LLM Group 1 Constitutional and Administrative Law National Security Public Order and Rule of Law 2025-26 Question Paper with Solutions

Article 22 is two codes in one article, and they point in opposite directions

Article 22 is placed in Part III among the rights to freedom, but it is the only fundamental right in the Constitution that expressly authorises the State to imprison a person who has committed no offence. Clauses (1) and (2) confer rights on a person arrested under the ordinary criminal law. Clauses (4) to (7) confer a much weaker set of protections on a person held in preventive detention, and clause (3) exists solely to switch the first set off for that person. Reading the article as a single scheme of protection is the commonest mistake in this paper. It is a scheme of protection for the ordinary arrestee and a scheme of authorisation, with limits, for the preventive detenu.

The punitive limb, clauses (1) and (2). A person arrested must be informed as soon as may be of the grounds of arrest; he has the right to consult and to be defended by a legal practitioner of his choice; he must be produced before the nearest magistrate within twenty four hours excluding the time of the journey; and he may not be detained beyond that period without the authority of a magistrate. These are the protections that make an arrest reviewable by a judicial officer almost immediately.

Three, the earliest opportunity of making a representation, clause (5), second half. The detenu must be afforded the earliest opportunity of making a representation against the order. This is the safeguard that does the most work in practice, because almost every successful habeas corpus petition in this field is decided on it, as shown below.

Four, the withholding of facts against the public interest, clause (6). The authority need not disclose facts which it considers to be against the public interest to disclose. This is a limitation on the third safeguard and not a safeguard at all, and a good answer says so.

Clause (7) then permits Parliament to prescribe the circumstances in which a person may be detained for longer than three months without an Advisory Board opinion at all, the maximum period of detention, and the Board's procedure.

What the courts have added to the text

The procedural safeguards are mandatory, and breach means release. In Ram Krishan Bhardwaj v. State of Delhi, AIR 1953 SC 318, one of the grounds supplied to the detenu was vague. The State argued that the other grounds were specific and sufficient. The Court rejected that: where a detention rests on several grounds and one is bad, the detention is bad as a whole, because the Court cannot know how much weight the detaining authority gave to the defective ground and the detenu's right to make an effective representation has been impaired. This rule of the bad ground vitiating the whole order has been the most productive single principle in this branch of law for seventy years, and it is now given statutory recognition in the severability provisions of COFEPOSA and the National Security Act.

The order must be the detaining authority's own satisfaction, and it must be alive. Where the detaining authority acts on the dictation of a superior, or reproduces a police proposal without applying its mind, the order fails for want of subjective satisfaction. Where there is a long unexplained gap between the last prejudicial act and the order, the "live and proximate link" between the two is broken and the order is quashed as punitive in substance. Both doctrines are judicial constructions, not text.

Public order is not law and order. The line that keeps preventive detention within bounds was drawn in Ram Manohar Lohia v. State of Bihar, AIR 1966 SC 740, where Hidayatullah J. described three concentric circles: law and order is the widest, public order is narrower and inside it, and security of the State is the narrowest and innermost. Every public order problem is a law and order problem, but not every law and order problem reaches public order. A detention order that recites facts amounting only to an ordinary crime is therefore made under the wrong head, and the courts have quashed detentions on this ground repeatedly, most recently in cases where a detenu had merely been involved in individual offences however serious.

The Court also refused to strike down the vagueness of expressions such as "security of the State" and "public order", holding them to be terms of settled judicial content. And it held that section 3 of the 44th Amendment, being unnotified, could not be enforced by mandamus, since bringing a law into force is a matter for the executive. That last holding is why the two month safeguard is still dead.

The limits of the Advisory Board as a safeguard are structural and should be stated. Its proceedings are not open. The detenu has no counsel unless the State takes one. There is no cross examination. The Board's report is confidential except for its opinion. And the Board is advisory in name: if it reports sufficient cause, the detention continues; if it reports otherwise, the Government must release, so the safeguard operates in only one direction. It is best described as an internal review conducted by judicially qualified persons, not as an adjudication.

