Of all the legal theorists you will encounter, Hans Kelsen is probably the least understood (not only by students) and most misrepresented. I have therefore devoted more space to Kelsen than other positivists. Th is refl ects my eff orts, over many years, to reduce the pain and suff ering that Kelsen has induced in my own students. Much of this is doubtless a con-sequence of his use of fairly diffi cult and abstract conceptual language which, especially to those unfamiliar with the Continental approach to philosophy, is not always congenial.
Very few question his remarkable facility for critical exposition and inquiry, indeed, it has been claimed that ‘no single writer . . . [has] made a more illuminating analysis of the legal process’.36 His pure theory of law has become as important (if not nearly as infl uential) as Hart’s theory, and represents a signifi cant strand in modern legal positivism.37 But, though the Kelsenian enterprise is not entirely painless, I hope the following discussion will assist your comprehension and appreciation of its principal features.
To the extent that he insisted on the separation of law and morals, what ‘is’ (sein) and what ‘ought to be’ (sollen), Kelsen may legitimately be characterized as a legal positivist, but he is a good deal more. Th e pure theory is a subtle and profound statement about the way in which we should understand law. And we should do so, he argues, by conceiving it to be a system of ‘oughts’ or norms. But Kelsen acknowledges that the law consists not merely of norms, but ‘is made up of legal norms and legal acts as determined by these norms’.38 In other words, legal norms (which include judicial decisions and legal transac-tions such as contracts and wills) when acted upon also describe actual human conduct.
Even the most general norms describe human conduct. Th us:
Kelsen’s observation that the legal scientist is not concerned with human conduct but is only concerned with norms may have obscured from view the important point that so far as human conduct features in a norm as condition or consequence such conduct falls four-square within the concerns of the Kelsenian legal scientist. . . . Kelsen permits of a greater degree of reference to actual human conduct than is sometimes perceived by those who would label his contribution as ‘sterile.’ 39
Indeed, it has been argued that Kelsen’s attempt to understand and explain the ‘science of the mind and of meaning and of values as instantiated in actual human societies’ is
‘the only jurisprudence ever to take sociology seriously’.40 He was, it has been suggested,
‘engaging in sociology when writing his Pure Th eory, notwithstanding his indignant denials’.41 Th e validity or otherwise of these claims should become clearer in the course of the following pages.42
As a follower of the great eighteenth-century philosopher, Immanuel Kant, Kelsen espouses the view that objective reality can be comprehended only by the application of certain formal categories like time and space. Th ese categories do not ‘exist’ in nature: we
36 Lloyd’s Introduction to Jurisprudence, 8th edn (London: Sweet & Maxwell, 2008), 305.
37 Dias describes Kelsen’s writings as constituting ‘the most refi ned development to date of analytical positivism’, RWM Dias, Jurisprudence, 5th edn (London: Butterworths, 1985), 258.
38 General Th eory of Law and State, transl Anders Wedberg (Cambridge, Mass: Harvard University Press, 1949), 39.
39 Tur and Twining (eds), Essays on Kelsen, 23–4.
40 R Tur, ‘Th e Kelsenian Enterprise’ in Tur and Twining (eds), Essays on Kelsen, 150 at 182.
41 G Sawer, Law in Society (Oxford: Clarendon Press, 1965), 5, quoted by Tur, op. cit.
42 I draw here on my essay, ‘One Country, Two Grundnormen? Th e Basic Law and the Basic Norm’ in R Wacks (ed), Hong Kong, China and 1997: Essays in Legal Th eory (Hong Kong: Hong Kong University Press, 1993).
Kelsen’s observation that the legal scientist is not concerned with human conduct but is only concerned with norms may have obscured from view the important point that so far as human conduct features in a norm as condition or consequence such conduct falls four-square within the concerns of the Kelsenian legal scientist. . . . Kelsen permits of a greater degree of reference to actual human conduct than is sometimes perceived by those who would label his contribution as ‘sterile.’39
use them in order to make sense of the world. Equally, to understand ‘the law’ we require similar formal categories, in particular the Grundnorm or basic norm which lies at the heart of the legal system (see 4.3.3).
