to presuppose the Basic Norm”. Yet in the second edition of The Pure Theory of Law, Kelsen withdraws his previous statement that the anarchist “refuses to presuppose the Basic Norm”, to argue that an anarchist can actually object to the law and yet “describe positive law as a system of valid norms, without having to approve of this law” (Kelsen 1967: 218, n. 82). This reversal may have been due to Kelsen realising that “disapproval of law in general, in virtue of a moral attitude (anarchism), or of a certain type of legal system, in view of a political attitude (communism), constituted, for this theory, a practical stance such as could co-exist with the juristic view, without affecting it or being affected by it” (Wilson 1982: 53). Surprisingly, although he is aware of Kelsen’s explicit change of mind on the question, Raz refers to Kelsen’s early statement – maintaining that the anarchist will “refuse to presuppose the Basic Norm” – to support his claim according to which “for an individual to presuppose the basic norm is to interpret the legal system as normative, i.e., as just” (Raz 1979: 138).
- 19 Those flaws are analysed in detail in Delacroix 2004.
19For our purposes, what matters is not so much the flaws in Raz’s interpretation of Kelsen,19 but rather the factors that lead both Bix and Raz to defend this ‘subjective endorsement’ account, given that the latter is at odds with Kelsen’s (1967: 218, n. 83) explicit rejection of what he called a ‘doctrine of recognition’. At the heart of these factors is a culturally-loaded misunderstanding of the very different explanatory goals that can be associated with an endeavour to account for law’s normativity.
- 20 It is doubtful whether this is more than a counterfactual stance, as law tends to provide reasons f (...)
20One may seek to explain how law’s normativity manifests itself through our linguistic and social practices. One of the challenges, from that perspective, is to account for the fact that, in theory at least,20 some of us may never consider law to be a source of reasons for action (while nevertheless referring to law as a normative system ‘from a detached perspective’), while others may regularly have (moral or prudential) reasons to defeat law’s prima-facie reasons. One may then analyse the contrast between the normativity of law and that of morality by pointing out, as Gardner (2007) does, that morality’s hold over us – ‘inescapable’ as it is – is rather different from law’s. A large proportion of the Anglo-American literature on legal normativity proceeds within this explanatory framework, under Hart’s decisive influence.
- 21 If only from a terminological perspective, such a statement is exceptional. It is not only the sole (...)
- 22 For a detailed argument to this effect, see Delacroix 2006.
21Alternatively, one may seek to understand what makes law’s normative status possible in the first place. How can contingent, fallible, and ‘man-made’ laws have such a hold over us? This line of questioning is ancient. Ever since the progressive discrediting of the meta-referents (whether they be God or Nature) that are central to classical natural law models, this question has become ‘live’ again. One may adopt a multitude of strategies to answer it – or indeed circumvent it, a trend inaugurated by Montaigne (1991: 543): to ‘save’ law from the peril inherent in the discovery of its precarious origins, Montaigne posits a ‘law of pure obedience’, enjoining us to obey the law ‘because it is the law’, without asking any further questions. For, to inquire further into the sources of this law of pure obedience is to expose oneself to the risk of ‘disgust’ for the law and its authoritative force. Montaigne thus goes as far as using the term ‘mystical’21 to refer to the need to shroud the origins of law’s normative force in ‘mysteries’ available only to the ‘wise’, who are less likely to let the discovery of the contingent ‘tiny spring’ at the origins of law affect their respect for law’s authority. In many ways, I think there is a strong parallel between Montaigne’s circumvention strategy and Kelsen’s.22
- 23 According to Schmitt, the systematic “disregard of the independent problem of the realisation of la (...)
- 24 See for instance Kelsen 1967: 219, which has an equivalent in Kelsen 1949b: 401, except that any re (...)
- 25 In General theory of law and state, Kelsen characterises the content of the Basic Norm as “determin (...)
22Kelsen is a lot less candid about it than Montaigne, but Kelsen’s theory, through all its twists and turns, is attempting to answer the age-old ‘sources’ question – ‘what are the sources of legal normativity?’ (in the same way as Korsgaard asks ‘what are the sources of [moral] normativity?’) –, while at the same time ruling out any reference to social facts or moral norms. In other words, the source of legal normativity has to be found within legal normativity itself, a task that may well be akin to “Munchausen's attempt to pull himself out of the swamp by his own pigtail”. The solipsism inherent in such a normativist ambition was rightly emphasised by Schmitt, who criticises the trend to proceed as if legal normativity can be understood without any reference to its conditions of effectuation.23 The many twists and turns within Kelsen’s theory – including, most interestingly, the disappearance of the notion of the ‘content’ of the Basic Norm in the second edition of the Pure theory of law24 – do betray considerable tensions, which culminate in Kelsen’s turn to a characterisation of the Basic Norm as a fiction. The latter testifies, somewhat dramatically, to Kelsen's awareness of the impossibility for the Basic Norm to be at the same time this transcendental presupposition free of any ideological or factual consideration and the element that, via its content, intrinsically refers to the factual sphere.25
23In the light of these difficulties, Bix’s ‘subjective endorsement’ interpretation may be deemed a neat and elegant way of rescuing Kelsen’s account of normativity. The problem is that it underestimates the significance of the contrast between Hart’s and Kelsen’s respective explanatory ambitions. To state that law is a system that one may (or may not) choose to view in a normative way is tantamount, from Kelsen’s perspective, to a form of subjective voluntarism, as it effectively traces the source of legal normativity to the individual’s subjective endorsement of it. Clearly, the latter voluntarist interpretation is not what Bix has in mind. But if, like Kelsen, your explanatory ambition is not limited to accounting for the way law’s normativity manifests itself within our socio-linguistic practices, but instead aims to explain what makes it possible, then this voluntarist interpretation (with its aporetic implications) imposes itself, as exemplified in the passage quoted earlier:
The theory of recognition, consciously or unconsciously, presupposes the ideal of individual liberty as self-determination, that is, the norm that the individual ought to do only what he wants to do (Kelsen 1967: 218, n. 83).
24The point of this paper is not to argue about the merits of this or that interpretation of Kelsen’s account of legal normativity, but rather to emphasise the fact that one may associate one’s account of legal normativity with very different explanatory ambitions. An endeavour to trace the sources of legal normativity need not be doomed (or ‘shrouded in mysteries’ a la Montaigne). Granted, such an endeavour is unlikely to succeed if, like Bix, one rigorously upholds what some people refer to as ‘Hume’s Law’ – the idea that one may not derive an ought from an is. In some ways Kelsen’s theory is a formidable demonstration of the unavoidable circularity that is concomitant with a strict dualist methodology.
25Yet this need not mean that a Hartian, downstream account that takes legal normativity as a given (and proceeds from there) is the only way to go. On the upstream side, legal theory has a lot to learn from various forms of non-reductive naturalism. The latter may, among other things, help dislodge the idea that any account of the social practices that give rise to legal norms must presuppose intentional agency all the way through (as per conventionalist accounts).
RSS feed