An honest assessment

The pattern is clear. The safeguards that work are the two that are procedural and verifiable on the record: was the ground supplied, and was the representation considered promptly. The safeguards that are substantive, which is to say those that would test whether the person should be detained at all, do not work, because subjective satisfaction is not reviewable on merits and the Advisory Board does not test evidence. This is why the standard criticism of Indian preventive detention law is that it is strong on form and empty on substance, and why the National Human Rights Commission and successive Law Commission reports have pressed for the notification of section 3 of the 44th Amendment.

Conclusion. Article 22 is not a guarantee against preventive detention; it is the constitutional permission for it, hedged with procedure. Clauses (1) and (2) are switched off by clause (3) for exactly the class of person who most needs them, and what remains is a three month leash policed by a closed Advisory Board, a right to be told the grounds, and a right to be heard on paper.

For full marks, cover: the three proclamations with their dates, stated grounds and duration; the constitutional machinery each one switched on; why the 1975 proclamation is different in kind from the other two; ADM Jabalpur as the low point and its formal burial in Puttaswamy; the 44th Amendment as the constitutional answer; and a judgment on what the episode teaches about emergency powers.

The three proclamations

There have been exactly three proclamations of national emergency under Article 352 in the history of the Republic, and the dates matter because papers in this subject frequently misstate them.

A note on the older papers in this folder. The 2015, 2018 and 2019 papers of this subject ask about emergencies "declared in 1962, 1965 and 1970". No emergency was proclaimed in either of those years. A candidate who meets that wording should answer on the three real proclamations and say plainly that 1965 fell within the subsisting 1962 emergency and that there was no 1970 proclamation. That is not a quibble: the whole constitutional interest of the period lies in the fact that the 1962 emergency was allowed to run for over five years after the fighting stopped, and in the fact that 1975 was proclaimed under a different head altogether.

The leading decision of that period, Makhan Singh v. State of Punjab, AIR 1964 SC 381, is an important corrective to the picture of judicial helplessness. A Constitution Bench held that a Presidential order under Article 359 bars a petition alleging breach of the specified articles, but does not bar a challenge on any ground outside them: the detenu may still say that the detaining law itself is beyond legislative competence, or that the order is mala fide, or that it does not comply with the statute under which it purports to be made, or that the officer had no jurisdiction. The reasoning is that Article 359 suspends the remedy for specified rights, not the rule of law generally. That distinction was available in 1975 and was not used.

The 1971 emergency ran for over five years, well beyond the fourteen day war it was declared for, and the second and third proclamations overlapped, which is what made the 1975 declaration constitutionally possible on the same architecture.

Institutions. Press censorship was imposed under the Defence and Internal Security of India Rules. Around one hundred thousand people were detained. The Thirty ninth Amendment placed the election of the Prime Minister beyond judicial scrutiny; the Forty first raised immunities; and the Forty second Amendment, 1976, the most far reaching in the Constitution's history, extended the life of the legislatures, curtailed judicial review, added Articles 323A and 323B on tribunals, and inserted Article 368(4) and (5) purporting to place constitutional amendments wholly beyond challenge.

The rule of law. The point at which the ordinary legal order gave way is Additional District Magistrate, Jabalpur v. Shivkant Shukla, (1976) 2 SCC 521. High Courts across the country had entertained habeas corpus petitions from MISA detenus, using the Makhan Singh route: the detention did not comply with the statute, or was mala fide.

The constitutional answer: the 44th Amendment, 1978

The Constitution (Forty fourth Amendment) Act, 1978 rewrote the emergency provisions so that 1975 could not recur on the same architecture.

One, "internal disturbance" in Article 352 was replaced by armed rebellion, so that political disorder alone can no longer found a proclamation. Two, the President may proclaim only on the written advice of the Cabinet, not the Prime Minister alone. Three, a proclamation must be approved by both Houses within one month, by a majority of total membership and two thirds of those present and voting, and lapses after six months unless re approved.

Four, one tenth of the members of the Lok Sabha may requisition a sitting to move a resolution of disapproval, which needs only a simple majority. Five, and most important for this subject, Article 359 was amended so that the right to move a court for the enforcement of Articles 20 and 21 can never be suspended, and Article 358 was confined to emergencies declared on the ground of war or external aggression and to laws actually related to the emergency. Six, the right to property was removed from Part III altogether, ending its use as an argument for emergency economic powers.

