Every rule a court enforces has a pedigree. Salmond divides that pedigree into formal and material sources, and the material sources further into legal sources (legislation, precedent, custom — authoritative, claiming recognition as of right) and historical sources (juristic writing, foreign law, morality — persuasive only). For the judiciary and CLAT-PG aspirant the operative trio is legislation, precedent and custom. This article works through each, the hierarchy between them, the machinery of ratio decidendi and obiter dicta, the circumstances that destroy a precedent's binding force, and the conditions a custom must satisfy before a court will clothe it with the force of law.

The phrase "source of law" is ambiguous, so jurists divide it. The formal source is that from which a rule derives its validity and force — for the positivist this is the will of the State expressed through its organs. The material source is that from which the matter, the actual content, of the rule is drawn. Salmond subdivides material sources into legal and historical. Legal sources — legislation, precedent and custom — are recognised by the law itself as authoritative; a litigant may cite them as of right and a court is bound to give them their due operation. Historical sources — the opinions of jurists such as Bentham and Austin, foreign decisions, principles of morality and religion — influence the law's development but have no claim to recognition as of right; they are merely persuasive. The whole apparatus of sources is a sub-field of analytical jurisprudence, and you may find it useful to read it alongside our overview of the schools of jurisprudence and the hub of jurisprudence notes. The practical importance of the distinction is this: a rule traceable to a legal source binds; a proposition resting only on a historical source persuades.

Legislation as a source: meaning and kinds

Legislation, also called statute law, is the deliberate declaration of legal rules by a competent authority — Parliament, a State legislature, or a body to which legislative power has been delegated. Salmond defines it as the making of law by a body competent for the purpose. It is distinguished from precedent and custom in being direct: the rule is stated in advance, in fixed verbal form, before any dispute arises. Jurists classify legislation as supreme (enacted by a sovereign legislature and not subject to repeal by any other body) and subordinate or delegated (made under powers conferred by the supreme legislature — executive rule-making, by-laws of local authorities, autonomous legislation of universities and the like). Subordinate legislation is valid only so long as it stays within the four corners of its parent statute and is liable to be struck down as ultra vires if it strays beyond. A further refinement is the distinction between direct and indirect legislation: when the sovereign legislature itself lays down the rule the legislation is direct, but when it merely authorises some other body to make rules the resulting product is indirect or delegated. Some jurists, notably Salmond, would treat even precedent and custom as forms of legislation in a broad sense, since the State permits them to operate; but the orthodox view confines legislation to the express declaration of rules by an authority constituted for that purpose, and it is this enacted law that has become the standard form of law in the modern State, the earlier forms — precedent, and custom resting on religious faith or revelation — having lost much of their former efficacy. Salmond captured the modern dominance of legislation in his celebrated metaphor: case law is "gold in the mine — a few grains of the precious metal to the ton of useless matter," while statute law is "the coin of the realm ready for immediate use."

Virtues of legislation over precedent

Salmond ranks legislation above precedent for several reasons. First, its abrogative power: a statute can both make new law and abolish old law, whereas a precedent has only constitutive efficacy — it can create good law but cannot of its own force destroy an existing rule; its operation is, in that sense, irreversible. Second, division of labour and efficiency: the legislature concentrates on making law, the executive on operating it and the judiciary on interpreting and applying it, so each function is performed with greater competence. Third, prospective operation: a statute declares the rule before any act falling under it is done, satisfying the natural-justice maxim that law should be knowable in advance — whereas a judicial precedent necessarily creates and declares the rule in the very act of applying it retrospectively to conduct already past, as happened in Rylands v. Fletcher, where the rule of strict liability was announced upon the parties before the court. Fourth, legislation can anticipate the future, providing for circumstances that have not yet arisen and filling gaps the moment a doubt is noticed, while a bad precedent endures until another case happens to come up for overruling. Fifth, superiority of form: a statute is brief, accessible and authoritative, while case law lies buried in the bulky records of litigation.