Kelsen’s project is thus a fairly ambitious one. He seeks, to use his own words, to raise jurisprudence ‘to the level of a genuine science’. His theory is described by Richard Tur as
‘a thoroughgoing attempt to develop an epistemology for jurisprudence. It is a recipe for legal knowledge’.43 Th e ingredients are, however, oft en far from straightforward. And the result is not to everyone’s taste.
4.3.1 Unadulterated law
Few have diffi culty in grasping Kelsen’s insistence on excluding the ‘impurities’ of morality, history, politics, sociology, etc. If we are to arrive at a scientifi c (as opposed to a subjective, value-laden) theory of law, says Kelsen, we need to restrict our analysis to the
‘norms’ of positive law: those ‘oughts’ which provide that if certain conduct (X) is per-formed, then a sanction (Y) should be applied by an offi cial to the off ender. If X then Y.
Th e theory therefore rules out all that cannot be objectively known: the social purpose of law, its political functions, etc. Law has only one function: the monopolization of force.
Kelsen’s pursuit of a ‘science of law’ is premised on the claim that an account of law can be disinfected from ‘elements of psychology, sociology, ethics, and political theory’.44 In his words:
Th is adulteration is understandable, because [these] disciplines deal with subject-mat-ters that are closely connected with law. Th e pure theory of law undertakes to delimit the cognition of law against these disciplines, not because it ignores or denies the connec-tion, but because it wishes to avoid the uncritical mixture of methodologically diff erent disciplines . . . which obscures the essence of the science of law and obliterates the limits imposed upon it by the nature of its subject-matter.45
By ‘norms’ Kelsen means that ‘something ought to be or ought to happen, especially that a human being ought to behave in a specifi c way’.46 Th us the statement ‘the door ought to be closed’ or a red traffi c light are both norms. But a norm, in order to be valid (ie, binding), must be authorized by another norm which, in turn, is authorized by a higher norm in the system. Th e separation between law and morality means that the validity of legal norms can fl ow only from another legal, as opposed to a moral norm. Kelsen is profoundly rela-tivistic: he rejects the notion that there are values ‘out there’; all norms are relative to the individual or group under consideration. Th is point is well explained by Professor Raz47 who shows that Kelsen is not a sceptic (ie, he does not take the view that all normative statements are necessarily false), he is a relativist or subjectivist:
Normative statements can be true or false. It is merely that their truth depends on the existence of relativistic rather than absolute values: ‘relativistic . . . positivism does not assert that there are no values, or that there is no moral order, but only that the values in which men actually believe are not absolute but relative values.’48
43 R Tur, ‘Th e Kelsenian Enterprise’ in Tur and Twining (eds), Essays on Kelsen, 149–83, 157.
44 Pure Th eory of Law, transl Max Knight (Berkeley and Los Angeles: University of California Press, 1967), 1.
45Loc cit. 46 Pure Th eory of Law, 4.
47 J Raz, ‘Th e Purity of the Pure Th eory’ in Tur and Twining (eds), Essays on Kelsen, 79 at 87.
48 H Kelsen, What is Justice? (Berkeley and Los Angeles: University of California Press, 1957), 179.
Th is adulteration is understandable, because [these] disciplines deal with subject-mat-ters that are closely connected with law. Th e pure theory of law undertakes to delimit the cognition of law against these disciplines, not because it ignores or denies the connec-tion, but because it wishes to avoid the uncritical mixture of methodologically diff erent disciplines . . . which obscures the essence of the science of law and obliterates the limits imposed upon it by the nature of its subject-matter.45
Normative statements can be true or false. It is merely that their truth depends on the existence of relativistic rather than absolute values: ‘relativistic . . . positivism does not assert that there are no values, or that there is no moral order, but only that the values in which men actually believe are not absolute but relative values.’’48
As Raz remarks, a conspicuous diffi culty with this form of relativism is its assumption that any sincere moral statement I make about myself must be true; because I believe that there is a norm requiring me to perform a certain act such a norm exists and my state-ment is true. By the same process of reasoning insincere moral statestate-ments about myself are always false. And normative statements I make about others are true only if they con-form to their beliefs about themselves. ‘Th us’, concludes Raz, ‘it is true that a racist should behave in a racist way.’49 Th is is clearly unacceptable.