Consequences for public order and the rule of law: an assessment

The 1962 and 1971 emergencies show the duration problem: a proclamation justified at its making was allowed to continue for years after the justification lapsed, and during that time the extraordinary became routine. The 1975 emergency shows the capture problem: emergency powers designed for the survival of the State were used against political opposition, and the machinery for suspending rights, being drafted in general terms, made no distinction between the two purposes. In both cases the rule of law failed not because the text was ambiguous but because every institution that might have checked the executive, Parliament, the Court and the press, was disabled at the same time by the same instrument.

The counter argument deserves a hearing. Emergency powers exist in every constitution because a State facing invasion or armed rebellion cannot fight it with the ordinary law of arrest and trial, and the framers, having watched partition, drafted deliberately wide powers. Article 352 was used three times in seventy six years and has not been used since 1975, which suggests the political cost is now prohibitive. But the honest reply is that the restraint since 1977 is a political fact, not a legal one, and it is the 44th Amendment and not political virtue that makes a repetition constitutionally difficult.

For full marks, cover: what due process means, procedural and substantive; why the framers deliberately rejected it and wrote "procedure established by law"; how Maneka Gandhi brought it back through the back door; the specific features of anti terror statutes that displace it, taken one by one; TADA, POTA and UAPA with what the record shows; the bail jurisprudence from Watali to Najeeb; and a conclusion that takes a position.

What due process means

"Due process of law" is the guarantee in the Fifth and Fourteenth Amendments to the United States Constitution that no person shall be deprived of life, liberty or property without due process of law. It has two dimensions.

A.K. Gopalan v. State of Madras, AIR 1950 SC 27, gave that choice its logical effect. Gopalan was detained under the Preventive Detention Act, 1950 and argued that the Act was unreasonable. The Court held, by a majority, that "procedure established by law" means procedure enacted by a competent legislature, not fair procedure; that the Court cannot ask whether the procedure is reasonable; and that Articles 19, 21 and 22 are mutually exclusive compartments, so a law of preventive detention valid under Article 22 need not satisfy Article 19 or any test of reasonableness. Due process, in short, was excluded twice over.

Maneka Gandhi v. Union of India, (1978) 1 SCC 248, reversed that entirely. The petitioner's passport was impounded without reasons under the Passports Act. The Court held that Articles 14, 19 and 21 are not mutually exclusive but form a single golden triangle; that a law depriving a person of personal liberty must prescribe a procedure that is right, just and fair, and not arbitrary, fanciful or oppressive; and that a procedure which is arbitrary is no procedure at all, since Article 14 strikes at arbitrariness and Article 14 now informs Article 21.

POTA, 2002. The Prevention of Terrorism Act repeated the confession provision with more safeguards and was upheld in People's Union for Civil Liberties v. Union of India, (2004) 9 SCC 580. The Court's reasoning is instructive for this question: it held that the mere possibility of abuse is no ground to strike down a statute, and that the remedy for abuse is action against the abuser, not invalidation. POTA was repealed in 2004 after review committees found large scale misuse, and the confession provision was not carried into the UAPA when its substance was transferred there.

UAPA, 1967, as amended in 2004, 2008, 2012 and 2019. This is now the principal statute. It defines a terrorist act very widely, permits detention for up to one hundred and eighty days before a charge sheet, and by section 43D(5) provides that an accused shall not be released on bail if the court, on a perusal of the case diary or the report, is of the opinion that there are reasonable grounds for believing that the accusation is prima facie true. The 2019 amendment allows the Central Government to designate an individual as a terrorist, with only a departmental review committee by way of remedy, which is the sharpest current due process objection: a designation with reputational and practical consequences made without a hearing and without adjudication.

That reasoning has been applied repeatedly since, including in Vernon v. State of Maharashtra, (2023) SCC OnLine SC 885, where the Court also clarified that Watali does not forbid the court from looking at the probative value of the material at a surface level, and in later decisions holding expressly that Article 21 overrides the section 43D(5) embargo where incarceration has been prolonged. The Supreme Court has said in terms that bail is the rule even in these cases and that the special statute cannot become an instrument of indefinite detention.

