For most of Indian legal history custom was not the rival of statute but its master: the Privy Council famously declared in Collector of Madura v Moottoo Ramalinga that clear proof of usage will outweigh the written text of the law. The modern codes inverted that hierarchy. Once a matter is brought within a statute, custom survives only where the statute chooses to spare it. This chapter maps the full arc of that reversal — from the supremacy of custom under uncodified personal law, through the express override and savings clauses of the modern codes, to the constitutional ceiling that now caps every custom, however ancient. Read alongside the Customary Law hub, it shows how a usage that is valid in the abstract can still be defeated by a section number.
The pre-code position: custom above text
Before codification, custom did not merely fill gaps in the personal law; it sat at the apex of the hierarchy of sources. The governing proposition was laid down by the Privy Council in The Collector of Madura v Moottoo Ramalinga Sethupathy (1868) 12 MIA 397, the famous Ramnad adoption case, where their Lordships held that under the Hindu system of law "clear proof of usage will outweigh the written text of the law." A widow's adoption, made with the assent of her husband's sapindas but without his express authority, was upheld because the usage of the Dravida country sanctioned it even though the smriti texts, read literally, did not. The text was treated not as binding legislation but as evidence of usage, and where a clear contrary usage was proved the text yielded.
This was no isolated dictum. The colonial courts repeatedly treated the digests and commentaries as records of practice rather than as a frozen code, so that a proven local, family or class usage could displace the general rule of the school. The Privy Council's own jurisprudence on antiquity and continuity, examined below, was built precisely because custom carried such weight that its proof had to be carefully policed. The consequence was a legal order in which custom and statute rarely met head-on, because there was as yet little statute on the personal-law side to collide with.
The conflict examined in this chapter is therefore largely a creature of the twentieth century — it arises only once the legislature begins to codify, and in doing so chooses how much room to leave for the usages it found in possession of the field. The codifiers did not write on a blank slate; they wrote over a body of living custom that the courts had spent a century recognising, and the override and savings clauses examined later are best understood as the legislature's considered answer to the question of how much of that inheritance to keep. To understand which customs could even enter that contest, see Essentials of a Valid Custom.
First filter: only a valid custom can contend with a statute
A custom cannot conflict with a statute unless it is first a custom in law. The threshold conditions — antiquity, continuity, certainty, reasonableness and conformity with public policy and morality — operate as a gatekeeper, and many apparent custom-versus-statute disputes are decided at this stage without the court ever reaching the statute. The Privy Council in Hurpurshad v Sheo Dyal (1876) 3 IA 259 described a custom as a rule which in a particular family, caste, community or district has from long usage obtained the force of law, and insisted that it be proved to have been "acted upon in practice for such a long period and with such invariability" as to show acceptance by common consent.
Crucially, Indian courts softened the English requirement of immemorial antiquity. In Mst. Subhani v Nawab AIR 1941 PC 21 the Board held that it is not essential that a custom's antiquity be carried back beyond the memory of man; what antiquity must be shown depends on the circumstances, and certainty and reasonableness remain indispensable. The same decision affirmed that custom is a question of fact, to be established inductively from instances and not deductively from theory. The practical effect is that the party asserting a custom against a statutory rule must discharge a heavy evidentiary burden before the question of statutory override even arises. The component conditions are unpacked in Essentials of a Valid Custom.
Codification and the override rule
The decisive shift came with the codifying statutes of the 1950s, which reversed the Ramnad hierarchy by an express override clause. Section 4 of the Hindu Marriage Act 1955 provides that, save as otherwise expressly provided in the Act, any text, rule or interpretation of Hindu law, and any custom or usage as part of that law, "shall cease to have effect" with respect to any matter for which provision is made in the Act, and any other inconsistent law in force ceases to have effect to the extent of the inconsistency. Section 4 of the Hindu Succession Act 1956 is in materially identical terms for matters of intestate succession.
Two features of this drafting are decisive. First, the override is not at large: it bites only on matters "for which provision is made" in the Act. A custom touching a subject the code leaves untouched is not abrogated by the mere existence of the code. Second, where the override does operate, it is absolute — the custom does not merely yield to a contrary statutory rule but "ceases to have effect" altogether, so that it cannot even be pleaded as a survival. This is the structural inversion of Collector of Madura: custom no longer outweighs the text; the text extinguishes the custom unless the text itself preserves it.
The same legislative technique reappears across the codes — the Hindu Adoptions and Maintenance Act and the Hindu Minority and Guardianship Act carry parallel override clauses — so that the analyst's first question in any conflict is narrow and textual: does the Act make "provision" for this matter, and does it "expressly provide" otherwise for custom? The phrasing repays close reading. "Provision is made" directs attention to the subject-matter the code actually occupies, not to the breadth of its title, so that a custom on an ancillary or untouched point can survive even within a codified field. And "save as otherwise expressly provided" confirms that the savings are exhaustive: a custom not expressly preserved is not impliedly preserved. The override thus operates as a default rule of abrogation against which the savings clauses are read as deliberate, closed exceptions.
