Hart versus Kelsen

Why is a hundred-rupee note worth a hundred rupees? One answer: because in fact everyone, above all the banks and shops, treats it as such; the practice is the foundation, and it makes no sense to ask whether the practice itself is 'valid'. That is Hart. Another answer: because if you reason about money at all, you must assume at the start that the issuing authority's notes ought to be accepted; that assumption is not a fact but the presupposition of the whole system of monetary reasoning. That is Kelsen. Both answers support the same working system. They differ about what lies at the bottom of it. This note compares them; their theories in full are in Notes 11 and 12.

1. Parallel Systems

The resemblance is striking. Both reject Austin's sovereign and habit of obedience (Notes 61 and 62). Both see a legal system as a set of norms or rules linked by chains of validity: a byelaw is valid because a statute authorises it, a statute because the constitution authorises it. Both separate law from morality. Both say the whole system exists only if it is, by and large, effective. And both must answer the same question: what makes the top of the chain valid? Their different answers are the heart of the comparison.

2. Hart's Answer: The Rule of Recognition as a Fact

For Hart (The Concept of Law, 1961), the ultimate rule of recognition is not validated by any further rule. Its existence is a matter of fact: it exists only as a complex, but normally concordant, practice of the courts, officials and private persons in identifying the law by reference to certain criteria (Hart's point, paraphrased). It is therefore neither valid nor invalid; it is simply accepted as appropriate for use. Hart compared it to the standard metre bar once kept in Paris: it makes no sense to ask whether the bar itself is a metre long, because it is the standard by which other lengths are measured. Statements of validity are internal statements made within the system; statements that the rule of recognition exists are external statements of fact.

Hart also distinguished two ideas that students often merge. A criterion is supreme if rules identified by it prevail over rules identified by other criteria (in the United Kingdom, enactment by the Queen in Parliament ranks above custom and precedent). A rule is ultimate if it is not itself validated by any other rule. The rule of recognition is ultimate; the criteria it contains are ranked, and one of them is supreme.

3. Kelsen's Answer: The Grundnorm as a Presupposition

For Kelsen, an ought can be derived only from another ought, never from an is. The chain of validity must therefore end in a norm, not a fact. But this last norm cannot have been enacted by anyone, since then we would ask what authorised its enactment. So it is presupposed: legal science, when it treats a given effective order as valid, presupposes that one ought to behave as the historically first constitution prescribes. In his classic works (Pure Theory of Law, 1934 and 1960; General Theory of Law and State, 1945), Kelsen described the basic norm in neo-Kantian terms as a transcendental-logical presupposition of legal cognition. In the posthumous General Theory of Norms (1979), he changed his account and called it a fiction, in the sense used by Vaihinger's philosophy of 'as if': a thought known to be contrary to reality but useful for thinking.

Static and dynamic systems

Kelsen explained why the chain of validity is formal. A static system, such as morality, derives its norms from the content of a basic norm: from 'love your neighbour' one deduces 'do not injure him'. A dynamic system, such as law, derives its norms from authorisation: a norm is valid because it was made in the way a higher norm prescribes, whatever its content, although a constitution may also forbid or require certain content. Hart's rule of recognition likewise identifies law chiefly by pedigree, but in the Postscript he allowed that it may include moral criteria, which Kelsen's pure theory would treat as a matter for the positive constitution alone.

4. Primary and Secondary: Opposite Meanings

For Hart, primary rules impose duties on ordinary people (do not steal), and secondary rules are rules about rules (recognition, change, adjudication). For Kelsen, the primary norm is the norm directing an official to apply a sanction (if a person steals, the judge ought to punish), and the secondary norm is the derivative norm addressed to the citizen. The same words carry reversed meanings, a favourite examination trap.

This is also the root of Hart's main criticism of Kelsen in chapter 3 of The Concept of Law. Hart argued that treating every law as a direction to officials to apply sanctions distorts the social function of law. The criminal law is meant chiefly to guide citizens, most of whom never meet an official; to read it as a set of instructions to judges is to see it only from the point of view of the bad man and the court. Power-conferring rules, such as the rules for making a will, become mere fragments of sanction norms, when in life they are facilities that people use to shape their affairs.