Precedent: meaning and the doctrine of stare decisis

A precedent is a judicial decision that furnishes a rule for the determination of later cases involving similar facts — "anything said or done furnishing a rule for subsequent conduct." In England a precedent is not merely evidence of the law but a source of it, and courts are bound to follow the law so established. The doctrine of precedent (the principle of stare decisis — to stand by decided matters) has a loose and a strict sense. In the loose sense it means only that reported decisions may be cited and will probably be followed; in the strict sense it means that in defined circumstances a decision must be followed whether the later court approves of it or not. The unwritten or common law of England is, from the thirteenth century onward, essentially a product of decided cases, and the doctrine secures certainty, consistency and equality of treatment. The justifications usually offered are fourfold: it secures certainty, so that lawyers can advise clients with confidence; consistency and equality, since like cases are decided alike; convenience, sparing courts the labour of reasoning every question afresh; and a measure of justice through impartiality, because a judge constrained by prior authority cannot bend the rule to the parties before him. Against these stand its dangers — rigidity, the perpetuation of error until a higher court intervenes, and the sheer bulk of reported decisions through which the rule must be hunted. A precedent that is merely unsatisfactory may be reversed by Parliament in legislating, and judges retain power to overrule their own earlier decisions and so correct mistakes. Its complement is the principle that the rule is found not in the whole judgment but only in its operative core, the analytical identification of which we turn to below.

Authoritative and persuasive precedents

Precedents are broadly of two kinds. An authoritative precedent is one which judges must follow whether or not they approve of it; it is a legal source of law. A persuasive precedent is one which a judge is under no obligation to follow but must take into consideration and accord such weight as it deserves; it is a historical source only. In England the decisions of the House of Lords (now the UK Supreme Court) are authoritative. In India, by virtue of Article 141 of the Constitution, "the law declared by the Supreme Court shall be binding on all courts within the territory of India"; a High Court's decision binds the subordinate courts within its territorial jurisdiction. Persuasive precedents in England include the decisions of foreign courts such as the United States Supreme Court, decisions of superior courts in Commonwealth countries, judgments of the Privy Council, and judicial dicta. For the Indian Supreme Court and High Courts, decisions of the Privy Council, the U.S. Supreme Court and other foreign courts are persuasive only. Courts of equal authority cannot overrule one another; where two High Courts conflict, the anomaly can be resolved only by the Supreme Court.

Ratio decidendi — the binding rule

It is not the whole of a judgment that binds, but only its ratio decidendi — the rule of law on which the decision is actually based, applied to the parties in respect of the live issues argued on both sides. What is decided merely as between the parties binds only them, who are thereafter barred by res judicata from reopening it; but the proposition of law underlying the result binds later courts. The ratio is the rule the judge necessarily treated as a step to his conclusion. Two classic tests are offered for extracting it. Professor Wambaugh proposed the "reversal test": take the proposition of law, reverse it, and ask whether reversal would change the decision — if it would, the proposition is the ratio; if the decision stands regardless, it was not. Dr. Goodhart proposed the "material facts" theory: the ratio is determined by ascertaining the facts treated as material by the judge together with his decision upon them. Both have limits — Salmond thought Goodhart's test more theoretical than practical — and where several judges give separate concurring judgments, Lord Dunedin observed it is not the duty of a later court to labour to extract a single ratio by which to be bound.

Obiter dicta and the technique of distinguishing

Obiter dictum means "something said by the way" — a statement of law not necessary to the decision, made on hypothetical facts or by way of general reasoning. Obiter dicta are not binding; the later court will treat them only as persuasive. The classic illustration is the contrast between Bridges v. Hawkesworth (1851) and South Staffordshire Water Co v. Sharman (1896) (the notes wrongly abbreviate the latter as "S.S. Water Company v. Sharman"). In Bridges a customer found a bundle of banknotes on the floor of a shop; the Queen's Bench held the finder, not the shopkeeper, entitled, the notes having been dropped in a part of the shop to which the public had access. In Sharman the defendant, employed to clean out a pool on the plaintiff company's land, found two gold rings embedded in the mud; the court distinguished Bridges, holding that an occupier in possession of land with the manifest intention to control it and everything in or upon it has a better title than the finder, and so the rings went to the landowner. The two cases together show how a later court, by isolating a material difference in fact, confines an earlier ratio rather than overruling it — the everyday craft of the common law. One exception to the no-binding-force rule is conventional: the dicta of Blackburn J. are followed with peculiar respect.