Savings clauses: where the statute spares custom
The override is heavily qualified by deliberate savings. The legislature, conscious that it was codifying a living law, carved out specific reservations in favour of usage. The Hindu Marriage Act itself is studded with them. Section 5(iv) and 5(v) prohibit marriages within the degrees of prohibited relationship and between sapindas, but each prohibition is expressly subject to a custom or usage governing the parties that permits such a marriage. Section 7(2) recognises saptapadi but only "where such rites and ceremonies include" it, leaving customary forms of solemnisation intact. Most importantly, Section 29(2) preserves any right recognised by custom, or conferred by special enactment, to obtain the dissolution of a Hindu marriage — the statutory anchor of customary divorce.
These savings show that override and recognition are not opposites but two settings of the same dial. Where the legislature wished custom to survive, it said so in terms; its silence elsewhere is read as an intention to override. The interpretive consequence is important: because the general rule is abrogation and the saving is the exception, the saved custom must be brought squarely within the words of the saving clause and is not extended by analogy. A party relying on Section 29(2) must therefore prove a genuine customary right of dissolution, not merely a community practice of separation. The taxonomy of which usages — local, class or family — can occupy these saved spaces is set out in Kinds of Custom.
Strict construction: custom in derogation of general law
Where a custom is pleaded against the general law — including a codified general rule — the courts construe it strictly. The principle, repeatedly affirmed, is that a custom being in derogation of the general law must be construed strictly and the burden of proving it lies heavily on the party who sets it up. The Supreme Court has applied this to defeat loosely framed or over-broad claims of usage: a custom cannot be extended by analogy beyond the instances actually proved, and must be established inductively, instance by instance, rather than inferred from a general theory.
This canon does real work in conflict cases. A community may prove a custom for one limited situation, yet a litigant will seek to stretch it to cover a different transaction that the statute otherwise governs. Strict construction confines the custom to its proven scope and lets the statute occupy the remainder. Combined with the rule that custom is a question of fact (Mst. Subhani v Nawab), it means that a custom asserted against a code must be both narrowly pleaded and amply proved. The doctrine is the procedural counterpart of the override: even within the spaces the statute saves for custom, the saving is read narrowly and the asserted usage is held to the precise contours it can evidentially support. The deeper jurisprudential character of custom as a binding usage is examined in Definition and Nature of Custom.
Customary divorce: the leading battleground
No area illustrates the interplay better than customary divorce under Section 29(2). The savings clause preserves the custom, but the override and strict-construction canons police it. In Yamanaji H. Jadhav v Nirmala (2002) 2 SCC 637 the Supreme Court held that customary divorce is an exception to the general law of monogamous, court-dissolved marriage, and that a party asserting it must plead and prove the custom as a fact; courts cannot assume its existence. The Court remitted the matter precisely because the alleged custom had not been established by evidence.
The principle was reinforced in Subramani v M. Chandralekha (2005) 9 SCC 407, where the Supreme Court reiterated that the prevalence of a customary divorce in the community, contrary to the general law of divorce, must be specifically pleaded and established by the person propounding it. A deed of divorce executed between the parties, without proof of a community custom sanctioning such dissolution, is insufficient. The result is a stable rule: Section 29(2) does not confer a free-standing right to divorce by private arrangement; it merely refuses to abolish a customary right that is independently proved to exist with all the attributes of a valid custom.
Where the proof fails, the override in Section 4 reasserts itself and the only available route to dissolution is the statutory one. This is the cleanest modern demonstration of how a saving clause, the override and strict construction operate together on a single set of facts. The saving clause keeps the door open; the override stands behind it as the default; and strict construction polices the threshold so that only a genuine, fully proved custom may pass. The practical lesson for litigants is unforgiving: a community's informal practice of separation, a panchayat's sanction, or a registered deed of mutual release will not by themselves satisfy Section 29(2). What must be shown is a customary right of dissolution — ancient, continuous, certain and reasonable — recognised in the community as having the force of law. Absent that proof, the parties remain married in the eyes of the code, whatever they may have arranged between themselves, and any second marriage contracted on the strength of the supposed customary divorce is liable to be held void.
Custom defeated by public policy and morality
Some customs fail before any statutory override is reached, because they offend the internal limits of custom itself. The statutory definition of custom in Section 3(a) of the Hindu Marriage Act codifies the common-law position: a custom or usage is a rule continuously and uniformly observed for a long time which has obtained the force of law among Hindus, "provided that the rule is certain and not unreasonable or opposed to public policy." An immoral usage is therefore not a weak custom that statute overrides — it is no custom at all.