5. The Exchange between Them

The two met in debate at Berkeley in 1961, the year The Concept of Law appeared, and Hart published his reflections as 'Kelsen Visited' (1963). The essay examined several of Kelsen's doctrines, including his idea that legal science describes norms in 'ought' statements of its own, and his view that a legal norm and a conflicting moral norm cannot both be regarded as valid from a single point of view. The exchange showed how close their projects were in aim and how different in method: Hart was an Oxford linguistic philosopher describing a social practice; Kelsen was a neo-Kantian constructing the presuppositions of a normative science.

6. International Law

For Hart (The Concept of Law, ch. 10), international law is a set of rules, binding and treated as law, but not a system: it lacks a legislature, courts with compulsory jurisdiction and, above all, a unifying rule of recognition. It resembles the simple regime of primary rules. Hart dismissed the suggested basic norm that States should behave as they have customarily behaved as an empty repetition of the fact that they accept certain rules. For Kelsen, international law is a genuine though primitive system, part of a single monist legal order with national law.

7. The Two Compared

8. Revolutions and the Indian Constitution

Both theories explain revolutions, in different language. For Kelsen, when a new order becomes by and large effective, a new basic norm is presupposed; this is the reasoning applied in State v Dosso (Pakistan, 1958), later overruled in Asma Jilani (1972), and in Matovu (Uganda, 1966), while the Privy Council in Madzimbamuto (1968) refused to treat the Rhodesian revolution as complete (see Note 62). For Hart, a revolution succeeds when the officials, above all the judges, come to accept a new rule of recognition in their practice. Hart's account is arguably more realistic here: the Pakistani judges in Dosso were not describing a presupposition but deciding whether to accept a new ultimate rule.

Some Indian scholars suggest, as an academic characterisation only, that after Kesavananda Bharati (1973) the basic structure doctrine functions as part of India's ultimate criterion of validity, since even constitutional amendments are tested against it. The Supreme Court has not described the doctrine in Hartian or Kelsenian terms, and the characterisation should be presented in an answer as a scholar's reading, not as a holding. On Hart's view, the Indian rule of recognition is whatever criteria Indian courts and officials in fact accept, which today include the basic structure. On Kelsen's view, the Constitution is the highest posited norm, and the Grundnorm is the presupposition behind it.

9. Evaluation

Hart's account is praised for grounding law in something observable and for explaining the internal attitude of officials; it is criticised for moving from a fact (a practice) to a normative conclusion (obligations) without explaining how. Kelsen's account is praised for keeping validity strictly normative; it is criticised because the Grundnorm has no content of its own and is chosen only after efficacy is established, so fact returns by the back door. Kelsen's own late move to calling it a fiction shows how hard the problem is.

The two may be less opposed than they look. Kelsen's legal scientist presupposes the basic norm only in order to describe a legal order as valid, without endorsing it morally; Hart's observer describes the officials' acceptance of the rule of recognition without sharing it. Both try to explain how law can be described as binding by someone who need not approve of it. They differ on whether that explanation needs a presupposed norm (Kelsen) or only a social fact seen from the internal point of view (Hart). An examination answer that ends on this point shows real understanding of the debate.

10. Frequently Asked Questions

Q. What is the main difference between the rule of recognition and the Grundnorm?
A. The rule of recognition is a social fact: it exists as the practice of officials in identifying the law, and is neither valid nor invalid. The Grundnorm is a presupposed norm: no one makes it, and legal thought assumes it so that an effective order can be treated as valid.

Q. Why does Kelsen refuse to ground validity in a practice?
A. Because he holds that an ought can be derived only from another ought, never from an is. Grounding validity in the mere fact of a practice would, for him, confuse the normative science of law with sociology.

Q. Did Kelsen's view of the basic norm change?
A. Yes. In his classic works he called it a transcendental-logical presupposition. In the posthumous General Theory of Norms (1979) he described it as a fiction in Vaihinger's sense, an assumption known to be unreal but necessary for thought.

Q. What is the difference between a supreme criterion and the ultimate rule?
A. A supreme criterion is one whose rules prevail over rules identified by other criteria, such as legislation over custom and precedent. The ultimate rule is the rule of recognition itself, which is not validated by any further rule and exists only as the accepted practice of officials.

See also: Note 11 (analytical school), Note 12 (Kelsen), Note 61 (Austin versus Hart), Note 62 (Austin versus Kelsen), Note 38 (sovereignty).