Circumstances destroying or weakening a precedent

Several circumstances rob a precedent of its binding force. (1) Abrogation by overruling or reversal: a higher court may declare an earlier decision wrongly decided, in which case it becomes null and void and (subject to exceptions) the overruling operates retrospectively, governing intermediate transactions by the new rule. The leading Indian illustration is the Supreme Court's overruling of I.C. Golak Nath v. State of Punjab (AIR 1967 SC 1643) — which had held that Parliament could not abridge fundamental rights — in Kesavananda Bharati v. State of Kerala (AIR 1973 SC 1461), which replaced that view with the basic-structure doctrine. A precedent cannot, however, be overruled by an obiter dictum. (2) Reversal on a different ground on appeal weakens the original decision's authority. (3) A decision given per incuriam — in ignorance of a relevant statute or binding rule — is not binding, the principle settled by the Court of Appeal in Young v. Bristol Aeroplane Co Ltd [1944] KB 718, which also held that the Court of Appeal is bound by its own previous decisions save in defined exceptions. (4) A High Court decision overlooking a Supreme Court precedent is a bad precedent. (5) Inconsistency among a court's own earlier decisions releases it from the conflicting line. (6) Precedent sub silentio — where a point of law was assumed without argument — does not bind, the stock illustration being Gerard v. Worth of Paris Ltd. (7) Decisions of an equally divided court settle nothing and have no precedential force.

Custom: the oldest source of law

Custom is the earliest source of law and, historically, the soil from which much of the common law and the personal laws grew. Salmond writes that "custom is to society what law is to the State" — each is the expression, realisation and measure of a community's insight into right conduct. A custom may be legal or conventional. A conventional custom (usage) operates only because, and to the extent that, the parties have impliedly incorporated it into their agreement; it binds as a term of the contract, not of its own force. A legal custom has the force of law in itself, independently of any agreement, and binds all persons within its ambit. Legal customs are again of two kinds: general customs, which prevail throughout the realm and constitute a part of the ordinary law of the land, and local customs, which obtain only in a particular locality, class or family and operate as an exception to the general law for that limited area. The Historical School, treated in our note on Savigny and Maine, elevated custom to the primary source, Savigny tracing it to the Volksgeist or spirit of the people.

Essentials of a valid custom

Before a court clothes a custom with legal force it must satisfy a series of tests. (i) Immemorial antiquity: the custom must be ancient, observed beyond the memory of any living person; in English law the technical date is the year of legal memory (1189), while Indian courts require proof that it is of long usage, ancient and invariable. (ii) Continuity: it must have been enjoyed continuously and without interruption; a break in observance raises a presumption that it never existed as of right. (iii) Enjoyment as of right (nec vi, nec clam, nec precario): it must have been enjoyed openly, not by force, not by stealth and not by mere licence. (iv) Certainty: a custom that is vague or indefinite fails. (v) Reasonableness — the most difficult requirement — the custom must not be immoral or opposed to public policy, the test being applied as at its origin and measured against the court's notions of natural justice and public utility — a custom valid at its inception is not defeated merely because changed circumstances later make it inconvenient. (vi) and (vii) Conformity with law: a custom must not conflict with the fundamental principles of the law of the land, nor with any statute (a statute can always abrogate a custom, never the reverse). (viii) Consistency with other established customs of the same locality. (ix) Opinio juris sive necessitatis: the necessary mental element — those who observe it must do so under a conviction that it is obligatory, not as a matter of mere choice or convenience.

The hierarchy: how the three sources rank

Where the three sources compete, the order of priority is settled. Legislation prevails over both precedent and custom: a valid statute can abrogate any prior precedent and any custom, however ancient, and a custom inconsistent with a statute is void pro tanto. Precedent prevails over custom in the sense that a custom must run the gauntlet of judicial recognition — it acquires the force of law only when a court, applying the requirements above, accepts it; and a custom contrary to a binding precedent or to the general law will be rejected. Yet the relationship is not purely hierarchical: custom feeds precedent (courts recognise and crystallise customs into rules) and precedent in turn feeds legislation (Parliament codifies settled case law). This dynamic interplay is precisely what the Historical School emphasised against the command theory of the analysts. For the modern State, however, Salmond's verdict holds: legislation is "the most powerful and the latest instrument in legal growth," the others having lost much of their former efficacy.

