What is law? — Hart and positivism

This first week asks the fundamental jurisprudential question: what makes something law? The central answer considered here is HLA Hart’s modern legal positivism, especially as developed in The Concept of Law. Hart’s object was not to produce a moral defence of law, nor a manual for judges, but an explanatory account of the distinctive structure of a legal system. His claim is that law is a matter of social sources: whether a norm is legally valid depends on its pedigree within a recognised legal system, not on its moral merit.

That claim needs careful handling. Positivism is often misstated as the view that law has no connection with morality, or that judges never reason morally, or that citizens are morally bound to obey all valid laws. Hart maintained none of these crude propositions. His separation thesis is conceptual and limited: the existence and content of law can be identified without necessarily deciding whether it is morally good, just, or legitimate. A wicked enactment may be law; a morally attractive principle may fail to be law unless it satisfies the system’s criteria of legal validity.

Hart’s distinctive contribution was to move beyond John Austin’s command theory. Austin explained law as the command of a sovereign backed by sanctions. Hart thought this account illuminated some criminal prohibitions but failed to explain large parts of ordinary legal life: powers to make wills, contracts and marriages; rules conferring legislative or adjudicative authority; continuity through changes of government; and the internal attitude of officials who treat rules as standards, not merely as predictions of punishment.

Hart’s replacement is the union of primary and secondary rules. Primary rules impose duties. Secondary rules are rules about rules: they identify valid law, enable change, and provide institutions for adjudication. At the foundation of a mature legal system stands a rule of recognition: a social rule accepted from the internal point of view by officials, supplying criteria by which valid legal rules are identified. In England and Wales those criteria include enactment by Parliament, authoritative judicial precedent, and other recognised sources. The rule of recognition is not enacted by any superior legal authority. It exists as a matter of social fact.

For Cambridge Part II purposes, the point is not to recite Hart’s vocabulary but to test its explanatory power. Does Hart account for constitutional adjudication, human rights interpretation, common law development, and hard cases? Does he succeed in separating legal validity from moral evaluation without making law normatively inert? Does his account survive Dworkin’s attack that law includes principles, not merely rules? A strong essay treats Hart as the beginning of a sophisticated debate, not as a slogan about law and morality.

Modern legal positivism is best understood as a reaction to two traditions: natural law theory and nineteenth-century command theory. Classical natural lawyers, in different ways, connected law to reason, justice, or the common good. On a strong version, an unjust law lacks the character of law in a central or focal sense. Positivists resisted that move. They sought to explain law as an institutional and social phenomenon whose existence can be determined without first resolving contested moral questions.

The immediate English background is John Austin’s The Province of Jurisprudence Determined. Austin’s analysis was deliberately austere. Laws properly so called were commands issued by a sovereign to whom the bulk of a political society habitually obeyed, and who was not in the habit of obedience to anyone else. Sanctions were central: law was a command backed by the threat of evil in the event of non-compliance. Austin’s virtue was analytical clarity. He separated positive law from positive morality and refused to confuse the question whether a rule is law with the question whether it ought to be obeyed.

Hart accepted that ambition but rejected Austin’s architecture. The command model could not explain law’s rule-governed character. It treated citizens as if they were merely obliged, in the sense of being threatened, rather than under an obligation according to a recognised standard. It could not explain power-conferring rules: a rule allowing a person to make a valid will is not naturally described as a threat-backed command. Nor could it explain the persistence of law across sovereign succession. The law made under one monarch or Parliament does not disappear simply because the persons occupying political office change.

Hart’s intellectual setting also included Hans Kelsen’s pure theory of law, which explained legal validity through a hierarchy of norms terminating in a presupposed Grundnorm. Hart shared Kelsen’s anti-natural-law ambition and his interest in legal validity, but grounded the ultimate criterion not in a transcendental presupposition but in social practice: the rule of recognition exists because officials accept and use it.

The mid-twentieth century setting matters. The collapse of legality under Nazi Germany gave renewed force to the question whether extreme injustice deprives enactments of legal status. Gustav Radbruch famously argued after the war that statutory law must yield where injustice reaches an intolerable degree. Hart replied, in his debate with Lon Fuller, that intellectual clarity and moral criticism are better served by admitting that wicked laws may be valid law and then condemning them as wicked. The positivist separation of law and morals was, for Hart, not moral complacency but a way of avoiding false consolation. If evil can be law, then legal validity cannot itself carry moral vindication.

