The Triple Talaq Act rests on a design that no other Indian family statute shares. The Muslim Women (Protection of Rights on Marriage) Act, 2019 declares that a husband’s pronouncement of instant triple talaq is void and illegal, which means it changes nothing: the marriage it purports to end continues in law exactly as it stood the moment before the words were spoken. The same statute then makes the speaking of those words a criminal offence punishable with imprisonment for up to three years and a fine. A single short Act, eight sections in all, therefore does two things that pull in opposite directions. It denies the pronouncement every legal consequence a divorce is supposed to have, and it attaches to that legally empty act the most serious consequence the state can attach to any act, the loss of the speaker’s liberty.

Everything argued about this statute, in Parliament, in the press and in the petitions that challenge it, is in the end an argument about that pairing. Supporters say the pairing is the point: a declaration of invalidity without a sanction had already been tried, because the Supreme Court set the practice aside in 2017, and the pronouncements continued, leaving women married in law but abandoned in fact, with no instrument in their hands to compel support. Critics say the pairing is incoherent: a legal system does not ordinarily imprison a man for doing something it simultaneously insists he has not done, and a wife whose interest lies in maintenance, residence and the children’s custody gains nothing from a husband in custody who cannot earn. Both sides agree on the underlying practice. Almost nobody in the debate defended instant triple talaq as a good thing. The disagreement is about the instrument, and it deserves to be stated that precisely, because much public writing on the statute blurs it.

A second confusion runs through the public account and this article exists partly to clear it. The statute did not ban triple talaq. The ban, in the only sense the law recognises, came from the Constitution Bench in Shayara Bano v. Union of India in 2017, which set the practice aside by three votes to two. What Parliament added in 2019 was not prohibition but penalty, together with a subsistence allowance, a custody entitlement and a set of procedural safeguards that summaries of the Act routinely omit and that materially change how the offence works. The safeguards matter to both sides of the argument. They are the answer the statute gives to the charge of a free-standing police power over a family dispute, and they are also the measure critics use when they ask why, if the machinery is so carefully confined to the wife’s complaint, the offence needed to be criminal at all.

This article takes the statute apart in the order a reader needs it. It begins with the practice and the legal landscape before 2017, because the Act cannot be understood without knowing what talaq-e-biddat was and what the courts had already done to it. It then reads the Shayara Bano judgment as what it actually was, three separate opinions that agreed on an outcome and disagreed on almost everything else. It follows the legislative sequence of bills, ordinances and a fresh bill through two Parliaments. It then sets out the Act itself, provision by provision, grouped by function, with the safeguards given the same weight as the offence. The arguments for and against the design are then stated at the strength their proponents would recognise, with the holders named, and the article closes with the remedies a woman in this position actually has across the surrounding statutes and an assessment of what the Act has and has not settled.

The practice and the law before the judgment

Talaq is the husband’s power of unilateral divorce in Muslim law, and it is not one practice but a family of them, distinguished by form and by revocability. In talaq-e-ahsan, the husband pronounces talaq once, in a period of purity, and the pronouncement is followed by the waiting period of iddat; the divorce takes effect only when that period expires, and it may be revoked at any point before it does, expressly or by the resumption of married life. In talaq-e-hasan, the pronouncement is made once in each of three successive periods of purity, and the divorce becomes irrevocable on the third. Both forms build time, and with it the possibility of reconciliation and arbitration, into the structure of the divorce itself. Talaq-e-biddat is the departure from that structure. The word talaq is pronounced three times in a single sitting, or a single pronouncement is made in terms that declare the divorce instant and irrevocable, and the marriage is treated as dissolved on the spot, with no interval for reconsideration and no mechanism for taking the words back even if both parties wish it. Biddat carries the sense of an innovation, and the form has been regarded within the tradition itself as a divorce that is effective in law but disapproved in theology, a divorce in its worst form that nonetheless dissolves the marriage. It was that form, and that form alone, that the litigation and then the statute addressed.

The statutory frame around the practice was thin, and its thinness became legally decisive. The Muslim Personal Law (Shariat) Application Act, 1937, Act 26 of 1937, directed that in questions of marriage and its dissolution, including talaq, the rule of decision where the parties are Muslims shall be the Muslim personal law. The Act codified almost nothing of substance. It pointed the courts to a body of uncodified law, drawn for most Indian Sunnis from the Hanafi school, and it was that pointing, the fact that a statute recognised and enforced the practice, on which the constitutional attack of 2017 would eventually fasten. The Dissolution of Muslim Marriages Act, 1939, Act 8 of 1939, supplied the corresponding route for wives, setting out grounds on which a Muslim woman could obtain a decree of dissolution from a court. The asymmetry is part of the background the reader needs: a husband’s divorce under the uncodified law was an extra-judicial act requiring no court, no ground and no notice, while a wife’s judicial divorce required her to sue and to prove a statutory ground, and the husband’s instant form required not even the discipline of time.

The instant form’s place within the community’s own law was narrower than the public debate usually allowed. Talaq-e-biddat was a doctrine of the Hanafi school, to which most Indian Sunnis adhere, and even within that school it was classified as a form sinful in its manner though effective in its result. The Ithna Ashari law that governs Shia Muslims takes a different position altogether: it requires the pronouncement to be made in a prescribed form in the presence of witnesses and does not treat three pronouncements in a single sitting as an irrevocable divorce at all. The practical consequence is that the practice the statute addresses was, even before 2017, the practice of one school within one community, a fact that mattered to both sides of the later argument. Supporters of legislation used it to deny that the Act legislated against Islam in general. Critics used it in the opposite direction, to ask why a penal statute of the Union should be built around the doctrine of a single school and what it meant for the criminal law to define its offence by reference to a form most of the community’s traditions did not practise. The Act’s definition, which speaks of talaq-e-biddat or any similar form having the effect of instantaneous and irrevocable divorce pronounced by a Muslim husband, is drafted broadly enough to cover the pronouncement wherever made, so the sectarian geography of the doctrine is a matter for the debate rather than a limit on the offence.

The practice had forced itself on public attention in episodes that preceded the petitions and prepared the politics. In 2005 the case of Imrana, a woman in Uttar Pradesh assaulted by her father-in-law, drew national notice when a religious ruling treated her marriage to her husband as dissolved in consequence, and the ensuing controversy, about clerical authority over women’s lives and about a community law administered outside the courts, ran for months in the press. In Vishwa Lochan Madan v. Union of India in 2014, the Supreme Court addressed the parallel system directly and held that a fatwa or a decision of a Dar-ul-Qaza has no legal sanction and cannot be enforced against anyone, that such bodies are not courts and their pronouncements bind nobody in law. The judgment did not touch talaq itself, but it fixed a premise the later litigation used: the only legal consequences in a Muslim marriage are those the law of the land recognises, and a pronouncement’s power over a woman’s status comes from the state law that receives it, not from any authority inherent in the words. That premise is the seed of Justice Nariman’s route in 2017, and the reason the constitutional attack was aimed at the receiving statute of 1937.

Reform inside the courts began well before the Constitution Bench. In Shamim Ara v. State of Uttar Pradesh in 2002, the Supreme Court held that a talaq, to be effective, must be pronounced in the sense the law requires: it must be for a reasonable cause and it must be preceded by attempts at reconciliation between the husband and the wife by arbiters drawn from their families, following the procedure the Quran prescribes. A plea raised for the first time in a written statement in maintenance proceedings, the Court held, is not by itself proof of a divorce. Shamim Ara did not strike anything down, and it operated case by case, but it withdrew judicial recognition from the casual, unexamined pronouncement and it supplied later benches with a bridge: the practice as actually exercised, instantaneous and without arbiters, could be described as one the tradition’s own texts did not sanction.

How did Shamim Ara change the position before Shayara Bano?

It held in 2002 that a talaq is effective only if pronounced for a reasonable cause and preceded by attempts at reconciliation through arbiters from both families, and that a bare plea of divorce in a written statement proves nothing. It invalidated no statute, but it let courts refuse recognition to instant pronouncements case by case.

The maintenance background matters as much as the divorce background, because the two have been entangled in this field for four decades. In Mohd. Ahmed Khan v. Shah Bano Begum in 1985, a Constitution Bench held that the general maintenance provision, Section 125 of the Code of Criminal Procedure, 1973, applied to a divorced Muslim woman like any other citizen, and upheld her claim beyond the iddat period. Parliament answered with the Muslim Women (Protection of Rights on Divorce) Act, 1986, Act 25 of 1986, which built a separate scheme of provision and maintenance for divorced Muslim women. A Constitution Bench in Danial Latifi v. Union of India in 2001 upheld that Act while reading it to require a reasonable and fair provision for the divorced woman’s future, made within the iddat period. The consequence for the present subject is direct. A woman divorced by instant talaq before 2017 entered a maintenance regime that the legislature and the courts had fought over for a generation, and the political memory of 1985 and 1986, of a judgment answered by a statute widely read as a retreat, hung over every later step in the triple talaq story. When the legislative debate of 2017 to 2019 came, both sides argued from that memory: the government, that Parliament had once failed Muslim women by legislating against a judgment; the opposition, that the same Parliament was being asked to legislate past a judgment in the opposite direction and with the community’s own representatives largely against it.

Two features of the instant form explain why it, and not the slower forms, became the object of a constitutional challenge and then of a penal statute. The first is its finality without process. A divorce by talaq-e-ahsan or talaq-e-hasan unfolds over months, in the open, with the involvement the tradition prescribes of arbiters from both families, and either form can be halted by reconciliation. A divorce by talaq-e-biddat required nothing: no ground, no notice in advance, no writing, no witness the law insisted upon, and no opportunity for the wife to be heard, to negotiate the terms of separation, or even to know it was coming. Cases that reached the courts and the press involved pronouncements made in anger, in intoxication, by telephone, by text message and by letter from abroad, and the common thread was that the wife learned of her divorce as a completed fact. The second feature is what followed from the finality. An irrevocable divorce of this kind carries, within the doctrine of the school that recognises it, a consequence for remarriage between the same parties: the divorced wife may not lawfully remarry her former husband unless she has first married another man, that marriage has been consummated, and it has itself ended, the intervening marriage known as nikah halala. Whatever the doctrine’s internal rationale, its practical operation on a woman divorced in a moment of anger, where both parties regretted the pronouncement afterwards, was to place a degrading condition between the family and its own restoration. Campaigners against the practice returned to that consequence repeatedly, because it converted a husband’s unilateral word into a lasting injury that neither party could undo by agreement. The Constitution Bench was not asked to rule on polygamy or on nikah halala, which were the subject of separate challenges, and it confined itself to talaq-e-biddat; but the halala consequence formed part of the factual atmosphere in which the instant form was judged, and it forms part of the reason the practice’s defenders within the community were so few.

The pressure that produced the Constitution Bench came substantially from Muslim women themselves, organised. The Bharatiya Muslim Mahila Andolan, an organisation of Muslim women founded in 2007 and led by figures including Zakia Soman and Noorjehan Safia Niaz, campaigned against instant triple talaq on two grounds held together deliberately: that the practice destroyed women’s security, and that it had no warrant in the Quran, whose model of divorce is staged, spaced and arbitrated. The Andolan surveyed Muslim women across ten States and recorded that a large majority of the women interviewed wanted the practice abolished, and it pressed for a codified Muslim family law. Individual women carried the question into court. Shayara Bano’s husband pronounced talaq on her in 2015, ending a marriage of fifteen years, the pronouncement reaching her by post, and in 2016 she petitioned the Supreme Court to have the practice declared unconstitutional. The facts of the petitioners’ own marriages became the debate’s recurring illustrations, because they displayed the practice’s modes in their starkest form: a marriage ended by a letter sent through the post, in Ishrat Jahan’s case a marriage ended by a telephone call from a husband working abroad, in others by a message on a telephone screen, each without warning, negotiation or any forum in which the wife could answer before the fact was accomplished. When supporters of legislation spoke of a practice that operated like a sword held over a marriage, it was those facts they were describing, and when the Act came to enumerate its modes, spoken, written, electronic or in any other manner whatsoever, it was legislating against the catalogue those petitions had assembled. Petitions by other women, Ishrat Jahan among them, and by the Andolan, raised the same challenge, and the Court also registered a suo motu petition under the title Muslim Women’s Quest for Equality. The suo motu proceeding had an origin worth noting: in October 2015 a two-judge bench of the Court, dealing with a question of Hindu succession, recorded its concern that the promise of a uniform civil code remained unredeemed and that women of different communities continued to suffer discrimination sanctioned by their personal laws, and it directed that a public interest proceeding be registered on the question of gender discrimination within personal law, instant triple talaq among the practices named. The women’s petitions and the Court’s own proceeding met in the Constitution Bench, so the case that resulted was simultaneously a claim by divorced and threatened wives, a campaign’s destination, and the Court’s own reference to itself. The Chief Justice constituted a bench of five judges to decide the question, and it is to that bench’s fractured answer that the statute of 2019 is the sequel.