And the successor to sedition. Section 124A of the Penal Code was kept in abeyance by the Supreme Court's order in S.G. Vombatkere v. Union of India of 11 May 2022. It has been replaced, from 1 July 2024, by section 152 of the Bharatiya Nyaya Sanhita, 2023, "acts endangering sovereignty, unity and integrity of India", which drops the word sedition but covers exciting secession, armed rebellion and subversive activities, and carries life imprisonment.

Three, the inversion of the presumption. Statutory presumptions of guilt and admissible police confessions together reverse the two rules the common law regards as constitutive of a fair criminal trial.

Four, the executive as adjudicator. Individual designation under the 2019 amendment, and executive banning of organisations, place a decision with penal consequences outside any court at the point at which it is made.

Five, the mere possibility of abuse doctrine. The reasoning in PUCL that abuse is no ground for invalidation is respectable in the abstract but sits badly with the evidence, because where a power is used tens of thousands of times and convicts one per cent, the abuse is not a possibility but a pattern.

For full marks, cover: the text of Article 358 and how it operates automatically; the crucial difference between suspending a right and suspending a remedy, which is what separates 358 from 359; the two limitations the 44th Amendment added; the case law on revival and on acts done during the emergency; the criticism; and a conclusion on whether the article as it now stands is defensible.

The text, and the mechanism

Article 358 provides that while a Proclamation of Emergency declaring that the security of India or any part of it is threatened by war or external aggression is in operation, nothing in Article 19 shall restrict the power of the State to make any law or to take any executive action which the State would but for Article 19 be competent to make or take. It further provides that any law so made ceases to have effect, to the extent of the incompetency, as soon as the Proclamation ceases, except as respects things done or omitted to be done before the law ceased to have effect.

Why the article exists

The justification offered by the framers, and repeated by the Court, is that Article 19 is a set of freedoms subject to reasonable restrictions in the interests of, among other things, the security of the State and public order; and that in wartime the ordinary judicial calculus of reasonableness cannot be applied case by case to measures such as censorship, requisitioning, restrictions on movement in border areas or the control of assembly. Article 358 removes the litigation risk from wartime administration by pre empting the question rather than by asking a court to answer it under pressure.

The 44th Amendment: two real limitations

Before 1978 Article 358 applied to an emergency on any ground, including internal disturbance, and to any law made during it, whether or not the law had anything to do with the emergency. Both features were used in 1975 and 1976. The Constitution (Forty fourth Amendment) Act, 1978 imposed two limitations, and a candidate who does not state both has answered only half the question.

The case law

On the automatic character and its consequences, Makhan Singh v. State of Punjab, AIR 1964 SC 381. During the 1962 emergency, detenus under the Defence of India Rules challenged their detention. The Constitution Bench explained the architecture with precision: Article 358 suspends the operation of Article 19 while the emergency lasts, so a law that would otherwise be bad for breach of Article 19 cannot be attacked on that ground; Article 359 does something different and narrower, barring the enforcement of the specified rights; and neither provision prevents a detenu from arguing that the impugned order is outside the statute, or mala fide, or made by an authority without jurisdiction, because those grounds do not depend on any fundamental right.

The importance of the case is that it establishes that an emergency suspends particular rights, not the rule of law, and it supplied the route that the High Courts used in 1975 and that ADM Jabalpur then closed.

On the relationship with Article 21. After the 44th Amendment, Article 359 cannot suspend the enforcement of Articles 20 and 21 at all. Since Article 358 touches only Article 19, the position now is that during any emergency, on any ground, a person deprived of life or personal liberty may go to court under Article 21, and ADM Jabalpur could not be decided the same way today. Puttaswamy (2017) confirmed this by overruling that decision in terms.

The criticism

One, the automatic operation remains objectionable in principle. Article 358 requires no application of mind to which freedom needs to be curtailed and why. All six freedoms in Article 19 go together, including freedoms with no conceivable connection to a war, such as the freedom to form associations for trade union purposes or to practise a profession. A proportionate provision would suspend the freedoms shown to be an obstacle, not the article.

