Custom is the oldest and, historically, the most important source of law. Long before legislatures drafted codes or judges spun precedent, communities ordered their affairs by usages repeated so consistently that disobedience felt unnatural and obedience felt obligatory. For the judiciary and CLAT-PG aspirant, the topic is deceptively simple: everyone can recite that custom is “ancient, certain and reasonable,” yet the examiner rewards the candidate who can explain why a mere habit becomes a rule of law, how Indian courts treat custom as overriding even the written Smritis, and where the statutory definition in Section 3(a) of the Hindu Marriage Act, 1955 fits into the older common-law learning. This chapter fixes the definition and dissects the nature of custom — its sources of binding force, its relationship with statute and morality, and the evidentiary burden that has sunk many a pleaded custom. For the wider map of the subject, see the Customary Law hub and the companion introduction.

What is custom? The working definition

In its barest sense, a custom is a long-established practice that a community treats as obligatory. The classic juristic formulation belongs to Sir John Salmond, who described custom as “to society what law is to the State.” Each, he said, is the expression and realisation, to the measure of men’s insight and ability, of the principles of right and justice. Custom, on this view, is the embodiment of those principles that have commended themselves to the national conscience as principles of justice and public utility; and of that conscience, national custom is the external and visible sign.

The Indian courts adopted a more operational definition. In Hurpurshad v. Sheo Dyal the Privy Council held that a custom “is a rule which, in a particular family, or in a particular district, has from long usage obtained the force of law.” That single sentence carries the whole topic: a custom is (i) a rule, not a stray instance; (ii) confined to a defined group or place; (iii) sustained by long usage; and (iv) one that has obtained the force of law, so that the community regards it as binding rather than merely convenient. The statutory echo appears in Section 3(a) of the Hindu Marriage Act, 1955, examined below. A practice that satisfies only the first three limbs but lacks the binding quality is usage or habit, not custom in the legal sense — a distinction that separates a legal custom from a conventional one, treated in the chapter on the kinds of custom.

Salmond and the juristic conception of custom

Salmond’s analysis remains the backbone of the Indian syllabus. He divided custom into two great classes. A legal custom operates per se as a binding rule of law, independently of any agreement on the part of those subject to it; its authority is absolute. A conventional custom (often called usage) operates only because the parties have, expressly or impliedly, incorporated it into their agreement; its authority is conditional on acceptance. The trade usage that supplies an unwritten term to a mercantile contract is conventional; the rule that governs succession in a particular caste irrespective of consent is legal. Legal custom Salmond further split into general custom, having the force of law throughout the territory, and local custom, binding only in a defined locality — a taxonomy developed in the sibling chapter on kinds of custom.

The juristic value of this framing is that it isolates the feature that converts practice into law: the sense of obligation, what the civilians called opinio necessitatis — the conviction that the practice is followed because it must be, not merely because it is habitual. A custom of always dining at a particular hour binds no one; a custom that a widow may adopt a son with the consent of her husband’s kinsmen binds the whole community and the courts. Salmond’s insight, repeatedly invoked in Indian judgments, is that custom owes its force not to the will of the sovereign but to the long, uniform and conscientious observance of the people themselves.

The statutory definition: Section 3(a), Hindu Marriage Act, 1955

Indian legislatures have codified the common-law learning. Section 3(a) of the Hindu Marriage Act, 1955 provides that the expressions “custom” and “usage” signify any rule which, having been continuously and uniformly observed for a long time, has obtained the force of law among Hindus in any local area, tribe, community, group or family. The provision then attaches three controlling provisos: the rule must be certain and not unreasonable or opposed to public policy; and, in the case of a rule applicable only to a family, it must not have been discontinued by the family.

Read closely, the section is a statutory crystallisation of every requisite the courts had developed: continuity, uniformity, antiquity (“for a long time”), the force of law, certainty, reasonableness, and conformity to public policy. The definition is mirrored in cognate statutes — Section 3(a) of the Hindu Succession Act, 1956 and Section 3(b) of the Hindu Adoptions and Maintenance Act, 1956 are in identical terms — so a candidate who masters the Hindu Marriage Act formulation has mastered the codified definition across the personal-law statutes. Crucially, the statute does not displace custom; Section 29(2) of the Hindu Marriage Act expressly preserves any right recognised by custom to obtain the dissolution of a marriage, so that a validly proved customary (panchayati) divorce survives the Act. The detailed essentials drawn from this definition are unpacked in essentials of a valid custom.