Shayara Bano: one outcome, three opinions

Shayara Bano v. Union of India, (2017) 9 SCC 1, was decided on 22 August 2017 by a Constitution Bench of five judges: Chief Justice J. S. Khehar and Justices Kurian Joseph, R. F. Nariman, U. U. Lalit and S. Abdul Nazeer. By three votes to two the Court set aside the practice of talaq-e-biddat. That sentence, which is how the case is almost always reported, conceals the feature of the judgment that matters most for everything that followed, including the statute. The three judges in the majority did not share a reasoning. Two of them reached the result through the Constitution’s equality guarantee. One of them reached it through the Quran. And the two judges in the minority, who would have upheld the practice against constitutional attack, themselves ordered that it be injuncted for six months and directed the Union to consider legislating against it. A reader who holds those four propositions separately understands both why legislation followed and why the legislation took the argumentative shape it did.

The first opinion, of Justice Nariman for himself and Justice Lalit, is the one with the longest doctrinal reach. Its opening move was about the source of law. Section 2 of the Application Act of 1937, a statute, directs that Muslim personal law shall be the rule of decision in matters including dissolution of marriage. A practice that a statute recognises and enforces is, to that extent, law in force within the meaning of Article 13(1) of the Constitution, and law in force that violates a fundamental right is void. That move carried the practice past the barrier erected in 1952 by the Bombay High Court in State of Bombay v. Narasu Appa Mali, which had held that uncodified personal law is not law in force for the purposes of Article 13 and so stands outside fundamental rights review. Justice Nariman’s opinion did not need to dismantle that barrier in general, because it routed the challenge through the 1937 statute. On the merits, the opinion held talaq-e-biddat to be manifestly arbitrary: the marital tie could be broken capriciously and whimsically by the husband alone, at his will, without any attempt at reconciliation, and a form of divorce so structured violated the equality guarantee of Article 14. Section 2 of the 1937 Act was accordingly struck down to the extent that it recognises and enforces instant triple talaq. Having decided the case on that narrower ground, the opinion expressly declined to enter the further ground of discrimination under Article 15 that the Attorney General had argued. The doctrinal question whether uncodified personal law, as opposed to a practice routed through a statute, is itself testable against fundamental rights was thus moved but not settled, a distinction the series examines in the complete guide to personal law in India, and it remains one of the largest open questions in the field.

The second opinion, of Justice Kurian Joseph, agreed in the result and disagreed about nearly everything else. It held, after an examination of the Quranic text and the precedent, that instant triple talaq is not sanctioned by the Quran, whose injunctions structure divorce as a spaced and considered process, and that a practice the faith’s own sources do not approve cannot claim the protection the Constitution gives to religion. The opinion drew directly on Shamim Ara, observing that the law laid down there, that a talaq must rest on reasonable cause and follow attempts at reconciliation, already captured the position, and it stated its principle in a sentence that has been quoted ever since: what is bad in theology is bad in law. On this reasoning, the practice fell not because the Constitution overrode religion but because religion, properly read, did not contain the practice at all. The difference between the two majority opinions is not a technicality. Justice Nariman’s route subjects a statutorily recognised practice to equality review and would in principle reach other practices the same way. Justice Kurian Joseph’s route decides this case from within the sources of the faith and reaches nothing beyond a practice those sources condemn. When Parliament came to legislate, it was legislating into a space where the Court had set the practice aside by two incompatible logics, one constitutional and one theological, and where no majority had endorsed any particular legislative design.

The minority opinion, of Chief Justice Khehar for himself and Justice Nazeer, is the third document the reader needs. It held that talaq-e-biddat is a matter of personal law, practised by the Hanafi school for well over a thousand years, and that it forms part of the religious faith of Sunni Muslims of that school, protected by the freedom of religion in Article 25. Personal law, on this view, is not law in force for Article 13 purposes and cannot be tested against the fundamental rights in Part III; the Constitution’s framers had left the reform of personal law to the legislature, pointing to Article 44 and its uniform civil code aspiration as the assigned route. The minority therefore declined to strike the practice down. But it did not leave the matter there. Recording that the practice was, even on the materials before the Court, a matter of serious concern, the minority directed the Union of India to consider appropriate legislation, and in the interim, invoking the Court’s power under Article 142 to do complete justice, it injuncted Muslim husbands from pronouncing talaq-e-biddat until such legislation was enacted, for a period of six months in the first instance. The practical consequence was that for six months after the judgment, all five judges had produced an order under which the practice could not lawfully be pronounced, the majority because it no longer existed in law, the minority because its exercise was suspended while Parliament decided.

What did the All India Muslim Personal Law Board argue in Shayara Bano?

The Board, a respondent in the case, argued that matters of religious practice such as talaq are for the legislature and not the judiciary, that personal law stands outside fundamental rights review, and that the practice should be discouraged from within the community; it told the Court it would issue advisories against instant triple talaq.

The minority’s interim injunction deserves a paragraph of its own, because it was the immediate legislative trigger and its fate illustrates the judgment’s odd geometry. Chief Justice Khehar’s direction suspended the practice for six months to give Parliament room to legislate, a period that would have run to February 2018. Within that window the government introduced the first Bill and the Lok Sabha passed it. But because the majority had set the practice aside outright, the operative legal position never depended on the injunction at all: the practice was void from the day of judgment by the majority’s order, and the minority’s suspension was, as a practical matter, overtaken by the very result it dissented from. Its significance is therefore political and constitutional rather than operative. Two judges who believed the Court had no power to invalidate the practice nonetheless used the Court’s power under Article 142 to suspend it and to direct the legislature to act, which tells the reader something important about the temper of the whole Bench: no judge on it was prepared to leave the practice simply standing. When the government later described the statute as the completion of the Court’s work, it was quoting the dissent as much as the majority, and when the opposition replied that the Court had given no direction for a penal law, it was equally right about the majority. Both descriptions are true of different opinions in the same case.

One analytical consequence of the judgment governs the whole design of the 2019 Act and should be fixed before the statute is examined. Once the practice was set aside, a pronouncement of instant triple talaq had no legal effect. It did not dissolve the marriage. A husband who spoke the words after 22 August 2017 remained, in the eyes of the law, married, and his wife remained his wife, with every claim that status carries. That is why the frequent description of the 2019 Act as the law that banned triple talaq is a category error. There was nothing left to ban. The civil consequence had already been removed, completely and by the highest court. What remained undecided in 2017, and what alone a statute could supply, was a sanction for the pronouncement itself, machinery to convert the wife’s continuing status into money and custody in the individual case, and an answer to the reported fact, on which the government would rest its entire legislative case, that pronouncements continued to be made and women continued to be turned out of their homes in reliance on them, judgment or no judgment. A wife so treated was married in law and abandoned in fact, and the ordinary remedies of a married woman, maintenance proceedings, a protection order, a suit, were the slow instruments the law offered her. The legislative question was whether the criminal law should be added to that list.

The arguments before the Bench and the doctrine left open

The hearing in Shayara Bano ranged well beyond the validity of one form of divorce, and the positions taken before the Bench explain the shape of each opinion. The Bench’s own composition was remarked upon at the time and has entered the case’s folklore: five judges drawn from five different faiths, a Sikh Chief Justice, and Hindu, Muslim, Christian and Parsi judges beside him, constituted to decide the religious practice of one community against the Constitution of all. The composition was not accidental, and it lent the fractured result a particular authority, since no opinion in the case can be read as the view of one faith about another. The petitioner whose name the case carries, the women’s organisation that had campaigned for a decade, the Union of India and the Personal Law Board were all heard over a hearing in the summer of 2017 that ranged from Quranic exegesis to the Constituent Assembly debates. For the petitioners, the case was that a power of instant, unilateral, irrevocable divorce vested in the husband alone was a standing denial of equality within marriage: the wife could neither exercise an equivalent power nor resist its exercise, and a legal order that enforced the result enforced the inequality. The Union of India supported the challenge. The Attorney General argued for the Union that the practice was vulnerable under the equality and non-discrimination guarantees and invited the Court to test personal law practices against Part III of the Constitution more broadly, a submission whose ambition Justice Nariman’s opinion noted and deliberately did not take up, resting instead on the narrower statutory route and on manifest arbitrariness. That judicial economy matters for how the judgment is cited afterwards. Shayara Bano is authority for the proposition that a practice recognised and enforced by the Application Act of 1937 is law in force and may fall for manifest arbitrariness under Article 14. It is not authority for the larger proposition that uncodified personal law as such is subject to fundamental rights review, because no majority adopted that proposition, Justice Kurian Joseph’s concurrence proceeded on theology rather than on Article 13, and the minority rejected the proposition outright. Examination answers and commentary that report the case as having subjected all personal law to the Constitution overstate it by exactly the width of the route the Court declined to travel.

For the All India Muslim Personal Law Board, senior counsel Kapil Sibal argued the case on the terrain the minority eventually occupied: that personal law is a matter of faith protected by Article 25, that its reform is constitutionally assigned to the legislature and to the community’s own processes, and that the Court should be slow to reconstruct by judgment what the framers left to politics. The Board’s position contained a concession that shaped the politics afterwards. It accepted that the practice was undesirable, pointed to its own efforts to discourage it through advisories and through the model marriage contract it had circulated, and argued that discouragement from within, not invalidation from without, was the legitimate instrument. When Parliament later legislated, supporters of the Bill used that concession twice over: the Board had told the Court that legislation was the proper route, and the minority had directed the Union to consider legislating, so the statute could be presented as the judgment’s own unfinished business rather than as a defiance of it. The Board, for its part, maintained the distinction between a law regulating the civil consequences of divorce, which it had contemplated, and a law imprisoning the husband, which it opposed as an interference of a different order. Both readings of the concession are faithful to the record, which is why the argument about who invited the statute has never been conclusively won by either side.

The doctrine of the essential religious practice, by which the Court asks whether a practice is essential and integral to a religion before extending Article 25 protection to it, runs through the opinions without commanding them. Justice Kurian Joseph’s concurrence is in substance an application of that doctrine from inside the faith’s texts: the practice failed the test because the Quran does not sanction it. The minority treated the practice as a constituent of the Hanafi faith as lived for centuries and therefore within the protection. Justice Nariman’s opinion made the doctrine unnecessary to the result, since a law that is manifestly arbitrary falls under Article 14 without any inquiry into its religious centrality. The coexistence of the three approaches in one case is the reason the judgment is taught as a specimen of the Court’s methods and not only as a result. It also explains why the case settled less doctrine than its fame suggests. On the justiciability of personal law, on the place of theology in constitutional adjudication, and on the scope of manifest arbitrariness beyond this practice, the opinions point in different directions, and a later bench wishing to move on any of those questions will find in Shayara Bano materials for more than one answer. The statute of 2019 was written in the shelter of that uncertainty: because the Court had not held that the criminal law could not touch the pronouncement, and had not held that it must, the choice of instrument returned to Parliament with the constitutional questions about it left for a future challenge to raise.