Nevertheless Kelsen’s relativist, value-free theory of law seeks to locate legal science in a world free of the ‘impurities’ of social science. It provides, in Stewart’s words:
the basic forms under which meanings can be known scientifi cally as legal norms—which will have a content, although the particular content is empirically contingent, and which, once determined as having a particular content, can be morally evaluated. Far from being an attempt to exclude considerations of experience, content, and justice, the pure theory is intended to make attention to them more rigorously possible.50
Th e hierarchy of legal norms that forms a legal system is ultimately traced back to the Grundnorm or basic norm of the legal system. Its nature, function, and relationship to other norms will be examined later.
Th e law consists of norms used as a ‘specifi c social technique’ by politicians to deter-mine how individuals ought to behave so as to promote social order and peace. Th is tech-nique consists in the acts of will of individuals authorized by the law to create norms which render the behaviour of individuals lawful or unlawful by providing sanctions for failure to comply with the norms. Th us legal norms diff er from other norms in that they prescribe a sanction. Th e legal system is founded on state coercion; behind its norms is the threat of force. Th is distinguishes the tax collector from the robber. Both demand your money. Both, in other words, require that you ought to pay up. Both exhibit a subjective act of will, but only the tax collector’s is objectively valid. Why? Because, says Kelsen, the subjective meaning of the robber’s coercive order is not interpreted as its objective mean-ing. Why not? Because ‘no basic norm is presupposed according to which one ought to behave in conformity with this order’.51 And why not? Because the robber’s coercive order lacks the ‘lasting eff ectiveness without which no basic norm is presupposed’. Th is illus-trates the fundamental relationship in Kelsen’s theory between validity and eff ectiveness which is discussed below.
Kelsen’s reduction of all legislation to the form ‘If X, then Y’ (where X is certain con-duct, and Y is a sanction) is widely regarded as unacceptably narrow. Th e form of law is given primacy over its meaning. It presumes (which, of course, Kelsen is content to do) that law is essentially coercion; many would want to argue that law has other functions, for example, regulatory purposes.
Other critics seek to show that Kelsen accords unwarranted importance to the role of sanctions in law. It results in a lopsided analysis of legal duty not only because a statute may impose duties without necessarily providing a sanction, but because, on the other hand, certain conduct may be made the condition of a sanction even though it is not the subject of a duty. Th us JW Harris points out that to measure eff ectiveness we need to know the content of the norm, that is, the nature of the duty involved. As he puts it, ‘Th e concept
49 Raz, ‘Th e Purity of the Pure Th eory’, 88.
50 I Stewart, ‘Kelsen and the Exegetical Tradition’ in Tur and Twining (eds), Essays on Kelsen, 123 at 128.
51 Pure Th eory of Law, 47.
the basic forms under which meanings can be known scientifi cally as legal norms—which will have a content, although the particular content is empirically contingent, and which, once determined as having a particular content, can be morally evaluated. Far from being an attempt to exclude considerations of experience, content, and justice, the pure theory is intended to make attention to them more rigorously possible.50
of “duty” must . . . stand on its own feet, as something distinct from the concept of sanc-tion. A theory of law must defi ne duty and sanction separately.’52
4.3.2 A hierarchy of norms
Kelsen represents a legal system as a complex series of interlocking norms which progress from the most general ‘oughts’ (eg, sanctions ought to be eff ected in accordance with the constitution); to the most particular or ‘concrete’ (eg, David is contractually bound to mow Victoria’s lawn). Each norm in this hierarchical system draws its validity from another—higher—norm. Th e validity of all norms is ultimately based on the Grundnorm (see 4.3.3).
Th is systemic, hierarchical model of law provides also the explanation for the dynamic creation of legal norms. Th e membership of norms in the legal system is determined by other norms in the hierarchy. Law is created by facts (eg, a judicial decision) which con-vey normative force from the authorizing norm to the authorized norm. Th e authorizing norm being valid and capable of endowing law-creating acts with status to create law, the norm so created is also valid. Law-creating acts thereby confer validity from one norm to another.