The Indian peculiarity: custom overrides the written text

The most distinctive feature of Indian customary law is that a proved custom prevails even over the written textual law. The proposition is anchored in the Ramnad case, Collector of Madura v. Mootoo Ramalinga Sethupathy (1868) 12 M.I.A. 397. The zamindar of Ramnad died sonless; his widow, with the assent of her husband’s sapindas, adopted a son. The Privy Council upheld the adoption and, in a celebrated passage, declared that “under the Hindu system of law, clear proof of usage will outweigh the written text of the law.”

This is the inverse of the English common-law instinct, where custom is tolerated only as an exception that proves the rule of the general law. In the Hindu system, the Smritis themselves command that approved usage be respected, so that custom is not a derogation from the law but a recognised source standing above the text where the two conflict. The principle was not a colonial curiosity confined to 1868; modern benches continue to apply it, holding under Section 29(2) of the Hindu Marriage Act that a marriage may be dissolved by a customary divorce where the customary right is clearly established. The lesson for the answer-script is precise: the Ramnad case did not invent a new law of custom; it recognised that, within the personal-law sphere, demonstrated usage is the superior authority. The doctrine’s deepest roots are traced in custom as a source of Hindu law.

Why is custom binding? Competing theories

The nature of custom is illuminated by the jurisprudential debate over why it binds. The Historical School, led by Savigny, held that custom is the spontaneous expression of the Volksgeist — the common consciousness or spirit of the people — and that law is found, not made; the legislator merely formulates what the popular conviction has already settled. On this view custom is binding because it is the authentic voice of the community, and its observance is itself the source of its authority.

The Analytical School, represented by Austin, took the opposite stance. Austin denied that custom is law in the strict sense until a sovereign — through a court or a statute — adopts it. A custom, he argued, is at most “positive morality” until clothed with legal sanction by the State; its binding force is therefore derivative, flowing from the sovereign’s explicit or tacit command. Salmond steered a middle course: custom is a material source from which courts draw rules, but it becomes a binding legal rule only when the requisites of validity are satisfied and a court recognises it. The Indian position is essentially Salmondian and pragmatic — custom is binding of its own force provided it is ancient, certain, reasonable and proved, the courts acting as gatekeepers rather than creators. This theoretical scaffolding is set out more fully in the introduction.

The essential attributes: an overview of valid-custom requisites

The nature of custom is best understood through the conditions a custom must satisfy to be enforced. Blackstone’s classical list required a custom to be immemorial, continuous, peaceable, reasonable, certain, compulsory (obligatory) and consistent. Indian courts compressed these into the trinity repeatedly invoked since Hurpurshad v. Sheo Dyal: a custom must be ancient, certain and reasonable. In Subramanian Chettiar v. Kumarappa Chettiar the Madras High Court restated that a custom, being in derogation of the general rules of law, must be construed strictly and established by clear and unambiguous evidence; it must not be opposed to morality or public policy and must not be expressly forbidden by the legislature.

Each attribute carries a distinct office. Antiquity ensures the practice predates living memory or, in India, is simply of long standing, so that it represents settled conviction and not recent fashion. Continuity and uniformity ensure that the practice has been observed without interruption and without contradictory instances, for a single break or a contrary decision destroys the custom. Certainty ensures the rule is definite in its content and the class it governs. Reasonableness measured at inception, and conformity to public policy and statute, ensure the custom is fit to be enforced today. The full anatomy of these requisites, with the leading authorities on each, is developed in essentials of a valid custom.

Antiquity: how old must a custom be?

The English rule fixed antiquity at a precise point: a custom had to be shown to have existed since “time immemorial,” legally fixed at the year 1189 (the first year of the reign of Richard I), so that the memory of man ran not to the contrary. Indian courts sensibly declined to import this arbitrary date. Here it is enough that the custom be ancient in the sense of long-established; the party need not prove existence from any fixed year, but must show usage of such duration as to evidence that the community has accepted it as governing.