Bills, ordinances and a fresh Parliament: the road to the Act

The legislative sequence that followed the judgment took twenty-three months, crossed two Lok Sabhas, and produced three ordinances, and each stage of it left a mark on the final text. The first Muslim Women (Protection of Rights on Marriage) Bill was introduced in the Lok Sabha in December 2017 and was passed by that House on 28 December 2017. In its original form the Bill declared the pronouncement void and illegal and made it an offence punishable with imprisonment of up to three years and a fine, and it framed the offence in the ordinary uncompromising way: cognizable, and with no special confinement of who could set the process in motion and no provision for compounding. In the Rajya Sabha the opposition parties pressed for the Bill to be sent to a select committee, the government declined, and the Bill was not carried in that House. It remained pending there, and the confrontation over it, between a government describing delay as the denial of justice to Muslim women and an opposition describing the penal clause as unexamined and disproportionate, set the pattern for the two years that followed.

The December 2017 debate in the Lok Sabha deserves recording because the arguments that would run for two years were all made in it on the first day. Shashi Tharoor, for the Congress, supported the objective of ending the practice and opposed the instrument, arguing that a Bill which declared a pronouncement void and then imprisoned the pronouncer punished a man for a marriage that in law continued, that it would leave the wife it claimed to protect without support, and that a measure of this character ought to receive a select committee’s examination rather than a single day’s passage. Asaduddin Owaisi moved amendments to strip the penal clause, on the grounds that desertion and abandonment, the harms at which the Bill was in substance aimed, were civil wrongs across the personal laws and that the criminal sanction selected one community’s form of the wrong for punishment. The government’s reply, made by the Law Minister, was that a judgment without a sanction had already been tested in the months since August and had not stopped the pronouncements, that the House had a duty to Muslim women that committees and delay had historically defeated, and that the safeguards the opposition sought could be considered as the Bill travelled. The House passed the Bill that day. The two Houses were, in effect, arguing about different risks: the Lok Sabha majority about the risk to women if the law stayed declaratory, the Rajya Sabha’s resistance about the risk to principle and to families if a penal law passed unexamined. Neither risk was imaginary, and the statute as finally enacted bears the imprint of both, a penal clause the first House always wanted and a safeguard triad the second House’s resistance forced the government to build.

In August 2018 the Union Cabinet approved a set of official amendments to the pending Bill which rebuilt the offence’s procedural frame, and those amendments, not the original clause, are the ancestor of the safeguards in the Act as enacted. The offence would be cognizable only on information given by the married woman herself or by a person related to her by blood or marriage. It would be compoundable at her instance, with the Magistrate’s permission and on terms the Magistrate determined. And an accused would be released on bail only after the Magistrate had heard the married woman and was satisfied that reasonable grounds for bail existed. The amendments were a direct response to the objection pressed in debate and outside it, that an unrestricted cognizable offence in a marital dispute invited intervention by strangers to the marriage and arrest as a first resort. They did not soften the punishment. They confined the trigger, and they gave the woman, whose complaint alone could start the process, the corresponding power to stop it by compounding.

With the Bill still not carried in the Council of States and Parliament not in session, the President promulgated the Muslim Women (Protection of Rights on Marriage) Ordinance, 2018, Ordinance 7 of 2018, on 19 September 2018, carrying the safeguard amendments into force at once; the constitutional mechanics and limits of legislating by ordinance, a device this sequence used three times in eleven months, are examined in the series article on the ordinance-making power in India. The government justified the step on the ground stated in the ordinance’s own recitals and in the Statement of Objects and Reasons of the Bills: that the practice of instant triple talaq had continued despite the Supreme Court’s judgment, with cases continuing to be reported, so that a deterrent backed by the criminal law could not await the Rajya Sabha’s timetable. A replacement Bill, the Muslim Women (Protection of Rights on Marriage) Bill, 2018, was introduced in the Lok Sabha on 17 December 2018 and passed by it on 27 December 2018. It was listed for consideration in the Rajya Sabha but was not taken up, and as the first ordinance approached its constitutional expiry the President promulgated the Muslim Women (Protection of Rights on Marriage) Ordinance, 2019, Ordinance 1 of 2019, on 12 January 2019, to give the provisions continued effect. When the replacement Bill again could not be taken up in the Rajya Sabha, a third instrument followed, the Muslim Women (Protection of Rights on Marriage) Second Ordinance, 2019, Ordinance 4 of 2019, promulgated on 21 February 2019. The sixteenth Lok Sabha was dissolved on 25 May 2019, and both pending Bills, of 2017 and of 2018, lapsed with it. The law in force throughout this period, it should be noticed, was never in doubt in its content: from 19 September 2018 onward the pronouncement was void, punishable and safeguarded in identical terms, first by ordinance and eventually by Act. What kept failing was not the policy but its conversion into a statute, and the repeated resort to Article 123 to bridge the gap is a separate criticism of the sequence, independent of the merits of the measure, which the closing assessment takes up. The criticism has doctrine behind it. In Krishna Kumar Singh v. State of Bihar in 2017, a seven-judge bench of the Supreme Court held that re-promulgation of ordinances is a fraud on the Constitution and impermissible, that the ordinance power is an emergency power to be used when the legislature cannot meet, and that the executive’s satisfaction in promulgating is open to judicial review. The triple talaq sequence sits close to the line that case drew. Each ordinance in the chain was promulgated while a Bill embodying the same provisions lay unpassed in one House, and the second and third promulgations were defended not on any new emergency but on the ground that the provisions needed continued effect, which is the justification Krishna Kumar Singh treats with suspicion. The government’s answer was that the first ordinance answered a documented continuance of the practice, that every ordinance was laid before Parliament as the Constitution requires, and that the replacement Bills were introduced and passed by one House promptly, so that the executive was bridging a gap in one chamber rather than evading the legislature. The sequence was never itself struck down, and the Act’s eventual passage by both Houses cured the process retrospectively in political if not in doctrinal terms, but the episode is the standard modern illustration of how close a determined government can sail to the re-promulgation prohibition when it commands one House and not the other.

What changed between the Bill of 2017 and the Act of 2019?

The punishment never changed: up to three years and a fine throughout. What changed was the procedure. The Act confines cognizance to information from the wife or her relatives by blood or marriage, makes the offence compoundable at her instance, and conditions bail on the Magistrate first hearing her.

The general election of 2019 returned the government with an enlarged majority in the Lok Sabha, and the measure was the first Bill the new Parliament took up in substance. The Muslim Women (Protection of Rights on Marriage) Bill, 2019 was introduced in the Lok Sabha on 21 June 2019 and passed by that House on 25 July 2019 by 303 votes to 82, the opposition again pressing, and again failing, to have it referred to a committee. In the Rajya Sabha on 30 July 2019 the motion to refer the Bill to a select committee was defeated by 100 votes to 84, and the Bill was then passed, 99 members voting in favour, the government’s task made lighter by the abstention or absence of members of parties that had opposed the measure in principle. The President, Ram Nath Kovind, gave his assent on 31 July 2019, and the Bill became the Muslim Women (Protection of Rights on Marriage) Act, 2019, Act 20 of 2019, published in the Gazette on the day of assent. Section 8 of the Act repealed the Second Ordinance of 2019, and Section 1 provided that the Act would be deemed to have come into force on 19 September 2018, the date of the first ordinance. The retrospectivity is of a particular and limited kind, and it is worth being exact about it. It does not reach back before the ordinances to the period after the judgment when the pronouncement, though void, was not punishable. The criminal law in this field has run continuously from 19 September 2018, first under the ordinances and then under the Act, so that no one is punished by the statute for an act done before the state had made it an offence. What the deeming provision does is weld the ordinance period and the Act period into a single unbroken tenure, which is why anything done under the repealed ordinance is deemed to have been done under the Act.

What the Triple Talaq Act actually says

The Act has three chapters and eight sections, and it can be read in minutes. Its brevity is itself part of the argument about it: supporters point to a short, confined statute aimed at one act; critics point out that the shortness is achieved because the Act borrows all of its machinery, courts, procedure, enforcement, from the surrounding general law and builds almost none of its own. Chapter I contains the title, extent and definitions. Chapter II, headed as the declaration of talaq to be void and illegal, contains the voidness provision and the punishment. Chapter III, headed as the protection of rights of married Muslim women, contains the subsistence allowance, the custody entitlement and the procedural code for the offence. Section 8, on repeal and savings, closes the Act.

Section 1 gives the short title, extends the Act, as enacted, to the whole of India except the State of Jammu and Kashmir, and contains the deeming commencement already described. On the reorganisation of that State into the two Union territories of Jammu and Kashmir and Ladakh at the end of October 2019, the Act was applied to both territories by notification, so the original exception belongs to the Act’s enacted form rather than to its territorial reach in operation. Section 2 carries three definitions and each does work. Electronic form takes its meaning from the Information Technology Act, 2000, Act 21 of 2000, which is how a pronouncement sent by message or email is gathered into the offence without a separate list of technologies. Magistrate means a Judicial Magistrate of the first class exercising jurisdiction under the Code of Criminal Procedure, 1973, Act 2 of 1974, in the area where the married Muslim woman resides; the forum is fixed by her residence, not by the place of the pronouncement or the husband’s convenience, a small provision with real consequences for a woman who has returned to her parental home in another district. And talaq itself is defined as talaq-e-biddat or any other similar form of talaq having the effect of instantaneous and irrevocable divorce pronounced by a Muslim husband. The definition is functional, not formal. It does not count the number of pronouncements or prescribe the words. Any form whose effect is instantaneous and irrevocable dissolution falls within it, and any form that preserves an interval and revocability, ahsan and hasan, falls outside it.

The Statement of Objects and Reasons that accompanied the Bill is worth reading as the statute’s official self-explanation, because it concedes the analytical point on which the whole design rests. It recites the Shayara Bano judgment and its 3:2 result. It records that the judgment, in the government’s assessment, had not worked as a deterrent against the practice, and that cases continued to come to light. It recalls that the Personal Law Board had itself told the Supreme Court that the legislature, and not the judiciary, was the proper author of reform in this field. And it states the Bill’s objects in terms that match the enacted text: to declare the pronouncement void, to make it an offence, to give the woman a subsistence allowance and custody, and to build the safeguards into the cognizance, bail and compounding of the offence. A court construing the Act will read its sections against that statement, and a reader of the debate should notice what the statement does not claim. It does not claim the judgment left the practice lawful. It claims the judgment left it undeterred, which is a different and narrower foundation, and the one on which the penal clause’s defence in any proportionality challenge will ultimately stand or fall.

Section 3 is the declaration: any pronouncement of talaq by a Muslim husband upon his wife, by words either spoken or written or in electronic form or in any other manner whatsoever, shall be void and illegal. Two features of the sentence repay attention. The modes are exhaustive by design, spoken, written, electronic, or any other manner whatsoever, so the old disputes about whether a talaq sent by letter, message or telephone call was effective are swept away at the level of the pronouncement’s validity as much as its punishability. And the section declares the pronouncement illegal as well as void, a pairing that prepares the ground for Section 4: the Act does not merely refuse the act a consequence, it characterises the act as unlawful, and then punishes it. Section 4 provides that any Muslim husband who pronounces talaq upon his wife shall be punished with imprisonment for a term which may extend to three years, and shall also be liable to fine. There is no minimum term and no stated ceiling on the fine, both being left to the sentencing court, and there is no separate provision for a repeat pronouncement. The punishment is of the same order as the terms Parliament has used for other marital wrongs it has chosen to criminalise, a comparison the government pressed in debate and the assessment below weighs.

Section 7 is the procedural heart of the Act, and it is where the safeguards live. It opens with a non obstante clause, notwithstanding anything contained in the Code of Criminal Procedure, 1973, to which the Act’s references are made and whose corresponding provisions in the successor code of criminal procedure, the Bharatiya Nagarik Suraksha Sanhita, 2023, carry the machinery. Clause (a) makes the offence cognizable only if information relating to the commission of the offence is given to an officer in charge of a police station by the married Muslim woman upon whom talaq is pronounced or by any person related to her by blood or marriage. The police power of arrest without warrant therefore exists, but its gate is narrow: a neighbour, an employer, a political organisation or a stranger who learns of a pronouncement cannot set the criminal process in motion under this Act. Clause (b) makes the offence compoundable at the instance of the married Muslim woman, with the permission of the Magistrate and on such terms and conditions as the Magistrate may determine. The woman who alone can start the case can also end it, on terms a court supervises, which converts the complaint into a negotiating instrument she holds. Clause (c) provides that no person accused of the offence shall be released on bail unless the Magistrate, on the accused’s application and after hearing the married Muslim woman, is satisfied that there are reasonable grounds for granting bail. The offence is in substance non-bailable in the Code’s classification sense that bail is not a matter of right, but the governing test is the Act’s own: the woman must be heard first, and reasonable grounds must appear. Read together, the three clauses describe an offence deliberately built unlike an ordinary cognizable offence, and any honest account of the Act, for it or against it, has to work with all three rather than with the punishment clause alone.