Four, the case for the article is nonetheless real. The alternative is that every wartime measure is litigated as it is taken, and that courts decide questions of military necessity on affidavits during hostilities. Comparative experience supports caution here: the United Kingdom's wartime jurisprudence in Liversidge v. Anderson, [1942] AC 206, showed a court declining to look behind a minister's satisfaction, with Lord Atkin's celebrated dissent that amid the clash of arms the laws are not silent; and the United States interned Japanese Americans with judicial approval in Korematsu v. United States, 323 US 214 (1944), a decision finally repudiated in Trump v. Hawaii, 585 US 667 (2018). Courts do not perform well as reviewers of wartime necessity, and Article 358 can be defended as an honest recognition of that, provided it is confined to war, which since 1978 it is.

Conclusion. Article 358 suspends Article 19 itself, automatically, for the duration of a war or external aggression emergency, and that is a far stronger measure than Article 359, which suspends only access to the courts and leaves the right intact. Its worst features were the ones the 44th Amendment removed: it no longer applies to an internal emergency, so the ground on which 1975 was proclaimed can never again suspend the freedoms of speech, assembly and association, and it now protects only laws that recite their connection with the emergency.

For full marks, cover: a definition and the three senses in which the phrase is used; the English doctrine with Dicey and the cases; the Indian position, Article 34 and the indemnity power; the sharp contrast with a proclamation of emergency, which is the comparison the examiner wants; Jallianwala Bagh and the Punjab Disturbances as the Indian instance; the modern substitutes, Chapter XI BNSS and AFSPA; and a conclusion.

What martial law is

Martial law is not a body of law at all. It is the suspension of ordinary law and its temporary replacement by the will of the military commander in an area where the civil authority has broken down, together with the use of military tribunals to try civilians. The phrase is used in three distinct senses and confusing them is the standard error.

One, there is no distinct martial law. What English law recognises is the common law right and duty of every subject, civilian and soldier alike, to use such force as is reasonably necessary to suppress a riot or repel an invasion. A soldier who fires on a mob is not exercising a special power; he is exercising the ordinary right of any citizen to prevent a felony, and he is answerable in the ordinary courts for anything beyond what was necessary. The Bill of Rights, 1689 reinforces this by making the keeping of a standing army in time of peace without the consent of Parliament unlawful.

Two, the courts retain jurisdiction, but they will not exercise it while the fighting continues. In R. v. Nelson and Brand (1867), arising out of the Jamaica rebellion, Cockburn C.J. directed that officers were answerable in the ordinary courts for excesses. In Ex parte Marais, [1902] AC 109, the Privy Council held, on an appeal from the Cape Colony during the Boer War, that where war is actually raging the civil courts will not interfere with the acts of the military authority, and famously observed that the fact that the civil courts are open is not decisive. In R. v. Governor of Wandsworth Prison, ex parte Allen and in the Irish cases the same line was taken. The principle that emerges is temporal: while war rages, the courts stand back; when it ends, they resume and the soldier must justify what he did.

Four consequences follow, and they are the heart of the answer.

One, Article 34 presupposes martial law; it does not create it. There is no article empowering the President, a Governor or any commander to proclaim martial law. Where the power comes from is therefore a genuine question, and the accepted answer is that it is a common law power, inherited with the English common law, to meet a situation where the civil authority has ceased to function.

Two, the indemnity must be by Parliament and by law. Not by ordinance in substance, not by executive fiat. This mirrors the English Act of Indemnity and imports the same public accountability: the acts done must be described in a Bill and defended in the House.

Three, Article 34 is an exception to the whole of Part III. An indemnity Act made under it is valid notwithstanding Articles 14, 19, 21 or 22, which is why it needs its own article. Absent Article 34, an Act validating a sentence passed by a military tribunal on a civilian would be struck down.

Four, Article 34 impliedly recognises that a court martial may have tried civilians, since it speaks of validating "any sentence passed, punishment inflicted, forfeiture ordered". That is the strongest textual indication that martial law in the full sense is contemplated.