The leading caution comes from Ramalakshmi Ammal v. Sivanantha Perumal Sethurayar (1872), where the Privy Council held that “it is of the essence of special usages, modifying the ordinary law of succession, that they should be ancient and invariable; and it is further essential that they should be established to be so by clear and unambiguous evidence.” The case concerned a special family custom governing succession to an impartible zamindary; the claim failed because the usage was neither ancient nor invariably observed. The decision teaches that antiquity is bound up with invariability: a custom advanced as ancient must be shown to have been continuously and uniformly followed, for occasional or recent observance betrays its want of immemorial standing.

Certainty, continuity and uniformity

A custom that is vague in its content or in the class it governs cannot be enforced, because a court cannot apply a rule it cannot state. The requirement of certainty demands that the custom be definite both in its terms and in the persons or locality it binds. Equally, the custom must be continuous: it must have been observed without interruption, for a discontinuance raises a presumption that the community has abandoned it. Section 3(a) of the Hindu Marriage Act captures both ideas by requiring observance “continuously and uniformly,” and by its proviso that a family custom is unenforceable once it “has been discontinued by the family.”

Uniformity is the partner of continuity. A custom destroyed by contradictory instances is no custom at all; the party must show that the alleged rule has been followed invariably whenever the occasion arose. This evidentiary stringency explains many failures. In Saraswathi Ammal v. Jagadambal, AIR 1953 SC 201, the Supreme Court rejected a pleaded custom of succession among the dasi (devadasi) community for want of satisfactory proof, observing that oral evidence of instances capable of documentary proof cannot safely be relied upon to establish a custom when the best evidence is withheld without explanation. The case is a standing warning that certainty and uniformity are proved by cogent, preferably documentary, evidence of repeated instances, not by assertion. A single judicial decision contrary to the alleged custom, or proof that members of the group have on occasion acted otherwise, is generally fatal, because it shows the supposed rule was never invariably accepted as binding. The certainty requirement also bars a custom whose very content shifts from instance to instance: if the witnesses cannot agree on what the rule is, the court has nothing definite to enforce and the plea fails for vagueness.

Reasonableness and the morality limit

A custom must be reasonable, but the test is applied with two refinements that examiners prize. First, reasonableness is judged as at the origin of the custom, not by contemporary fashion; a custom does not become unreasonable merely because modern sensibility would not adopt it afresh, though it may be struck down if it is positively unreasonable or has become so. Second, the burden lies on the party challenging the custom to show that it is unreasonable; the law leans in favour of an ancient usage rather than against it. The reasonableness limb shades into morality and public policy: a custom that sanctions an immoral or injurious practice cannot stand however ancient.

The morality limit is illustrated by the colonial-era treatment of customs of the dancing-girl communities, where courts refused to enforce usages founded on or facilitating immoral purposes. The same principle underlies the modern refusal to enforce customs offending constitutional values. Although framed in constitutional terms, the reasoning in cases striking down obnoxious customs proceeds on the footing that a custom repugnant to justice, equity, good conscience or public policy forfeits the force of law. A useful caution for the examinee is that reasonableness is not a roving licence for the judge to substitute personal opinion: the courts ask only whether the custom is so opposed to reason that to enforce it would work injustice, and they resolve genuine doubt in favour of the long-observed practice rather than against it. The point connects directly to the discussion in essentials of a valid custom, where the reasonableness and morality tests are examined against the leading authorities.

Custom against statute and the Constitution

However venerable, a custom cannot survive an express statutory prohibition or a constitutional command. Where the legislature has spoken, custom yields; where a custom is expressly abrogated by statute, it ceases to have the force of law. The most instructive authority is Sant Ram v. Labh Singh, AIR 1965 SC 314, where the Supreme Court considered the customary right of pre-emption. The Court held that a custom of pre-emption based on vicinage — the mere fact of being a neighbour — imposed an unreasonable restriction on the right to acquire, hold and dispose of property then guaranteed by Article 19(1)(f), was void under Article 13, and could no longer be enforced, even though it had previously been recognised as a valid custom.