What happens to a prosecution begun under the ordinances?

It continues under the Act. Section 8 repealed the Second Ordinance of 2019 and provided that anything done or any action taken under it is deemed to have been done or taken under the Act, so complaints, arrests and proceedings from the ordinance period carried across without a break.

Chapter III’s two entitlements are the provisions a woman in the situation actually uses, and their drafting rewards a close look. Section 5 provides that, without prejudice to the generality of the provisions contained in any other law for the time being in force, a married Muslim woman upon whom talaq is pronounced shall be entitled to receive from her husband such amount of subsistence allowance, for her and dependent children, as may be determined by the Magistrate. The opening words make the entitlement cumulative: it does not displace her claims under any other law, and she may pursue those as well or instead. The Act fixes no amount and no formula, leaving the quantum entirely to the Magistrate, which supporters describe as flexibility fitted to the family’s means and critics describe as the absence of the one thing a summary remedy should supply, a figure or a floor she can point to. The word chosen is also significant. The statute speaks of a subsistence allowance, not of maintenance in the developed sense the case law gives that word, and it speaks of it for the woman and her dependent children together. Section 6 provides that, notwithstanding anything contained in any other law for the time being in force, a married Muslim woman shall be entitled to custody of her minor children in the event of pronouncement of talaq by her husband, in such manner as may be determined by the Magistrate. Custody under the Act is an entitlement of the mother, again with the manner left to the Magistrate, and again framed to override contrary provisions elsewhere. Both sections proceed on the Act’s central premise, that the marriage subsists: they are not the incidents of a divorce, apportioned on a dissolution, but interim and continuing protections for a wife whose husband has attempted to cast her off and failed in law.

The Act among the family-law offences: calibration and the road not taken

The three-year term sits in a family of provisions, and its calibration was argued by reference to them. Cruelty by a husband or his relatives, the offence in Section 498A of the Indian Penal Code, itself carried into the Bharatiya Nyaya Sanhita, 2023, is punishable with imprisonment of up to three years and a fine. Bigamy, marrying again during the lifetime of a spouse where the second marriage is void for that reason, carries up to seven years under Section 494 of the Penal Code. Giving or taking dowry carries a minimum term of five years under the Dowry Prohibition Act, 1961, and causing a dowry death carries a minimum of seven. Against that table, the government argued, three years for a summary repudiation that destroys a woman’s home in a sentence is not an outlier but a midpoint, and the claim that India reserves imprisonment for other communities’ marital wrongs while sparing this one had the direction of the asymmetry backwards. The critics’ reply accepts the table and disputes its lesson. Each of those offences attaches to conduct with a victim who suffers a legal or physical harm the law can point to: cruelty inflicted, a second marriage actually contracted, property extracted, a death. The talaq offence attaches to words whose legal effect the same statute annuls, and the comparison therefore cuts as the two sides choose to frame it: for the government, the harm is the social fact of repudiation, which no annulment reaches; for the critics, the harm is one the law has already declared not to exist, and imprisonment for it stands alone in the table as punishment for an act with no legal consequence.

The road not taken had a name in the debate, and it was the Protection of Women from Domestic Violence Act, 2005. That statute, enacted for women of all communities, answered marital misconduct with a civil armoury: protection officers, protection orders, residence orders securing the woman’s right to live in the shared household, monetary relief and compensation, all obtainable quickly and without imprisoning anyone. A bill on the 2005 model, the critics argued, could have given a woman facing a triple talaq pronouncement an immediate residence order, a maintenance order with the force of the magistracy behind it and a protection order against repetition, and could have attached a criminal penalty where it belongs in such schemes, to the breach of the court’s order rather than to the pronouncement itself. Several opposition members urged precisely that architecture, a civil prohibition enforced by a penal sanction for disobedience. The government’s answer was that the 2005 Act already existed throughout the period in which pronouncements continued, and that its remedies, valuable as they are, had not supplied the deterrent signal the situation required. That answer identifies the real choice Parliament made. The question was never whether to protect the wife, since instruments for that existed and were enlarged by the Act’s own Sections 5 and 6. The question was whether to add the state’s condemnation in the form of an offence, aimed at the pronouncement itself, on top of protection. The Act is the record of that choice, and arguments about it that do not separate protection from condemnation miss the decision the legislature actually took.

The factual dispute beneath the argument

The value dispute dominated the debate, but a factual dispute ran underneath it and should be separated out, because parts of it are resolvable and parts are not. Three factual questions mattered. The first was whether pronouncements continued after August 2017. On that, the record is as good as such records get: cases were reported in the press, were cited in the Statements of Objects and Reasons of the successive Bills, and were not denied by the measure’s opponents, whose argument was about the response rather than the occurrence. The government put the instances that had come to light after the judgment at over a hundred when the first Bill was framed, and cited further instances in the ordinance period. The second question was how prevalent the practice remained relative to divorce generally. Here the record thins. Opposition members, P. K. Kunhalikutty among them, argued in the House that divorce rates among Muslims were low by comparison with other communities and that the practice, confined as it was to one school, touched a small number of marriages, so that a penal statute was a disproportionate response to a marginal phenomenon. The government did not rest its case on prevalence and did not need to: a deterrent statute aimed at an act its proponents regarded as unconscionable does not stand or fall by the act’s frequency, and both sides in truth used the numbers ornamentally. Behind the second question sits a measurement problem that neither side could cure, and it is structural rather than partisan. India has no register of Muslim marriages and divorces from which a rate could be computed. Pronouncements are private acts, frequently oral, and they enter any record only when a wife litigates, complains or is counted by a journalist or a researcher with a sample. Every prevalence figure offered in the debate was therefore an extrapolation from cases that had surfaced, and the two sides’ numbers differed less in their arithmetic than in the meaning each assigned to a surfaced case: for the government, each was an instance of a practice defying the Court; for the opposition, the same instances, set against the community’s population, showed a rarity that a penal statute could not improve upon. The honest position, which the article adopts, is that the occurrence of post-judgment pronouncements is established, their frequency is not measurable on the material publicly available, and arguments that need a frequency, in either direction, should be discounted accordingly.

The third question was what the sanction would do: whether it would suppress pronouncements, displace them into silent abandonment, or arm wives in negotiation. On that question there was and could be no settled record in 2019, and there is none in the statute’s text; it is the question the enforcement pattern over time answers, slowly and without the clean attribution either side would like, because a fall in reported pronouncements is consistent with deterrence, with concealment and with the judgment’s own delayed authority all at once. A reader equipped with that separation, what is established, what is thin and what is unknowable in advance, can follow the debate without being captured by its certainties.

What the pronouncement does and does not do

Consequence	Provision	Practical effect
Effect on the marriage	Section 3	The pronouncement is void and illegal; the marriage subsists and the parties remain husband and wife in law
Criminal liability of the husband	Section 4	Imprisonment for a term which may extend to three years, and liability to fine, on conviction
Who may set the criminal process in motion	Section 7(a)	The offence is cognizable only on information given by the wife or by a person related to her by blood or marriage
Bail	Section 7(c)	No release on bail unless the Magistrate has heard the wife and is satisfied there are reasonable grounds for granting it
Settlement of the case	Section 7(b)	The offence is compoundable at the wife’s instance, with the Magistrate’s permission and on terms the Magistrate determines
Subsistence allowance	Section 5	The wife is entitled to an allowance for herself and dependent children in the amount the Magistrate determines, without prejudice to her claims under other laws
Custody of minor children	Section 6	The wife is entitled to custody of her minor children in the manner the Magistrate determines, notwithstanding any other law
Which pronouncements are covered	Section 2(c)	Talaq-e-biddat and any similar form having the effect of instantaneous and irrevocable divorce; spaced and revocable forms are outside the Act
Forum	Section 2(b)	The Judicial Magistrate of the first class for the area where the wife resides
Period from which the offence runs	Sections 1(3) and 8	Deemed in force from 19 September 2018, the first ordinance; proceedings under the ordinances are deemed proceedings under the Act

Consequences that flow from the marriage subsisting

Because the pronouncement is void, the legal world after it is the legal world before it, and several consequences follow that neither the offence nor the entitlements exhaust. The husband remains bound by every obligation of the subsisting marriage, including the obligation to maintain his wife that the personal law and the general maintenance provision impose on a husband as such, and the wife remains entitled to reside in the matrimonial home with the protections the domestic violence statute gives a woman in a domestic relationship. If the husband, treating his pronouncement as effective, goes through a second marriage, that marriage is void, because a man with a living wife to whom he is still married has no capacity to marry again under the law that governs him in this respect as the courts apply it to a subsisting first marriage, and he exposes himself to the law of bigamy, which the Penal Code states as an offence where a person marries again during the lifetime of a spouse in a case where the second marriage is void by reason of its taking place during the life of that spouse. The same logic governs the wife’s position symmetrically: she cannot remarry on the strength of the pronouncement, and a woman who did so without a real dissolution would find the second union void. If the parties reconcile, nothing needs to be undone: there was no divorce to reverse, no fresh marriage to contract, and the halala barrier described earlier never arises, because it is a consequence of an irrevocable divorce and no divorce occurred. If the parties cannot reconcile, the marriage must still be ended by a lawful route if it is to end at all: a spaced and revocable form of talaq that ripens according to its terms, a negotiated dissolution, or a decree under the Act of 1939 at the wife’s suit. The criminal case, if there is one, proceeds on its own track in the meantime and neither dissolves nor preserves the marriage by its outcome; an acquittal does not make the pronouncement effective and a conviction does not make it more void. Keeping those tracks separate, the criminal, the declaratory and the matrimonial, is the single most useful discipline for a reader approaching any real dispute in this field, and it is the discipline the Act’s critics and supporters both observe when they are being careful.

The offence in the courtroom: elements, proof and the safeguards at work

A prosecution under Section 4 has to prove a short list of facts, and each of them has generated its own practical difficulty. The prosecution must show that the accused is a Muslim husband, that the complainant is his wife, that he pronounced talaq upon her, and that the pronouncement was of a form having the effect of instantaneous and irrevocable divorce. The marriage itself is ordinarily proved by the nikahnama or by the testimony of those present at the marriage, and disputes about the marriage’s existence are rare. The contest concentrates on the pronouncement. A spoken pronouncement is proved by the testimony of those who heard it, and a husband who denies the words puts the wife and her witnesses to their proof in the ordinary criminal way, with the presumption of innocence and the standard of proof beyond reasonable doubt entirely unaffected by anything in the Act. A written or electronic pronouncement travels with its document or its message, and an electronic record must be proved in accordance with the evidence law’s requirements for electronic records, including the certificate the law requires for their admissibility, a discipline the courts applied under the Indian Evidence Act, 1872 and which continues under the Bharatiya Sakshya Adhiniyam, 2023. The definition’s functional cast cuts both ways in the courtroom. It spares the prosecution any need to prove a ritual formula, since any pronouncement whose effect is instantaneous and irrevocable qualifies. It also opens a genuine line of defence: a single pronouncement in the ahsan mode, revocable by design and followed by iddat, is not within the Act at all, and a husband who can show that what he pronounced was a revocable talaq, or that the words were conditional, or spoken in a context the law has historically treated as negating a present intent to divorce, has an answer to the charge that goes to its root. The Act criminalises a specified legal act, not anger, not cruelty and not abandonment, and the trial has to find that act. The trial’s human geometry is the last feature to record. The principal witness is almost always the wife, whose testimony the evidence law has long treated as fully competent against her husband in criminal proceedings, while the accused husband cannot be compelled to testify and will ordinarily conduct his defence through the cross-examination of her account and of the relatives who support it. Because the marriage subsists and the parties’ families remain connected by it, the prosecution witness and the accused are frequently negotiating, through the same proceeding, the terms on which a continuing marriage will be lived or a lawful dissolution arranged, and the compounding power exists precisely because Parliament recognised that geometry. A court trying such a case is therefore doing two things at once that the criminal law usually keeps apart: determining guilt for a past utterance, and hosting the renegotiation of a marriage that the utterance failed to end.