As has been pointed out, the validity of each norm is dependent on a higher norm in the system whose validity is in turn dependent upon a higher norm in the system and so on. A point is eventually reached beyond which this climbing cannot go. Th is is the basic norm or Grundnorm. All norms fl ow from it in increasing levels of ‘concreteness’:
the basic norm expresses an ‘ought’ at the highest level of generality. Below it, in the hierarchy of norms, is the historically fi rst constitution. Below it are laws enacted—by the legislature or judiciary—which are more ‘concrete’, all the way down to the most con-crete, individualized norm such as: ‘the bailiff is empowered to seize the property of the defendant who has been found by a court to be liable to the claimant and who is unable to pay what he owes’. Th e coercive act of the bailiff (or the prison warder in incarcerating a prisoner) is the ultimate stage in the progression from general basic norm to particular individuated norm.
4.3.3 The Grundnorm
Since, by defi nition, the validity of the Grundnorm cannot depend on any other norm it must be presupposed. What does this mean? Kelsen seems to be saying (and this is a mat-ter of some controversy) that we need this assumption in order to understand the legal order. As he says, disclaiming any originality:
By formulating the Grundnorm, we do not introduce into the science of law any new method. We merely make explicit what all jurists, mostly unconsciously, assume when they consider positive law as a system of valid norms and not only as a complex of facts, and at the same time repudiate any natural law from which positive law would receive its validity. Th at the Grundnorm really exists in the juristic consciousness is the result of a simple analysis of actual juristic statements. Th e Grundnorm is the answer to the ques-tion: how—and that means under what condition—are these juristic statements concern-ing legal norms, legal duties, legal rights, and so on, possible?53
52 Legal Philosophies, 67.
53 General Th eory of Law and State, 117.
By formulating the Grundnorm, we do not introduce into the science of law any new method. We merely make explicit what all jurists, mostly unconsciously, assume when they consider positive law as a system of valid norms and not only as a complex of facts, and at the same time repudiate any natural law from which positive law would receive its validity. Th at theGrundnorm really exists in the juristic consciousness is the result of a simple analysis of actual juristic statements. Th eGrundnormis the answer to the ques-tion: how—and that means under what condition—are these juristic statements concern-ing legal norms, legal duties, legal rights, and so on, possible?53
Th is is a lucid statement of the role Kelsen assigns to his basic norm: it exists, but only in the
‘juristic consciousness’. It is a presupposition which facilitates an understanding of the legal system by the legal scientist, judge, or lawyer.54 But it is not chosen arbitrarily; it is selected by reference to whether the legal order as a whole is ‘by and large’ effi cacious. Its validity depends on effi cacy (see 4.3.4). Th e presupposed basic norm is characterized by Stewart as:
the nodal point at which the pure part of legal science passes over into the empirical part;
on the pure side, the basic norm stands in a relation of validity to the specifi c and generic formulations of the presupposition ‘basic norm’ and through them to the pure theory as a whole, while on the empirical side it stands in a relation of validity to the remainder of the legal order; its validity on the pure side cannot be questioned from the empirical side, since it is the condition of possibility, furnished by the pure side for the empirical side.55
In other words, the validity of the basic norm rests, not on another norm or rule of law, but is assumed—for the purpose of purity. It is therefore what Kelsen occasionally calls a
‘juristic hypothesis’, though it is sometimes described as a presupposition or even a fi c-tion. According to its creator, it ‘presents itself . . . not as a guess or hypothesis about the reality behind the law but explicitly as a methodological maxim, a norm of method which is ontologically neutral’.56
Consider his religious analogy:
A father addresses to his son the individual norm, ‘Go to school.’ Th e son asks his father,
‘Why should I go to school?’ Th at is, he asks why the subjective meaning of his father’s act of will is its objective meaning, i.e., a norm binding for him—or, which means the same thing, what is the basis of the validity of this norm. Th e father responds: ‘Because God has commanded that parents be obeyed—that is, He has authorised parents to issue
‘Why should I go to school?’ Th at is, he asks why the subjective meaning of his father’s act of will is its objective meaning, i.e., a norm binding for him—or, which means the same thing, what is the basis of the validity of this norm. Th e father responds: ‘Because God has commanded that parents be obeyed—that is, He has authorised parents to issue