Sant Ram demonstrates two propositions. First, a custom must conform not merely to ordinary statute but to the supreme law; a custom is “law in force” for the purposes of Article 13 and is therefore void to the extent of inconsistency with fundamental rights. Second, the validity of a custom is tested at the time of its enforcement, so that an ancient custom may be extinguished by a later constitutional or statutory norm. Because custom is in derogation of the general law, it is construed strictly, and any conflict with a clear statutory provision is resolved against the custom — subject only to those statutes, like Section 29(2) of the Hindu Marriage Act, that deliberately save custom.

How custom is proved: pleading and evidence

Custom is, in general, a question of fact that must be specifically pleaded and strictly proved; courts do not take judicial notice of a custom until it is so well established by repeated decisions that proof becomes unnecessary. The controlling modern authority is Dr. Surajmani Stella Kujur v. Durga Charan Hansdah, AIR 2001 SC 938. The parties belonged to the Oraon and Santhal Scheduled Tribes, to whom the Hindu Marriage Act did not apply in the absence of a notification, so that their matrimonial relations were governed by custom. The complainant alleged a custom prohibiting a second marriage but pleaded only a reference to a book on the community’s usages. The Supreme Court held that for a custom to acquire the colour of a rule of law, the party claiming it must plead and then prove that it is ancient, certain and reasonable; in the absence of pleading and evidence, reference to a text alone is insufficient, and the complaint failed.

The decision restates the evidentiary architecture of customary law. The custom must be (i) specifically pleaded, with its content and the class it governs identified; (ii) proved by clear and unambiguous evidence of repeated instances, the best evidence being documentary; and (iii) shown to satisfy the requisites of validity. The burden lies on the party asserting the custom, and a custom in derogation of the general law is construed strictly. The same discipline applies to the tribal customs preserved by the Constitution, discussed in tribal customary laws and constitutional recognition.

The nature of custom is sharpened by contrasting legal custom with conventional custom, the distinction Salmond drew and the Indian courts have applied. A legal custom binds absolutely and independently of consent; it is part of the law of the land for the group it governs, and a court must apply it whether or not the parties were aware of it. A conventional custom, or usage, binds only because it is taken to have been incorporated into the parties’ agreement; it supplies an implied term and yields to a contrary express stipulation. A mercantile usage that interest runs at a particular rate, or that a broker is entitled to a particular commission, is conventional; it governs the contract because the parties are presumed to have contracted with reference to it.

The practical consequences differ markedly. A conventional custom cannot override an express term of the contract and cannot operate if it is inconsistent with the manifest intention of the parties; a legal custom suffers no such limitation, subject only to statute and the requisites of validity. Conventional usages also need not be immemorial in the strict sense; reasonable notoriety in the relevant trade or locality suffices. The taxonomy — legal custom subdividing into general and local, conventional custom into general and local usage — is the organising scheme of the chapter on kinds of custom.

Custom in Muslim and tribal personal law

The recognition of custom is not confined to Hindu law. Before codification, custom played a substantial role among Indian Muslims, particularly in Punjab and the north-west, where agricultural communities often followed customary rules of succession in preference to the strict Shariat. The Muslim Personal Law (Shariat) Application Act, 1937 was enacted precisely to curtail this, providing that in the specified matters the rule of decision shall be the Muslim Personal Law (Shariat) notwithstanding any custom or usage to the contrary — an illustration of custom yielding to statute. The interplay is examined in custom as a source of Muslim law.

For the Scheduled Tribes, custom retains its primacy. As Surajmani Stella Kujur confirms, a tribe to which the Hindu Marriage Act has not been extended is governed by its own customs, which must nonetheless be pleaded and proved like any other custom. The Constitution reinforces this autonomy: the Fifth and Sixth Schedules empower Governors and Autonomous District Councils to preserve and regulate tribal customary practices, and Article 13(3)(a)’s definition of “law” as including custom both protects valid customs and subjects them to fundamental-rights scrutiny. The constitutional architecture is set out in tribal customary laws and constitutional recognition.

Exam synthesis: how to write the answer

A high-scoring answer on the definition and nature of custom moves in a settled sequence. Open with Salmond’s aphorism — custom is to society what law is to the State — and the operational definition from Hurpurshad v. Sheo Dyal: a rule which by long usage has obtained the force of law in a family, district or community. State the statutory definition in Section 3(a) of the Hindu Marriage Act, 1955 verbatim, noting that the same words recur in the Hindu Succession Act, 1956 and the Hindu Adoptions and Maintenance Act, 1956.