Two boundary questions complete the courtroom picture. The first is place. The Act extends to the whole of India, and its forum rule ties the Magistrate to the wife’s residence, so a pronouncement made by electronic form from outside the country, the letter or message from abroad that featured in so many of the reported cases, presents no shelter: the information is given where the wife resides, the Magistrate there is the Act’s Magistrate, and the electronic form is expressly within Section 3’s modes. The second is sentence. The Act fixes a maximum and no minimum, so a Magistrate who convicts may impose anything from a short term to three years, with a fine whose amount the Act leaves open, and the compounding power means that a large class of cases will end not in sentence at all but in supervised terms. Sentencing in such a structure will vary, as the allowance varies, with the facts and the court, and the variation is a designed consequence of a statute that chose discretion at every point where a tariff was possible.

The safeguard triad operates at three different stages, and its operation is best understood stage by stage. At the threshold, the information gate in Section 7(a) determines whose complaint the police may act on as a cognizable offence under this Act. Information from the wife herself is the paradigm case. Information from a person related to her by blood or marriage extends the gate to the family members most likely to be sheltering her, her parents, siblings and in-laws among them, while excluding the wider public. A pronouncement reported by anyone else, a neighbour, a journalist, a political worker, may be a scandal but it is not, without the wife or a specified relative, a cognizable offence under this statute, and a first information report founded on a stranger’s information is vulnerable at the threshold. At the middle stage, bail, Section 7(c) displaces the ordinary expectation that an offence of this gravity level is bailable as a matter of course. The Magistrate must hear the wife, a requirement that gives her notice and a voice at the first moment the accused seeks his liberty, and must record satisfaction that reasonable grounds for bail exist. In practice the hearing has made the bail stage the first real contest in most prosecutions, the point at which the wife’s position on reconciliation, on the allowance and on the children is first stated to a court, and the point at which settlements that mature into compounding often begin. At the final stage, compounding under Section 7(b) is not a private withdrawal. It requires the wife’s instance, the Magistrate’s permission and terms the Magistrate determines, so a compromise extracted by pressure can be examined by the court that is asked to endorse it, and terms addressing the allowance, the custody and the parties’ future conduct can be built into the order that closes the case. None of this makes the offence gentle, and it is not described here as doing so. It makes the offence wife-centred at every stage at which the state’s power is engaged, which is the feature both the government’s defence of the Act and the critics’ rejoinder, that a wife-centred civil remedy would have done the same work without the prison, must take as their common starting point.

The wife as gatekeeper: the Act’s theory of the complainant

The decision to confine the complaint to the wife and her relatives is the Act’s most considered design choice, and it embodies a theory about marital offences that contrasts with the general law’s usual approach. Under the cruelty provision of the Penal Code, information about the offence may come from a wider circle and the machinery, once engaged, has historically been difficult for the complainant to disengage, which is why the courts built an elaborate jurisprudence of quashing matrimonial prosecutions on settlement. The triple talaq offence inverts the sequence. It begins narrow, in the wife’s hand or her family’s, and it builds the exit, compounding on supervised terms, into the statute itself rather than leaving settlement to the High Courts’ inherent powers. The theory is that in a subsisting marriage, unlike a dissolved one, the complainant and the accused must usually go on living within the same family economy after the case, so the law’s pressure should be hers to apply, to hold and to release. The theory has a cost that the design accepts openly. A wife under pressure from the same family that shelters her may never complain at all, and a practice the state has declared criminal will then run unreported in precisely the households where the wife is weakest. Supporters answer that the alternative design, an offence any neighbour or official could trigger, would expose the family to interventions the wife never sought and would be worse. The Act’s answer to the dilemma is the relatives’ gate: her parents and siblings, the people to whose house she characteristically returns after a pronouncement, can set the process in motion even where she herself hesitates. The gate is therefore best read as a judgment about where, in the sociology of the event, the decision to prosecute is most safely lodged, and it is a judgment both sides can describe accurately while valuing it oppositely.

Subsistence and custody: the two entitlements in practice

Section 5’s subsistence allowance is a summary remedy, and its strengths and limits both follow from that character. It is summary because the same Magistrate who is seized of the criminal proceeding, or of the wife’s application in connection with the pronouncement, determines the amount, without the apparatus of a civil suit. It is limited because the Act supplies no scale. A Magistrate fixing the allowance has the husband’s means, the wife’s needs and the children’s needs before the court in whatever evidence the parties bring, and the resulting figures vary as such discretionary figures always vary. The section’s relationship to the general maintenance law is its most frequently misunderstood feature. The allowance is declared to be without prejudice to the provisions of any other law, so it neither replaces nor caps a claim under the general maintenance provision of the criminal procedure law, which a wife, divorced or not, may pursue on its own terms and with its own developed case law on quantum. The two claims answer different questions. The general provision asks whether a husband with sufficient means neglects or refuses to maintain his wife. The Act’s provision attaches to the pronouncement itself, as one of its consequences, and is available to a woman who remains a wife in law and whose complaint is that her husband attempted to cast her off. A practitioner advising in the situation will ordinarily run both, the Act for its speed and its connection to the proceeding already on foot, the general provision for the depth of its jurisprudence.

Section 6’s custody entitlement has to be read with the same care. It entitles the wife to custody of her minor children in the manner the Magistrate determines, notwithstanding anything in any other law. The notwithstanding clause overrides the personal law’s custody allocations for the purposes of the Act’s own order, and the entitlement is the mother’s by right within the proceeding. The manner of custody, the living arrangements, access for the father, the children’s schooling and residence, remains with the Magistrate, and the general law’s governing principle, that the welfare of the minor is the paramount consideration, which the courts administer under the Guardians and Wards Act, 1890 and the personal law alike, is the principle a Magistrate will in practice bring to the manner of the order even under a section that does not recite it. Custody under the Act is best understood as an immediate protective placement attached to the pronouncement proceeding, not as the final adjudication of guardianship between the parents for the whole of the children’s minority, and a parent dissatisfied with the long-term arrangement retains the ordinary guardianship forum. These two sections, read as machinery rather than as slogans, show the Act doing its most defensible work: converting the wife’s legal position, that her marriage subsists and the pronouncement was an offence against her, into interim money and interim custody in a single forum, at a speed the ordinary civil process rarely matches. The criticism that survives that reading is the one already noted, that the quantum of the allowance is unanchored and that the earner against whom it is fixed may be in the custody the same proceeding imposes. Both criticisms go to calibration rather than to existence, and both were available to Parliament in the figures and formulas it chose not to enact.

Regulating talaq elsewhere: the comparison the debate used

Both sides in the parliamentary debate reached for comparative law, and the comparison deserves a section of its own because it supports more than one conclusion. The government’s exhibit was the Muslim Family Laws Ordinance, 1961 of Pakistan, which Bangladesh retained, under which a husband who pronounces talaq in any form must send written notice to the chairman of the local Union Council and a copy to the wife, whereupon an arbitration council is constituted to attempt reconciliation, and the talaq does not take effect until ninety days have passed from the notice; a pronouncement in defiance of that procedure attracted penal consequences under the Ordinance itself, and a talaq not notified could be treated as ineffective. The minister’s argument from the comparison was that regulation of talaq, including regulation with teeth, is the norm across much of the Muslim world and not an Indian eccentricity directed at a minority. Sri Lanka’s Muslim Marriage and Divorce Act routes Muslim divorces through Quazi courts, so that dissolution is a supervised judicial act rather than a private utterance. Tunisia’s Code of Personal Status of 1956 went furthest in the region’s legal family, making divorce available only through a court for both spouses alike. Egypt took the interpretive route a generation earlier: its legislation of 1929, which several other states followed, provided that a talaq accompanied by a number, expressly or by implication, effects only a single revocable divorce, so that the triple pronouncement in one sitting is dissolved as a legal event into the first of the spaced pronouncements the tradition approves. The Union of India had placed this comparative material before the Constitution Bench in Shayara Bano itself, in an affidavit collecting the regulatory practice of other states, so the comparison was part of the case’s record before it was part of the Parliament’s rhetoric. The comparative record therefore shows a spectrum of instruments, from proceduralisation, which slows the divorce and surrounds it with reconciliation, through judicialisation, which moves it into a court, to the Indian choice of 2019, which left the divorce void and penalised the utterance.

The critics’ use of the same record is equally instructive. The Pakistani and Bangladeshi model, they pointed out, regulates the divorce itself: it does not pretend the pronouncement is nothing and then punish the nothing, it makes the pronouncement the start of a process at the end of which a real divorce, with its incidents settled, may occur. A husband in Karachi who complies with the Ordinance ends his marriage lawfully after ninety days and an attempt at reconciliation. A husband in India who pronounces the same words remains married and becomes a criminal. The comparison, on the critics’ reading, shows that the family of Muslim-majority states the government cited chose to govern the act by channelling it, while India chose to govern it by forbidding a form that the courts had already disabled and attaching imprisonment to the attempt. The deeper point the comparison exposes is that the Indian statute is not, despite its long title, a divorce reform law of the ordinary kind. It does not construct a procedure by which a Muslim husband may divorce. The routes to dissolution in Muslim law after 2019 remain what they were: the spaced forms of talaq outside the Act’s definition, divorce by mutual consent and by the wife’s delegated or negotiated routes, and the wife’s suit under the Act of 1939. The statute of 2019 is an offence and two entitlements wrapped around a void act, and the comparison with jurisdictions that built procedures throws that distinctive, narrow character into relief.

The case for the statute, at its strongest

The government’s case, made in the House by the Law Minister, Ravi Shankar Prasad, who piloted the Bill in both Parliaments, began from the record after the judgment. The Supreme Court had set the practice aside in August 2017, and pronouncements continued to be reported, of wives divorced by message, by letter and by utterance, turned out of their homes by husbands who treated the words as effective whatever the Court had said. When the first Bill was framed the government put the reported instances since the judgment at over a hundred. On that record, the supporters’ argument runs, a declaration without a sanction had been given its trial and had failed the women it was meant to protect, because the audience for the pronouncement, the husband, his family and the local community, did not take its instructions from the law reports. The criminal law communicates in a register the declaration could not reach. It tells the husband in advance that the words will cost him his liberty, and it tells the wife that the state stands behind the judgment’s promise that her marriage cannot be ended by them.

The second strand of the case is about leverage within the marriage rather than punishment after it. Because the marriage subsists, the wife’s real needs are money, a roof and the children, and the Act is built to give her instruments toward all three. Her complaint alone can start the process, so the threat of a complaint is hers to deploy or withhold. Compounding lies at her instance, so a husband who wishes the case ended must come to terms she accepts and a Magistrate approves, terms that can be expected to address the allowance and the children. The subsistence and custody sections then give the Magistrate summary powers over precisely those two subjects in the same proceeding. Supporters including Meenakshi Lekhi, who spoke for the Bill in the Lok Sabha, presented the measure in these terms: not as a law about religion, in which the state would have no business, but as a law about gender justice and the dignity of Muslim women, a class of citizens entitled to the same protection against summary repudiation that the Constitution promises everyone. The Bharatiya Muslim Mahila Andolan, whose campaign had carried the question to the Court, supported legislation to close the gap the judgment left, on the ground its members had argued for a decade: that the pronouncement’s social power survived its legal death and had to be met by the state.