The Cambridge examiner will expect awareness of this background because Hart is often misread when detached from it. He is not simply saying that officials can do whatever they like. He is replacing sovereign-command positivism with a social-rule account of legal systems, while preserving the central positivist thought that legal existence depends on sources, not merits. The question for the week is therefore historically precise: whether Hart’s reconstruction gives a more adequate account of legal order than Austin, and whether it can withstand the natural lawyer and interpretivist objections that followed.

Hart’s theory turns on five ideas: social rules, the internal point of view, the union of primary and secondary rules, the rule of recognition, and the separation thesis.

First, Hart distinguishes habits from social rules. A habit is regular conduct: people go to the cinema on Fridays, or drive a particular route to work. A rule has a normative aspect. Those who accept it use it as a standard for criticism and justification. If an official says that a purported Act is invalid because it has not received Royal Assent, that official is not merely predicting behaviour. She is invoking a standard treated as authoritative within the practice. Hart’s point is not that every citizen morally endorses the legal system. Rather, a legal system exists where the general population broadly obeys primary rules, and officials accept the secondary rules that identify, change and apply law.

Secondly, the internal point of view is Hart’s most important methodological innovation. An external observer may describe legal behaviour in predictive terms: courts usually punish theft; officials usually apply statutes. That misses the way participants understand legal standards. From the internal point of view, a rule is used as a reason for action, criticism and decision. Hart does not say that the internal point of view must be morally sincere. Officials may accept the rule of recognition for many reasons: moral conviction, professional duty, prudence, tradition, self-interest, or simple conformity. What matters is that they treat it as a common standard for official conduct.

The cases are not authorities for Hart’s theory in the way Donoghue v Stevenson is authority for negligence. Jurisprudence uses cases as evidence of legal practice and as pressure points for theory. They show how English law identifies valid sources, treats official power as legally limited, and manages the relation between enacted law, common law principle and morality.

Entick v Carrington remains the classic illustration of legality as source-based authority. State officials could not justify trespass by appealing to executive necessity or public interest. They needed legal authority. For Hart, this illustrates the rule-governed nature of official power. Officials are not merely powerful actors issuing commands; their actions are assessed by standards recognised as law. The case also supplies a useful bridge to rule-of-law theory: legal validity and official authority require ascertainable legal sources.

Shaw v DPP is useful because it exposes the tension between positivism and judicial moralism. The House of Lords upheld a conviction for conspiracy to corrupt public morals. To positivists, the case raises the question whether courts are identifying existing common law or creating offences under cover of moral condemnation. Hart would not deny that common law can be law; precedent and judicial practice are recognised sources. But he would insist that the legal status of a judicially articulated offence depends on the recognised criteria of the system, not simply on the immorality of the conduct.

R v R illustrates legal development by adjudication. The House of Lords abolished the supposed marital rape exemption. A crude positivist might be tempted to say that judges merely discover pre-existing rules or else illegitimately legislate. Hart’s account is subtler: rules of adjudication confer authority on courts, and open texture may require development in hard cases. The theoretical question is whether that development is discretionary law-making, as Hart tends to suggest, or principled interpretation of existing law, as Dworkin would argue.

Factortame, Jackson and Miller are constitutional stress tests. Factortame showed that, while the European Communities Act 1972 remained in force, courts could disapply inconsistent domestic legislation within the area of EU law. A Hartian account explains this not as metaphysical limitation of Parliament but as a change in the United Kingdom’s criteria of recognition brought about through accepted constitutional practice and statute. Jackson contains dicta questioning orthodox Diceyan sovereignty and suggesting that the rule of law may impose deeper limits. Miller reasserts the centrality of statute and parliamentary authorisation in altering domestic rights. These cases invite the question whether the UK rule of recognition is exhausted by parliamentary sovereignty or includes common law constitutional principles.

Ghaidan v Godin-Mendoza is especially valuable for the Hart-Dworkin boundary. Section 3 of the Human Rights Act required an interpretation of rent legislation compatible with Convention rights. The court’s reasoning was morally and purposively charged, but its authority was source-based: Parliament instructed courts to interpret legislation in that way. A positivist can therefore accommodate rights reasoning. The deeper challenge is whether, even without statutory incorporation, principles of equality and dignity would already form part of law. That is the Dworkinian question.

Use these cases selectively. In a Part II essay, cases should not become doctrinal mini-lectures. Their function is theoretical: to test whether Hart can explain legal validity, adjudicative authority, official acceptance, constitutional change and the role of moral principles in legal reasoning.

See practice questions section below.

See further reading section below.