The Madras High Court applied this in Balusami Reddiar v Balakrishna Reddiar AIR 1957 Mad 97, where a custom said to sanction a marriage between a man and his own granddaughter was rejected as repugnant to decency and morality; such a usage, the Court held, could not be recognised even in primitive society, let alone a civilised one. The custom failed on the reasonableness-and-morality limb of its own definition. The lesson for conflict analysis is that the public-policy filter is logically prior to the override: a custom that cannot clear Section 3(a) never reaches the question of whether some other section displaces it. Where, however, a custom is otherwise valid and merely competes with a statutory rule, the override and savings analysis takes over.
The constitutional ceiling: custom as 'law' under Article 13
Above statute sits the Constitution, and custom answers to it directly. Article 13(3)(a) defines "law" to include any custom or usage having in the territory of India the force of law, with the result that a pre-Constitution custom that violates a fundamental right is void under Article 13(1) just as a statute would be. The Supreme Court so held in Sant Ram v Labh Singh AIR 1965 SC 1614, striking down a custom of pre-emption: the reasons that render a statute void for offending Articles 14 and 15 apply with equal force to a custom, because custom is itself "law" within Article 13.
The constitutional ceiling has since been used to subordinate even temple and ritual usages to equality norms. In N. Adithayan v Travancore Devaswom Board (2002) 8 SCC 106 the Supreme Court upheld the appointment of a non-Brahmin as priest, holding that a custom restricting priesthood to a particular caste could not prevail against the constitutional guarantee against discrimination; any custom or usage, irrespective of its antiquity, must yield to constitutional mandates and cannot be permitted to perpetuate inequality.
The hierarchy is now complete and inverted from the Ramnad era: custom below statute, and both below the Constitution. A usage that survives the override and the savings analysis can still be struck down if it collides with Part III. Two refinements are worth noting. First, because Article 13(1) speaks of laws "in force" at the commencement of the Constitution, it reaches pre-Constitution custom directly, while post-Constitution usages are tested as the "law" they purport to be. Second, the constitutional filter is not confined to spectacular cases of caste or pre-emption; it informs the reasonableness limb of the custom's own definition, so that a usage offending equality or dignity will often fail at the threshold under Section 3(a) before Article 13 is formally invoked. The constitutional ceiling and the morality filter thus reinforce one another, the one operating internally to the definition of custom and the other externally as a supreme-law override.
Custom and Muslim law: the Shariat Act 1937
The Hindu codes are not the only statutes that reorder the custom-statute relationship; the Muslim Personal Law (Shariat) Application Act 1937 did so a generation earlier, but in the opposite direction. Before 1937, large Muslim communities — particularly in Punjab and parts of the south — were governed in matters of succession and property by agricultural and tribal custom that often departed sharply from classical Islamic law, frequently to the disadvantage of women. Section 2 of the 1937 Act provides that, "notwithstanding any custom or usage to the contrary," the rule of decision in the enumerated matters — intestate succession, special property of females, marriage, dissolution of marriage, maintenance, dower, guardianship, gifts, trusts and wakfs — where the parties are Muslims, shall be the Muslim Personal Law (Shariat).
The "notwithstanding" formula is the mirror image of the Hindu override: there the code displaced custom in favour of a secular statutory rule, whereas here the statute displaced custom in favour of the personal law itself. In both, the technique is identical — an express legislative command abrogating custom within a defined field — and the same interpretive consequence follows: custom survives only outside the enumerated matters, and a litigant asserting a customary departure must show that the subject lies beyond the Act's reach. The fuller account of how usage was first absorbed into, and then largely abrogated from, Islamic law in India appears in Custom as a Source of Muslim Law.
Custom praeter legem and contra legem
The conflict resolves differently depending on whether the custom operates alongside the statute or against it. A custom praeter legem — beyond the law, supplementing it where the statute is silent — generally survives, because the override clauses bite only on matters for which the code makes provision. A custom contra legem — directly contrary to an applicable statutory rule — is abrogated unless the statute expressly saves it. This distinction explains the otherwise puzzling coexistence of override and recognition within a single Act.
Thus a customary form of marriage solemnisation survives because Section 7 of the Hindu Marriage Act leaves room for it (praeter legem), whereas a custom permitting bigamy does not, because Section 5(i) makes positive provision for monogamy and saves no contrary usage (contra legem). The analyst must therefore characterise the custom precisely: is it filling a statutory gap, or is it contradicting a statutory command? Only the latter triggers the override; the former lives in the interstices the legislature left open. Because the override extinguishes only what it expressly reaches, a large body of customary practice — ceremonial, ritual and procedural — continues to operate beneath the codes without ever coming into conflict with them. The role of custom as a continuing, gap-filling source within Hindu law is developed in Custom as a Source of Hindu Law.