The third strand answers the charge of singling out one community, and the Law Minister made it expressly in the Rajya Sabha debate. Indian family law is not innocent of the criminal sanction elsewhere. The Dowry Prohibition Act and the cruelty provision of the Penal Code punish conduct inside Hindu and other marriages with imprisonment; bigamy is an offence; Parliament has repeatedly used criminal law to police marital behaviour it judged oppressive, and the sky did not fall. The minister also pointed outward: talaq is regulated by law in a large number of Muslim-majority states, Pakistan and Bangladesh among them, where a husband’s divorce takes effect through a statutory process involving notice and an interval for reconciliation, so that a state which penalises the instant form is not legislating against Islam but alongside the practice of much of the Islamic world. Finally, supporters rest on the safeguards as the proof of tailoring. A genuinely communal or careless statute, they say, would have made the offence cognizable at anyone’s instance and non-compoundable. This one confines the complaint to the wife and her family, lets her settle it, and makes the Magistrate hear her before bail. The design, on the supporters’ account, is a deterrent held in the wife’s hand, and it should be judged as that, not as an engine of mass prosecution.

The case against the statute, at its strongest

The opposition case also begins from the judgment, and reaches the opposite conclusion from the same fact. If the pronouncement is void, the critics argue, then the husband who speaks the words has done an act without legal effect, and to imprison him for up to three years for it is to punish a nullity. Asaduddin Owaisi, the AIMIM member who led the attack on the penal clause in the Lok Sabha and moved amendments against it, put the objection in constitutional terms: the Bill, he argued, offended the equality and religious freedom guarantees, and it visited on Muslim husbands a punishment for desertion-like conduct that the law visits on no one else, since a husband of any other community who abandons his wife, however brutally, faces maintenance proceedings and matrimonial litigation but no freestanding offence of abandonment carrying three years. The asymmetry argument is the critics’ strongest doctrinal card, and it is sharpened by the immediate context in which the statute was passed: a period in which the Supreme Court had decriminalised adultery and consensual same-sex relations, reading the criminal law out of intimate life, while Parliament was writing it into this one corner of it.

The second strand is practical, and it is the one voiced most often by the women and lawyers who work in the field rather than in the chamber. A husband in custody cannot earn, and a husband who cannot earn cannot pay the subsistence allowance the same Act promises. The wife who complains sets in motion a process whose success, measured by her own interests, would remove her family’s earner for up to three years while her allowance claim is determined against a man without an income. Ghulam Nabi Azad, leading the opposition in the Rajya Sabha, made the political version of the point in that House: that a Bill titled as the protection of rights on marriage would operate in practice as an instrument for the destruction of the families it claimed to protect. The compounding clause is the supporters’ answer, and the critics reply to it directly: compounding under the shadow of arrest is negotiation with a weapon on the table, it locates the case’s end in the Magistrate’s discretion rather than the wife’s right, and it does nothing for the woman whose husband simply leaves, pronouncing nothing, a route the Act does nothing to close and may even reward by making the pronouncement the only dangerous way to abandon a wife. Desertion without talaq, the critics observe, attracts no provision of the Act at all.

The third strand concerns the community’s own voice and the manner of the law’s making. The All India Muslim Personal Law Board opposed the legislation as an interference in personal law that the judgment had made unnecessary, and argued that reform of divorce practice should come through the community’s institutions and the advisory route it had promised the Court. Opposition parties including the Congress, which supported the end of the practice and had voted for the judgment’s objective in principle, opposed the imprisonment clause and demanded throughout that the Bill be sent to a select committee for the scrutiny its penal design required; that demand was refused in both Parliaments, and the measure was carried instead, for most of its life, by ordinance. Members across the opposition, N. K. Premachandran among them, described the repeated re-promulgation of ordinances to sustain a contested penal law against the recorded resistance of one House as a fraud on the Constitution, borrowing the phrase the Supreme Court itself had used for ordinance abuse in earlier doctrine. That procedural objection stands independently of the merits: a reader may believe the offence is justified and still conclude that a statute of this sensitivity, aimed at the intimate law of one community, was owed the slower discipline of committee evidence and clause-by-clause examination in both Houses, and did not receive it.

Legal scholarship divided along lines that did not follow the community’s own division, and the critics’ side of it deserves naming in its academic form because it shaped the petitions that followed. Faizan Mustafa, the constitutional law scholar and vice-chancellor, argued in print that criminalisation was a legislative error: that the judgment had already given women the only legal consequence that mattered, the survival of the marriage with its claims, and that a penal provision would be used in the heat of matrimonial breakdown in ways that would leave the woman worse off, destroying the reconciliation the voidness of the pronouncement had made possible. Flavia Agnes, the lawyer and women’s rights advocate who had worked with Muslim women’s groups for decades, made the allied argument from practice: that women in her casework wanted maintenance, residence and safety enforced, that the criminal process is the instrument least fitted to deliver any of the three, and that the Act’s true constituency was electoral rather than female. Supporters answered the scholarship as they answered the politicians, by pointing to the safeguards and to the record of continued pronouncements, and by asking what remedy, on the critics’ own account, a wife abandoned in reliance on a void pronouncement is to use against a husband who simply ignores a maintenance order, a contempt the civil process handles slowly where it handles it at all. The exchange is recorded here because it marks the boundary of the Act’s defence: the design can be justified as leverage and as deterrence, and the critics’ best point is that leverage and deterrence are descriptions of a negotiating weapon, and that the state hands such weapons to one spouse in a marriage with consequences it cannot supervise case by case.

Beneath the three strands sits a disagreement about what the criminal law is for, which no evidence can settle. Supporters treat punishment as the state’s loudest available statement that a practice is over, and regard the wife’s bargaining power under the Act as a benefit in itself. Critics treat punishment as a tool that must attach to a harm the law recognises, and regard a void act as incapable of carrying it. Both positions were held in the debate by people who wanted the practice ended. The article records the disagreement in that form because it is the true one, and because the petitions that later carried the same arguments to the Supreme Court were framed in its terms.

What the Act leaves untouched

The Act’s narrowness is a deliberate feature, and an accurate map of the field has to mark the ground it does not occupy. It does not touch polygamy. A Muslim husband’s capacity to take a second wife during the subsistence of the first marriage, governed by the personal law, stands exactly where it stood, and the separate challenges to that practice were not before the Bench in Shayara Bano and are not answered by this statute. It does not touch nikah halala, the intervening marriage doctrine described earlier, which remains as the personal law states it; the irony that a void pronouncement cannot trigger the halala consequence at all, since no divorce has occurred, is one of the Act’s quieter effects, because a wife whose husband pronounces instant talaq and repents is, in law, simply still married to him, with no intervening-marriage barrier between the couple and the continuation of their own marriage. It does not create any route by which a husband may obtain a divorce, nor any ground on which either party may sue for one; the Dissolution of Muslim Marriages Act, 1939 remains the wife’s judicial route and the spaced forms of talaq the husband’s extra-judicial ones. It does not codify Muslim family law, supply a marriage registration regime, set a minimum age, or address maintenance after an actual divorce, which remains the province of the Act of 1986 as read in Danial Latifi and of the general maintenance provision. It also leaves the statute book’s older pieces exactly where they were, which a reader of the bare Acts should hold in mind. The Application Act of 1937 still stands, its Section 2 read subject to the striking down of 2017, so that the direction to apply Muslim personal law in matters of dissolution continues for every form of talaq except the one the Court set aside and the Act punishes. The Act of 2019 repeals nothing but the ordinance it replaced. The result is a field in which the governing texts are a two-section direction from 1937, a grounds statute from 1939, a maintenance statute from 1986 read through a judgment of 2001, a Constitution Bench decision of 2017 in three opinions, and an eight-section penal and protective statute of 2019, and in which no single document states the law of Muslim divorce whole. That fragmentation is the standing argument of the codifiers, within the community and outside it, and the Act, by choosing to be a fragment aimed at one practice, declined to answer it.

And it does not pronounce on the uniform civil code, whose shadow lay across the whole debate: the government presented the Act as a step of gender justice within personal law rather than as a move toward a common code, the opposition suspected it as the opposite, and the text itself, confined to one practice of one community, is consistent with either story. A reader who understands what the Act refuses to do understands why the quarrel over it was so intense relative to its eight sections: it was fought as a precedent about the state’s authority over personal law, not as the modest procedural measure its length suggests.

The remedies around the Act: what a woman in this position actually has

The Act does not stand alone in the situation it addresses, and a reader who wants the usable picture needs the surrounding remedies in one view, because the Act’s two entitlements are deliberately cumulative with them. A woman whose husband pronounces instant talaq remains his wife in law, and she may proceed on that footing. She may give information to the police under the Act, if she chooses, as may her relatives by blood or marriage on her behalf. She may apply to the Magistrate under the Act itself for the subsistence allowance and for custody of her minor children. In parallel, and without prejudice to the Act, she retains the general maintenance remedy, the provision of the criminal procedure law, Section 125 of the Code of Criminal Procedure, 1973 and its successor provision in the Bharatiya Nagarik Suraksha Sanhita, 2023, under which a husband with sufficient means who neglects or refuses to maintain his wife may be ordered to pay her a monthly allowance. That general provision is the one the Shah Bano litigation was fought over, and the article on Shah Bano and the legislative response traces how a maintenance claim by a divorced Muslim woman produced first a Constitution Bench and then a statute. The statute it produced, the Muslim Women (Protection of Rights on Divorce) Act, 1986, governs the distinct situation of a woman who has actually been divorced, and supplies, as Danial Latifi read it, a reasonable and fair provision for her future made within the iddat period, together with her mahr and the properties given to her at marriage; it is the neighbour provision, not a rival, since the woman under the 2019 Act has not been divorced at all.

Her position within the subsisting marriage carries one further asset that is often overlooked because it belongs to the personal law rather than to any statute in this story. The mahr, the dower fixed at the marriage, is a debt owed by the husband to the wife, enforceable by her as such, and where it remains unpaid it is recoverable independently of any question of divorce; a wife negotiating from within a marriage her husband has failed to end is, on the personal law’s own terms, frequently his creditor as well as his spouse. The pronouncement does not disturb that position either. Nothing in the Act touches the dower, the wife’s property, or her right to the marriage’s gifts and effects, all of which the law treats as hers. The point completes the picture the Act’s structure implies: the pronouncement’s legal harvest, for the husband who makes it, is a criminal exposure and two summary liabilities, while every pre-existing right of the wife, in maintenance, in dower, in property and in the children, survives intact and is in several respects easier to enforce because the marriage he tried to end is the foundation on which those rights stand.

Two further routes deserve a place in the map. Where the marriage itself is to be ended at the wife’s instance, the Dissolution of Muslim Marriages Act, 1939 gives her grounds on which a court may decree dissolution, including the husband’s cruelty, his failure to maintain her and his failure to perform his marital obligations, so that the wife who wants out of the marriage, rather than protection within it, has a judicial door of her own. Where the conduct around the pronouncement involves violence, threats or dispossession, the Protection of Women from Domestic Violence Act, 2005 supplies protection orders, residence orders and monetary relief through a civil process designed for speed, and the guide to the domestic violence statute sets out that machinery in full; its definition of domestic relationship and its remedies are independent of the marriage’s personal law. The practical sequence in a real case usually runs through several of these at once, and the evidence that decides them is mundane: the nikahnama, the message or letter by which the pronouncement was made, the witnesses to a spoken pronouncement, and the record of the household’s means. A woman building such a file is handling the papers on which her security depends, and keeping a complete private copy of the marriage record, the pronouncement and the correspondence in an encrypted workspace of the kind journalists use on VaultBook to protect sources and documents is the sort of elementary precaution that costs nothing and is regularly regretted only in its absence. This account is general information about the statutes and not advice on any individual case; limitation periods, forum rules and the facts of a particular marriage are unforgiving, and a woman in this situation should take advice on her own circumstances before choosing among the routes described.

Reading the eight sections as a system

It is worth pausing over the Act’s architecture, because its eight sections interlock more carefully than its reputation for haste suggests, and because the interlocking explains several features that puzzle first-time readers. The Act contains no procedural code of its own beyond Section 7, no appellate structure, no rule-making power of consequence and no administrative machinery: no board, no fund, no officer. Trials are conducted by the ordinary criminal courts under the ordinary criminal procedure, appeals and revisions travel the ordinary channels, and the sentence is an ordinary sentence. The statute is a shell of five operative ideas fitted onto the general law: a definition, a declaration, a punishment, a procedural frame and two entitlements. The definition does the work of scope. The declaration does the work of status. The punishment does the work of deterrence. The procedural frame allocates control of the punishment among the wife, her family and the Magistrate. And the entitlements convert the status the declaration preserves into money and children. Almost every question a reader can ask about the Act is answered by identifying which of the five ideas governs it, and almost every criticism of the Act is a criticism of the joins: between the declaration and the punishment, which sit in adjacent sections asserting opposite things about the act’s significance; between the punishment and the allowance, which assume an earner alternately imprisoned and paying; and between the wife’s control of the trigger and the state’s ownership of the prosecution once triggered, since a criminal case, once on foot, belongs in form to the state even though the Act returns its effective control to her at the compounding stage.