The essential difference is that an emergency is a constitutional state of exception, declared by the political executive, controlled by Parliament and leaving the courts open; martial law is an extra constitutional fact, recognised by the Constitution only for the purpose of cleaning up afterwards.

Chapter XI of the Bharatiya Nagarik Suraksha Sanhita, 2023 (formerly Chapter X of the Code of Criminal Procedure, 1973) governs the dispersal of unlawful assemblies. An Executive Magistrate may command an assembly to disperse and use civil force; and where that fails he may require the armed forces to disperse it, the officer then acting under the magistrate's direction and using as little force as possible. Prosecution of the officer or magistrate for acts done under these provisions requires the sanction of the Government. This is the "aid to civil power" model: the army acts, but the civil magistrate decides, and the ordinary law continues to apply.

The Armed Forces (Special Powers) Act, 1958 goes further. Once an area is declared "disturbed" under section 3, a commissioned officer may under section 4 fire upon and even cause the death of a person acting in contravention of an order prohibiting assembly of five or more persons or the carrying of weapons, may destroy arms dumps and fortified positions, may arrest without warrant on reasonable suspicion and may enter and search without warrant. Section 6 bars prosecution except with the previous sanction of the Central Government.

Assessment

The Indian and English positions are in substance the same and the differences are of form. Both refuse to recognise martial law as a source of legal authority; both treat the soldier as answerable in the ordinary courts; both provide for a retrospective statutory indemnity; and both accept that while disorder actually rages a court will not second guess the commander. India's difference is that it has written the indemnity power into the Constitution itself in Article 34, which is at once an acknowledgement that the situation may arise and a limitation on it, since the indemnity must come from Parliament by law and from nobody else. The deeper Indian difference is practical rather than doctrinal: because AFSPA and Chapter XI of the BNSS supply extensive military powers under a civil legal framework with statutory sanction requirements, the occasion for martial law proper has never arisen since 1950.

India inherited that position and constitutionalised its final step in Article 34, which empowers Parliament to indemnify persons for acts done to restore order where martial law was in force and to validate sentences passed under it, notwithstanding the rest of Part III. There is no Indian power to declare martial law, and it must be distinguished sharply from a proclamation of emergency, which is constitutional, political, parliamentary controlled and leaves the courts open. Independent India has never used it: the disturbances that would once have called for it are met instead by the aid to civil power provisions of Chapter XI of the BNSS and by AFSPA, both of which keep the armed forces inside a statutory framework and, after Naga People's Movement and Extra Judicial Execution Victim Families, inside the reach of the ordinary criminal courts.

So the Constitution expressly permits protective discrimination. Article 15(3) provides that nothing in Article 15 shall prevent the State from making any special provision for women and children. This is not an exception grudgingly allowed; it is a recognition that treating unequals equally reproduces inequality. It is the constitutional foundation of every protective statute discussed below, and it is what defeats the argument that a law protecting only women violates Article 14. Article 16(4) does similar work for backward classes; Article 15(3) is the women specific enabling provision.

Article 21 has become the principal engine, because after Maneka Gandhi the right to life means the right to live with human dignity, and the Court has read into it the right to be free from sexual harassment, the right to reproductive autonomy, the right to privacy and, since Puttaswamy, decisional autonomy over the body.

The statutory framework, arranged by the harm

Listing statutes earns few marks. Arranging them by the harm each addresses shows you understand the scheme.

Violence in the home. The Protection of Women from Domestic Violence Act, 2005 was the first Indian statute to define domestic violence to include physical, sexual, verbal, emotional and economic abuse, and to give a civil remedy: protection orders, residence orders, monetary relief and custody orders, obtainable from a Magistrate through a Protection Officer, without requiring the woman to leave or to prosecute.

Its most important innovation is the right to reside in the shared household irrespective of ownership, which addresses the practical reason women do not complain. In Satish Chander Ahuja v. Sneha Ahuja, (2021) 1 SCC 414, a three judge Bench overruled S.R. Batra v. Taruna Batra and held that the shared household includes a house belonging to the husband's relatives where the woman has lived in a domestic relationship, restoring the Act's protection to daughters in law living in the in laws' home, which is the commonest Indian situation.