Judicial decisions in the Indian system

In India the doctrine of precedent operates within a constitutional framework that gives it a sharper edge than the unwritten English original. Article 141 makes the law declared by the Supreme Court binding on all courts in the country, and the Supreme Court's pronouncements therefore have an authority no English decision ever possessed before the Practice Statement era. Three propositions follow. First, only the ratio decidendi of a Supreme Court decision binds; its obiter, however weighty, is persuasive only, though the obiter of the highest court is naturally treated with great respect. Second, the Supreme Court is not bound by its own decisions and may, sitting as a larger bench, overrule them — the device by which constitutional law has evolved, as when the seven-judge view in Golak Nath yielded to the thirteen-judge bench in Kesavananda Bharati. Third, a decision rendered per incuriam or sub silentio, or one in which the point was assumed without argument, does not attract Article 141, for it does not "declare" the law in the relevant sense. Decisions of a High Court bind the courts subordinate to it and, by convention, a single judge is bound by a Division Bench of the same court; a co-ordinate bench in doubt refers the matter to a larger bench rather than dissenting. The decisions of the Privy Council rendered before the Constitution remain persuasive, and those of foreign courts — the U.S. Supreme Court especially on questions of due process and free speech — are frequently cited but never binding. The result is a tightly tiered system in which precedent, anchored to the text of Article 141, functions as a genuinely authoritative source second only to legislation and the Constitution itself.

Interpretation: where legislation meets judicial decisions

Because a statute speaks through words that must be applied to unforeseen facts, the judicial function of interpretation is itself a quiet source of law. The oldest of the interpretive canons is the Mischief Rule, laid down in Heydon's Case (1584) by the Barons of the Exchequer, who directed that for the true construction of a statute four things be considered: what was the common law before the Act; what was the mischief and defect for which the common law did not provide; what remedy Parliament had appointed; and the true reason of that remedy. The judge must then suppress the mischief and advance the remedy. The rule's modern application is illustrated by Smith v. Hughes [1960] 1 WLR 830, where prostitutes solicited men in the street from balconies and windows; the Street Offences Act 1959 penalised soliciting "in a street." Lord Parker CJ, applying the mischief rule, held that since the object of the Act was to enable people to pass along the streets without being solicited, soliciting from a balcony fell within "in a street" and the women were guilty. Interpretation thus shows precedent and legislation operating not as rivals but as partners in the production of the living law — a theme developed further in our notes on the sociological school and law as social engineering.

Frequently asked questions

What are the legal sources of law according to Salmond?

Salmond's three legal (authoritative) sources are legislation, precedent and custom. They are distinguished from historical sources — juristic writings, foreign decisions and morality — which are merely persuasive. Legal sources may be cited in court as of right; historical sources only influence the law's development.

Why does Salmond regard legislation as superior to precedent?

Because legislation has abrogative power (it can both make and unmake law, whereas precedent only creates), it operates prospectively and so satisfies natural justice, it provides for the future and fills gaps at once, and it is brief and accessible. Salmond's metaphor: case law is "gold in the mine" while statute law is "the coin of the realm ready for immediate use."

What is the difference between ratio decidendi and obiter dicta?

The ratio decidendi is the rule of law on which the decision actually rests, applied to the live issues argued, and it binds later courts. Obiter dicta are observations said "by the way" — on hypothetical facts or general reasoning — not necessary to the decision, and they are only persuasive. Wambaugh's reversal test and Goodhart's material-facts test help locate the ratio.

When is a precedent not binding (per incuriam)?

A precedent decided per incuriam — in ignorance of a relevant statute or binding rule — does not bind, as held in Young v. Bristol Aeroplane Co Ltd [1944] KB 718. Other weakening circumstances include overruling or reversal, a High Court decision overlooking a Supreme Court precedent, inconsistency with the court's own earlier rulings, decisions sub silentio, and decisions of an equally divided court.

What are the essential requirements of a valid custom?

A valid custom must be of immemorial antiquity, enjoyed continuously and as of right (openly, not by force, stealth or licence), certain, reasonable, consistent with statute and the general law, compatible with other local customs, and accompanied by opinio juris sive necessitatis — the conviction that it is obligatory rather than optional. A custom conflicting with a statute is void.

How does Article 141 of the Constitution operate as a source of law in India?

Article 141 provides that the law declared by the Supreme Court is binding on all courts within the territory of India, making Supreme Court precedent an authoritative source. Only the ratio decidendi binds; obiter dicta are persuasive. The Supreme Court is not bound by its own decisions and may overrule them — as it overruled Golak Nath in Kesavananda Bharati.