A technical point about Section 8 completes the architecture. The repeal is of the Second Ordinance of 2019 alone, the instrument in force when the Act passed, and the savings clause carries forward everything done under that ordinance. The two earlier ordinances needed no repeal by then: each had spent its constitutional life being replaced by the next, and the chain they formed is preserved not by the repeal clause but by the commencement clause, which fixes the whole period from 19 September 2018 as the Act’s own tenure. A prosecution for a pronouncement made in, say, the winter of 2018 therefore charges the offence as an offence under the Act, although the words were spoken when only an ordinance ran, and the accused can take no point on the substitution, because the deeming provisions at both ends of the chain were designed to meet exactly that objection. Retrospectivity of this welded kind, where a statute adopts a period during which identical law already operated, is constitutionally far safer than retrospectivity that creates an offence for a period when none existed, and the distinction explains why the Act’s backward reach, so prominent in the political argument, has never been the centre of the legal challenge to it.

The heading of Chapter III is part of the architecture and repays notice. The long title promises protection of the rights of married Muslim women, and the chapter that bears the promise contains the allowance, the custody entitlement and, significantly, the offence’s procedural code. The safeguards are placed by the draftsman among the woman’s rights, not among the accused’s protections, and that placement is the statute’s own answer to the question whose safeguards they are. A bail hearing at which the wife must be heard is, in the Code’s usual arrangement, a protection of the accused’s liberty interest balanced against the prosecution. In this Act it is drafted and placed as one of the wife’s entitlements, a right to be present and heard before the man who pronounced talaq on her is released. Whether one approves that inversion or distrusts it, it is the clearest single evidence that the Act was designed from the wife’s procedural position outward, and it is the feature a reader should carry into the arguments on both sides: the statute’s defenders are describing a real design, and its critics are criticising that design itself, not a caricature of it.

The Act in the federal frame

One structural feature of the Act is easy to miss and explains part of its administration. Marriage and divorce sit in Entry 5 of the Concurrent List in the Seventh Schedule, among the subjects on which both Parliament and the state legislatures may legislate, and criminal law and criminal procedure sit beside them in the same List. The Act is therefore a central statute on a concurrent subject, competent for Parliament to enact for the whole country, and it displaces nothing that a state had enacted because no state had legislated in this corner at all. Its enforcement, however, runs through institutions the Union does not command. Police and public order are state subjects, so the information the wife gives is given to a state police station, the investigation is a state investigation, and the prosecution is conducted in the state’s courts by the state’s prosecutors under a central offence. The allowance and custody orders are made by the ordinary magistracy. The consequence is that a national prohibition is administered at the pace and with the priorities of thirty-six different police and prosecution systems, and the Act’s lived reality, the speed of the first information report, the conduct of the bail hearing, the seriousness with which compounding terms are framed, varies by district in the way every concurrent-subject penal statute’s reality varies. The Union’s role after enactment is confined to the law’s text and to answering for it in Parliament. A reader assessing the Act’s uniformity should therefore expect, as with the dowry and cruelty provisions whose administration has the same federal structure, a national rule and a local experience, and should attribute unevenness in the second neither wholly to the statute’s design nor wholly to its administrators.

The political afterlife of the Act

The Act was enacted weeks after a general election in which it had already featured as an argument, and its political life since passage has tracked the debate that produced it. The government presented the statute to Muslim women as a constituency in its own right, and the Prime Minister and the Law Minister both framed its passage as the state discharging a debt that earlier Parliaments, and specifically the Parliament of 1986, had declined to pay. That framing made the Act a standing exhibit in a larger contest about which political formation speaks for Muslim women’s interests, a contest in which the women’s organisations that had campaigned against the practice occupied an uncomfortable position: the Bharatiya Muslim Mahila Andolan had wanted the practice ended and a codified family law enacted, and it supported the end while continuing to press for the wider codification the Act does not attempt. The community organisations that had opposed the penal clause continued to oppose the statute in court and in public, and the opposition parties that had resisted it in Parliament largely let the question rest after 2019, neither repealing it in the states they governed, which in any event a state legislature could not do for a central Act, nor making its repeal a platform. The durable political fact is that the Act settled into the statute book with its controversy intact but its repeal off every agenda, a pattern Indian legislative history shows repeatedly with contested social statutes: enactment ends the parliamentary argument long before it ends the argument about the enactment. For the enforcement picture the same caution applies as elsewhere in this series. Prosecutions under the Act depend entirely on complaints by wives and their relatives, so their volume measures willingness to invoke the Act as much as it measures the practice’s persistence, and the figures given in debate and in answer to parliamentary questions at different times move with definitions and periods. The structural facts are the stable ones: the complaint gate keeps the caseload in the hands of the women concerned, the bail stage is where most contests are fought, and compounding is the exit through which a substantial share of the pressure the Act creates is designed to be released.

How the case and the Act are studied

For the law student, who meets this material in constitutional law and family law papers alike, the value of the Shayara Bano and statute sequence is that it compresses three methods of legal change into a single episode short enough to hold in the mind. The first method is constitutional adjudication on a statutory hook: Justice Nariman’s opinion, which found the practice inside a statute, tested it for manifest arbitrariness, and struck the statute down to the extent of the practice, an exercise in deciding much by assuming little. The second method is interpretation from within a tradition: Justice Kurian Joseph’s opinion, which treated the faith’s own text as the authoritative measure of the practice and found the practice wanting by it. The third method is legislative supremacy over personal law, asserted by the minority as the only legitimate route and then exercised by Parliament in a form, criminalisation, that no opinion in the case had proposed. A candidate who can state which judge used which method, what each method decided and what each deliberately left undecided, and how the 2019 Act then selected a fourth instrument altogether, the penal sanction, has the episode. The traps in the material are the ones this article has marked as they arose: reporting the 3:2 outcome as a single majority reasoning; describing the Act as the instrument that banned the practice, when the ban was the judgment’s and the Act’s contribution is the offence and the entitlements; omitting the safeguards from the description of the offence; and treating the retrospectivity as though it reached back to the judgment, when it reaches back only to the first ordinance and to a period in which the identical offence already ran. The sequence also supplies the standard illustration, in the ordinance chapter of any polity course, of re-promulgation sustaining a measure against one House’s resistance, and the standard illustration, in the criminal law chapter, of the question what a sanction attaches to when the act it punishes has been declared void.

Assessment: what the Act settled and what it did not

What the Act settled is the state’s position, stated as clearly as a statute can state it. The pronouncement is void. It is also punishable. The wife holds the trigger, the settlement power and the first word on bail, and she has summary claims to subsistence and custody in the same forum. No later dilution has touched those propositions, and the design has survived its first political test, which was the election fought partly over it. What the Act did not settle is everything that made it controversial. Petitions challenging the statute, brought among others by the Jamiat Ulama-i-Hind and by the Samastha Kerala Jamiyyathul Ulama, carried the parliamentary objections into constitutional form: that the criminalisation of a pronouncement already deprived of legal effect is disproportionate, that it selects the marital wrong of one community for a punishment no equivalent wrong attracts, and that the penal clause sits uneasily with the religious freedom the minority in Shayara Bano would have protected. The Supreme Court issued notice on those petitions, and the questions they frame are the questions this article has set out, awaiting a bench’s answer in the durable form such questions take in this field: not whether the practice should exist, on which the argument is over, but whether the criminal law was a constitutionally permissible instrument for ending it.

The challenge, when a bench hears it, will be argued in the grammar the Court has built for such questions since the privacy judgment in Justice K. S. Puttaswamy v. Union of India: proportionality. The state will say the aim, ending a practice that destroys marriages in a sentence and leaves women destitute, is legitimate and indeed commanded by the equality the same Constitution promises; that the sanction is suitable because deterrence is the mechanism by which pronouncements were to be stopped where declarations had failed; that it is necessary, or at least within the legislature’s margin, because the civil alternatives had been on the field since 2017 and had not sufficed on the government’s record; and that the balancing favours the Act because its safeguards confine the sanction’s incidence to cases the wife herself brings. The petitioners will answer at the necessity and balancing stages: that a measure which imprisons a man for a legally ineffective act, and whose deterrent object could be pursued by the civil and procedural means the Act itself also enacts, fails the requirement that a rights-limiting measure be the least restrictive adequate one, and that the burden it imposes falls asymmetrically on one community’s marital wrongs. On the religious freedom limb, the Act has a structural defence the 2017 minority would have recognised: the statute does not regulate a religious practice’s civil validity on the Court’s contested authority but exercises Parliament’s undoubted power to create offences, and it was the minority itself that assigned personal law reform to Parliament. None of that forecloses the equality limb, which is where the petitions are at their strongest and where the analogy with desertion and abandonment in the other personal laws will be pressed hardest. This article predicts nothing about the outcome. It records that the questions are framed, that they are the same questions the two Houses argued in political form, and that a constitutional answer to them, whenever it comes, will be the final document in the sequence this article has traced.

The honest verdict on the design is two-sided, as the subject requires. As a piece of drafting aimed at a defined mischief, the Act is unusually disciplined: it names one act, confines its trigger to the person wronged, arms her with the settlement and a hearing before bail, and attaches the two civil entitlements she most needs to the same proceeding. Its safeguard triad is a genuine legislative response to the misuse objection, produced by the parliamentary process working as it should, between the Bill as first introduced and the Bill as finally passed. As a piece of criminal law theory, it remains exposed to the objection its critics never had answered in the chamber: that it punishes an act the law itself declares to be without effect, and that the spouse it imprisons is the same spouse its own allowance section expects to pay. Whether deterrence has in fact reduced pronouncements is an empirical question on which the responsible course is the one this series follows with all enforcement figures, to describe the pattern as the evidence matures rather than to borrow a number from the debate. The statute’s contribution to the series’ running question, what criminalisation adds when the legislature reaches for it, is the purest case the series contains: a civil consequence already exhausted by a court, a sanction added by a legislature, and a design whose entire content is the sanction and the bargaining power the sanction creates. The pillar articles around this one carry the neighbouring ground: the 1986 Act and its reading in Danial Latifi, the Shah Bano judgment and the politics of its reversal, and the doctrinal map of the personal law system in which this eight-section statute sits as the most contested single enactment of its generation.

Frequently Asked Questions

Q: What did the Supreme Court decide in Shayara Bano?

In Shayara Bano v. Union of India, (2017) 9 SCC 1, decided on 22 August 2017, a five-judge Constitution Bench set aside the practice of talaq-e-biddat by three votes to two. Justices Nariman and Lalit held that the Muslim Personal Law (Shariat) Application Act, 1937, which recognises and enforces the practice, is law in force under Article 13, and that instant triple talaq is manifestly arbitrary and violates the equality guarantee in Article 14. Justice Kurian Joseph concurred on a different ground, holding that the practice is not sanctioned by the Quran and is therefore bad in theology and in law. The result was that the practice ceased to have legal effect from the day of the judgment.

Q: Which pronouncement does the Triple Talaq Act criminalise?

The Act criminalises a pronouncement of talaq by a Muslim husband upon his wife where the pronouncement is talaq-e-biddat or any other similar form of talaq having the effect of instantaneous and irrevocable divorce, as Section 2(c) defines it. Section 3 covers the pronouncement however made: by words spoken or written, in electronic form, or in any other manner whatsoever. The test is the pronouncement’s effect, not its exact words or the number of times talaq is uttered. A pronouncement that is spaced over time and remains revocable, as in talaq-e-ahsan or talaq-e-hasan, does not produce an instantaneous and irrevocable divorce and so falls outside the definition and outside the offence.