Burden and proof in conflict cases
Because custom in derogation of the general law is construed strictly, the evidentiary architecture of a conflict case usually decides it. The party asserting the custom carries the burden, and must establish it by clear and cogent evidence of repeated instances, not by general assertion. This burden is heavier where the custom is pleaded against a codified rule, because the override is the presumptive position and the custom is the exception the litigant must affirmatively bring home.
Three evidentiary propositions recur. First, custom is a question of fact and must be proved inductively from instances (Mst. Subhani v Nawab). Second, it cannot be extended by analogy beyond the situations actually proved. Third, in the divorce context, a deed or community practice is not self-proving; the existence of a customary right of dissolution must be separately pleaded and established (Yamanaji H. Jadhav v Nirmala; Subramani v M. Chandralekha). The practical upshot is that many custom-versus-statute disputes are won or lost on the pleadings and the quality of the evidence rather than on abstract questions of hierarchy: a poorly proved custom simply collapses into the statutory default.
A decision framework for conflict
The cases yield a sequential framework that can be applied to any custom-versus-statute problem. Step one: is the asserted usage a valid custom at all — ancient, continuous, certain, reasonable and not opposed to public policy or morality under Section 3(a) and Balusami Reddiar? If not, the inquiry ends; there is no custom to weigh. Step two: does a statute make "provision" for the matter? If the field is uncovered, the custom survives praeter legem and there is no conflict.
If the field is covered, step three asks whether the statute expressly saves the custom — as Sections 5(iv), 5(v) and 29(2) of the Hindu Marriage Act do, or as Section 2 of the Shariat Act conspicuously does not. A saved custom survives but is construed strictly and held to its proven scope; an unsaved custom is extinguished by the override in Section 4. Step four, finally, tests even a surviving custom against the Constitution: under Article 13 and Sant Ram v Labh Singh, a custom that violates a fundamental right is void regardless of antiquity or statutory recognition, as N. Adithayan confirms.
Run in order, the four steps reconcile the apparent contradiction between override and recognition and locate every custom in its proper place beneath statute and Constitution alike. The framework also explains why so few custom-statute disputes are decided on grand questions of hierarchy: most fall at step one for want of proof, or at step two because the statute never reached the matter, long before the override or the Constitution is engaged. For the examiner, the discipline is to take the steps in sequence and resist the temptation to leap straight to the override — a custom that is never proved, or that operates in a field the code left open, raises no conflict at all. For the foundational vocabulary that the framework assumes, return to the Introduction or the Customary Law hub.
Frequently asked questions
Did custom always rank below statute in Indian law?
No. Under uncodified personal law custom ranked above the written text. The Privy Council in Collector of Madura v Moottoo Ramalinga (1868) held that clear proof of usage outweighs the written text of Hindu law. The hierarchy was reversed only by the codifying statutes of the 1950s, whose override clauses (e.g. Section 4 of the Hindu Marriage Act) made custom yield to the code.
What exactly does Section 4 of the Hindu Marriage Act do to custom?
It provides that, save as expressly provided in the Act, any custom or usage as part of Hindu law ceases to have effect with respect to any matter for which the Act makes provision. The override is absolute where it applies, but it bites only on matters the Act actually covers; customs on subjects the Act leaves untouched are not abrogated.
How can a custom still survive after codification?
Through express savings. The Hindu Marriage Act preserves custom in specific places — Sections 5(iv) and 5(v) permit marriages within prohibited degrees or between sapindas where custom allows, and Section 29(2) preserves customary rights of divorce. A saved custom survives but is construed strictly and confined to its proven scope.
Why do courts construe custom strictly when it conflicts with a statute?
Because a custom in derogation of the general law is an exception, the burden of proving it lies heavily on the party asserting it, it cannot be extended by analogy, and it must be proved inductively from instances. Mst. Subhani v Nawab (AIR 1941 PC 21) confirms that custom is a question of fact, so a vaguely pleaded or loosely proved custom simply collapses into the statutory default.
Can a valid, ancient custom still be struck down?
Yes — by the Constitution. Article 13(3)(a) treats custom as "law," so a custom violating a fundamental right is void. In Sant Ram v Labh Singh (AIR 1965 SC 1614) a custom of pre-emption was struck down under Articles 14 and 15, and in N. Adithayan v Travancore Devaswom Board (2002) a caste-restrictive temple custom yielded to the equality guarantee.
How does the Shariat Act 1937 fit the custom-statute picture?
Section 2 of the Muslim Personal Law (Shariat) Application Act 1937 provides that, notwithstanding any custom or usage to the contrary, the rule of decision in the enumerated matters for Muslims shall be the Shariat. It is the mirror image of the Hindu override: an express statutory command abrogating custom, but in favour of the personal law itself rather than a secular code.