Then capture the nature of custom through four propositions, each with an authority. One: custom is binding of its own force where valid — the theoretical debate between Savigny’s Volksgeist and Austin’s positivism, resolved on Salmondian lines. Two: in the personal-law sphere custom overrides even the written text — Collector of Madura v. Mootoo Ramalinga Sethupathy (1868) 12 M.I.A. 397, “clear proof of usage will outweigh the written text.” Three: custom must satisfy the requisites of antiquity, continuity, certainty and reasonableness — Ramalakshmi Ammal v. Sivanantha Perumal (1872) and Saraswathi Ammal v. Jagadambal, AIR 1953 SC 201. Four: custom yields to statute and the Constitution and must be strictly pleaded and proved — Sant Ram v. Labh Singh, AIR 1965 SC 314, and Dr. Surajmani Stella Kujur v. Durga Charan Hansdah, AIR 2001 SC 938. Close by linking forward to the essentials of a valid custom, and you have an answer that demonstrates both definition and nature with verified authority.

Frequently asked questions

What is the legal definition of custom?

A custom is a rule which, by long and uniform usage, has obtained the force of law in a particular family, community, locality or district. The classic judicial definition comes from Hurpurshad v. Sheo Dyal, where the Privy Council described custom as a rule that has “from long usage obtained the force of law.” The statutory definition in Section 3(a) of the Hindu Marriage Act, 1955 adds that the rule must have been “continuously and uniformly observed for a long time” and must be certain, not unreasonable, and not opposed to public policy.

Can a custom override the written text of the law in India?

Within the Hindu personal-law sphere, yes. In Collector of Madura v. Mootoo Ramalinga Sethupathy (1868) 12 M.I.A. 397, the Ramnad case, the Privy Council laid down that “clear proof of usage will outweigh the written text of the law.” A proved, valid custom therefore prevails over a conflicting Smriti text. The principle survives today: under Section 29(2) of the Hindu Marriage Act a marriage may be dissolved by a customary divorce where the customary right is clearly established. Custom cannot, however, override an express statutory prohibition or the Constitution.

What is the difference between a legal custom and a conventional custom?

Salmond drew the distinction. A legal custom operates of its own force as a binding rule of law, independently of any agreement; it binds whether or not the parties consented or were aware of it. A conventional custom, or usage, binds only because it is taken to be incorporated into the parties’ agreement; it supplies an implied term and yields to a contrary express stipulation. Trade and mercantile usages are conventional; rules of caste or family succession are legal customs.

How old must a custom be to be valid in Indian law?

Indian law does not adopt the English fixed date of 1189. It is enough that the custom be ancient in the sense of long and continuous standing, sufficient to show settled acceptance by the community. In Ramalakshmi Ammal v. Sivanantha Perumal Sethurayar (1872), the Privy Council held that usages modifying the ordinary law must be “ancient and invariable” and proved by clear and unambiguous evidence; antiquity is therefore bound up with continuity and uniformity of observance.

Can a custom be struck down for being unreasonable or against statute?

Yes. A custom that is unreasonable, immoral, opposed to public policy, or contrary to an express statute or the Constitution is void. In Sant Ram v. Labh Singh, AIR 1965 SC 314, the Supreme Court held that the customary right of pre-emption based on vicinage imposed an unreasonable restriction on property rights under Article 19(1)(f), was hit by Article 13, and could no longer be enforced. The validity of a custom is tested at the time of enforcement, so a later constitutional or statutory norm can extinguish an ancient custom.

Does a court take judicial notice of custom, or must it be proved?

As a rule, custom is a question of fact that must be specifically pleaded and strictly proved by clear evidence of repeated instances; courts do not take judicial notice until a custom is so well settled by repeated decisions that proof becomes unnecessary. In Dr. Surajmani Stella Kujur v. Durga Charan Hansdah, AIR 2001 SC 938, the Supreme Court held that a mere reference to a book on community usages, without pleading and proof that the custom is ancient, certain and reasonable, is insufficient. The burden lies on the party asserting the custom.