Q: Why is a void talaq pronouncement also a crime?

Because Parliament treated the pronouncement and its legal effect as two different things. The Supreme Court’s judgment of 2017 had already removed the effect: the pronouncement dissolved nothing and the marriage continued. The government’s case for the statute was that pronouncements nevertheless continued to be made and acted upon socially, so a declaration without a sanction had failed to protect women in practice. The Act therefore declares the pronouncement void and illegal in Section 3 and punishes the making of it in Section 4, attaching the criminal law to the act of pronouncing rather than to any consequence. Critics answer that punishing an act the law itself treats as ineffective is incoherent, and that disagreement, deterrence against nullity, is the centre of the debate about the Act.

Q: Who can report an offence under the Triple Talaq Act?

Only two classes of person can set the criminal process in motion. Under Section 7(a), the offence is cognizable only if information relating to its commission is given to the officer in charge of a police station by the married Muslim woman upon whom talaq is pronounced, or by any person related to her by blood or marriage. A neighbour, an employer, a political organisation or any other stranger who learns of a pronouncement cannot found a cognizable case under this Act. The confinement was a deliberate safeguard, introduced by official amendments in 2018 after the first Bill had left the information gate open, and it places the decision to prosecute in the hands of the wife and the family members most likely to be sheltering her.

Q: What subsistence allowance and custody does the Triple Talaq Act provide?

Section 5 entitles a married Muslim woman upon whom talaq is pronounced to receive from her husband such amount of subsistence allowance, for herself and her dependent children, as the Magistrate determines. The Act fixes no amount and no formula, and the entitlement is expressly without prejudice to her claims under any other law, so it can be pursued alongside the general maintenance remedy. Section 6 entitles her, notwithstanding anything in any other law, to custody of her minor children in the event of the pronouncement, in the manner the Magistrate determines. Both entitlements proceed on the Act’s premise that the marriage subsists: they are protections for a wife whose husband attempted to cast her off, not the financial and custodial incidents of a completed divorce.

Q: Why was the Triple Talaq Act preceded by ordinances?

Because the government could pass its Bill in the Lok Sabha but not in the Rajya Sabha, where the opposition pressed for reference to a select committee. The first Bill was passed by the Lok Sabha in December 2017 and stalled in the Council of States. With Parliament not in session and the government maintaining that pronouncements were continuing, the President promulgated the first ordinance on 19 September 2018, a second on 12 January 2019 and a third on 21 February 2019, each carrying the same provisions into force while the replacement Bills failed to clear the Rajya Sabha. When the sixteenth Lok Sabha was dissolved the pending Bills lapsed, and a fresh Bill in the new Parliament became the Act, which repealed the last ordinance and adopted its commencement date.

Q: What are the main criticisms of criminalising triple talaq?

The criticisms fall into three groups. The doctrinal criticism is that imprisoning a man for up to three years for a pronouncement the same statute declares void punishes an act without legal effect, and that desertion or abandonment in other personal laws attracts no equivalent freestanding offence, so the provision selects one community’s marital wrong for criminal punishment. The practical criticism is that a husband in custody cannot earn and therefore cannot pay the subsistence allowance the Act promises, and that the wife’s real interests, maintenance, residence and the children, are served by civil remedies rather than by prosecution. The procedural criticism is that the measure was sustained through three successive ordinances and never received select committee scrutiny in either Parliament, a route members across the opposition described as a fraud on the Constitution.

Q: Does the Act apply to other forms of talaq?

No. The Act’s definition in Section 2(c) is confined to talaq-e-biddat and any other similar form of talaq having the effect of instantaneous and irrevocable divorce. Talaq-e-ahsan, a single pronouncement followed by the iddat period and revocable within it, and talaq-e-hasan, pronouncements spaced across three successive periods and irrevocable only on the third, do not take effect instantaneously and are outside the Act. The Act also does not touch the other routes out of a Muslim marriage: khula and mubaraat, which end a marriage at the wife’s instance or by mutual agreement, talaq delegated to the wife, and the wife’s suit for a decree of dissolution under the Dissolution of Muslim Marriages Act, 1939 all continue to operate as they did before the Act.

Q: When did the Triple Talaq Act come into force?

The Act received the President’s assent on 31 July 2019 and was published in the Gazette that day, but its commencement is earlier than its enactment. Section 1(3) provides that it shall be deemed to have come into force on 19 September 2018, the date on which the first of the three ordinances carrying the same provisions was promulgated. The deeming welds the ordinance period and the Act period into a single continuous tenure, and Section 8 reinforces it by providing that anything done under the repealed Second Ordinance of 2019 is deemed to have been done under the Act. The Act is accordingly cited as Act 20 of 2019, with a commencement date in September 2018.

Q: What is the punishment for pronouncing triple talaq under the Act?

Section 4 provides that a Muslim husband who pronounces talaq upon his wife in the form the Act covers shall be punished with imprisonment for a term which may extend to three years, and shall also be liable to fine. The Act prescribes no minimum term of imprisonment and no ceiling for the fine, leaving both to the sentencing Magistrate on the facts of the case. There is no separate or enhanced provision for a second pronouncement. The three-year maximum places the offence alongside the Penal Code’s cruelty provision for husbands and their relatives, which carries the same maximum, and below bigamy, which carries seven years, comparisons that were pressed on both sides of the parliamentary debate about the term’s calibration.

Q: Is the offence under the Triple Talaq Act bailable?

Bail under the Act is governed by Section 7(c) rather than by the ordinary classification in the Code of Criminal Procedure. It provides that no person accused of an offence under the Act shall be released on bail unless the Magistrate, on an application filed by the accused and after hearing the married Muslim woman upon whom talaq is pronounced, is satisfied that there are reasonable grounds for granting bail to that person. Two features distinguish the provision: the wife must be heard before the accused is released, which gives her notice and a voice at the first contest in the case, and the Magistrate must record satisfaction that reasonable grounds exist, rather than bail following as a matter of course for an offence of this level.

Q: Can a case under the Triple Talaq Act be settled or withdrawn?

Yes, through compounding, which the Act builds into Section 7(b). The offence is compoundable at the instance of the married Muslim woman upon whom talaq is pronounced, with the permission of the Magistrate and on such terms and conditions as the Magistrate may determine. It is not a simple private withdrawal: the wife must initiate it, the court must approve it, and the terms are the court’s, which allows terms addressing the subsistence allowance, the custody of the children and the parties’ future conduct to be built into the order that closes the case. The design answers the criticism that a criminal complaint in a subsisting marriage leaves the wife trapped in a prosecution she started, by giving her a supervised exit that can carry the settlement’s substance.

Q: Does a triple talaq sent by phone, message or email count under the Act?

Yes. Section 3 declares a pronouncement void and illegal whether it is made by words spoken or written, in electronic form, or in any other manner whatsoever, and Section 2(a) gives electronic form the meaning it bears in the Information Technology Act, 2000. A pronouncement delivered by a telephone call, a text message, an email or a messaging application is therefore within the Act in exactly the same way as one spoken in the wife’s presence, and many of the cases that built the campaign and the litigation involved precisely those modes, including a divorce communicated by post and one by telephone from abroad. In court the electronic record must be proved in accordance with the evidence law’s requirements for electronic records, including the certificate those rules require.

Q: What happens to the marriage when a husband pronounces triple talaq?

Nothing in law. Section 3 declares the pronouncement void and illegal, and the Supreme Court had already set the practice aside in 2017, so the pronouncement does not dissolve the marriage and the parties remain husband and wife with every obligation of that status intact, including the husband’s duty to maintain his wife. The consequences run the other way: a second marriage contracted by the husband in reliance on his pronouncement is void and exposes him to the law of bigamy; the parties can reconcile without any remarriage, since no divorce occurred for the halala rule to attach to; and if the marriage is to end, it must end by a lawful route, such as a spaced form of talaq outside the Act or a decree under the Dissolution of Muslim Marriages Act, 1939.

Q: Which Magistrate hears a case under the Triple Talaq Act?

The Act defines Magistrate in Section 2(b) as a Judicial Magistrate of the first class exercising jurisdiction under the Code of Criminal Procedure, 1973 in the area where the married Muslim woman resides. The forum follows the wife’s residence rather than the place of the pronouncement or the husband’s residence, a deliberate choice for a situation in which the woman has frequently returned to her parental home in another town or district after the pronouncement. The same Magistrate conducts the trial of the offence, hears the wife before any grant of bail under Section 7(c), supervises compounding under Section 7(b), and determines the subsistence allowance under Section 5 and the manner of custody under Section 6, so the whole of the Act’s machinery is concentrated in the one court nearest to her.

Q: Has the constitutionality of the Triple Talaq Act been challenged?

Yes. Petitions under Article 32 were filed in the Supreme Court by organisations including the Jamiat Ulama-i-Hind and the Samastha Kerala Jamiyyathul Ulama, challenging the Act and in particular its criminal provisions, and the Court issued notice on them. The petitions frame in constitutional terms the arguments made in the parliamentary debate: that imprisoning a husband for a pronouncement already deprived of legal effect is disproportionate, that the offence selects the marital wrong of one community for a punishment equivalent wrongs elsewhere do not attract, and that the penal clause sits uneasily with the freedom of religion. The questions those petitions raise await a bench’s determination, and the Act operates in full while they do. The challenge is to the statute’s instrument, not to the 2017 judgment, whose setting aside of the practice no party seeks to reopen.

Q: What should a woman do if her husband pronounces triple talaq on her?

The first thing to understand is that the pronouncement has not divorced her: the marriage subsists, and with it her rights to maintenance and residence. If she chooses to invoke the Act, she or a person related to her by blood or marriage may give information to the officer in charge of the police station, and she may apply to the Magistrate for the area where she resides for the subsistence allowance under Section 5 and custody of her minor children under Section 6. In parallel she may pursue maintenance under the general provision of the criminal procedure law, protection and residence orders under the Protection of Women from Domestic Violence Act, 2005, and recovery of her unpaid mahr, which is a debt owed to her. Preserving the nikahnama and the message, letter or record of the pronouncement is essential evidence in each of these routes.

Q: How does the Triple Talaq Act 2019 differ from the Muslim Women Act 1986?

The two statutes govern opposite situations. The Muslim Women (Protection of Rights on Divorce) Act, 1986, Act 25 of 1986, addresses a woman who has actually been divorced, and prescribes the provision and maintenance owed to her within and after the iddat period, a scheme the Supreme Court in Danial Latifi in 2001 read to require a reasonable and fair provision for her future made within that period. The Act of 2019 addresses a woman who has not been divorced at all, because the pronouncement is void: her marriage subsists, and the Act gives her a criminal complaint against the husband, a subsistence allowance and custody while the marriage continues. The 1986 Act creates no offence and imprisons nobody; the 2019 Act’s distinctive contribution is precisely the offence, together with entitlements designed for a subsisting marriage.

Q: Does the Act apply to a pronouncement made before Parliament passed it?

It applies to pronouncements made on or after 19 September 2018, the date of the first ordinance, because the Act is deemed to have come into force on that date and the identical offence ran continuously under the three ordinances before the Act replaced the last of them. A pronouncement made during the ordinance period is charged under the Act by the combined operation of the commencement clause and the savings in Section 8. It does not apply to a pronouncement made between the Supreme Court’s judgment in August 2017 and the first ordinance in September 2018: during that interval the pronouncement was void, because the judgment had set the practice aside, but it was not yet a criminal offence, and the Act creates no liability for that window.

Q: What did Chief Justice Khehar’s dissent hold in Shayara Bano?

Chief Justice Khehar, writing for himself and Justice Nazeer in the minority, held that talaq-e-biddat is a matter of personal law forming part of the religious faith of the Hanafi school, protected by the freedom of religion in Article 25, and that personal law is not law in force for the purposes of Article 13 and so cannot be tested against the fundamental rights. The dissent therefore declined to strike the practice down, holding its reform to be the legislature’s province. But it did not leave the practice standing untouched: it directed the Union of India to consider appropriate legislation on the subject, and in the interim injuncted Muslim husbands, in exercise of the Court’s power under Article 142, from pronouncing talaq-e-biddat for a period of six months.