Harassment at work. Vishaka v. State of Rajasthan, (1997) 6 SCC 241, is worked out below; it produced the Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act, 2013, which requires an Internal Committee in every workplace with ten or more employees, Local Committees at the district level for the unorganised sector, and a ninety day inquiry.

Work and pay. The Equal Remuneration Act, 1976, now subsumed in the Code on Wages, 2019, which prohibits discrimination in wages on the ground of sex; the Maternity Benefit Act, 1961, amended in 2017 to extend paid leave from twelve weeks to twenty six weeks and to require a creche in establishments with fifty or more employees; and the Factories Act restrictions on night work, which are increasingly criticised as protective in form and exclusionary in effect.

Reproductive autonomy. The Medical Termination of Pregnancy Act, 1971, amended in 2021 to extend the limit from twenty to twenty four weeks for specified categories, and the Pre Conception and Pre Natal Diagnostic Techniques Act, 1994 against sex selective abortion.

Trafficking and indecency. The Immoral Traffic (Prevention) Act, 1956 and the Indecent Representation of Women (Prohibition) Act, 1986.

The Supreme Court held that sexual harassment at the workplace violates Articles 14, 15, 19(1)(g) and 21; that in the absence of domestic legislation the Court could look to international conventions to construe the fundamental rights, in particular the Convention on the Elimination of All Forms of Discrimination against Women, which India had ratified in 1993; and that a vacuum in legislation does not excuse the State. It laid down binding guidelines with the force of law under Article 141 until Parliament legislated. Parliament took sixteen years, enacting the 2013 Act. The case therefore establishes three things at once: the horizontal reach of Part III into employment, the use of international law as an interpretive source, and the price of legislative inaction.

Personal law and equality: Shayara Bano v. Union of India, (2017) 9 SCC 1. The petitioner was divorced by talaq-e-biddat, instantaneous triple talaq, after fifteen years of marriage. A Bench of five judges of five different faiths held the practice invalid by 3:2. The majority reasoning is not uniform and a good answer says so: Nariman and Lalit JJ.

Autonomy: Joseph Shine v. Union of India, (2018) 2 SCC 189. Section 497 of the Penal Code punished a man who had sexual intercourse with the wife of another man without that man's consent, and expressly provided that the wife was not punishable even as an abettor. A Constitution Bench struck it down unanimously. The reasoning is what matters for this question: the provision treated the wife as her husband's property, since the offence disappeared with the husband's consent; it denied the woman sexual agency by making her incapable of committing the offence; and it was based on a stereotype about women's sexual autonomy that Article 15 forbids.

Chandrachud J. added that a law which deprives a woman of sexual autonomy within marriage is inconsistent with dignity under Article 21. The same reasoning underlies the pending challenges to the marital rape exception, on which a two judge Bench split in Hrishikesh Sahoo in 2022 and which is now before a larger Bench; a current answer should say it is undecided.

The two most recent constitutional developments

Reservation in the legislatures. The Constitution (One Hundred and Sixth Amendment) Act, 2023 reserves one third of the seats in the Lok Sabha, the State Legislative Assemblies and the Delhi Assembly for women. It came into force on 16 April 2026, and it is still not operative, because Article 334A conditions its operation on a delimitation exercise undertaken after the publication of the first census taken after its commencement. A candidate must not write that it is "pending in Parliament": it is enacted and in force and inapplicable, which is a different and more interesting position, and the Constitution (One Hundred and Thirty First Amendment) Bill, 2026, which sought to advance its operation, was defeated in the Lok Sabha, securing 298 votes of 528 against the 352 required.

Sub classification and the creamy layer. State of Punjab v. Davinder Singh, decided 1 August 2024 by seven judges, 6:1, overruled E.V. Chinnaiah and held that States may sub classify within the Scheduled Castes to give preference to the more disadvantaged among them. Four judges would extend the creamy layer principle to Scheduled Castes and Tribes. This bears on this question because women within the most disadvantaged sub groups are the intended beneficiaries of any such classification.

Personal law. Despite Article 44, succession, marriage, divorce, guardianship and maintenance continue to be governed by religion specific rules with unequal effects. Shayara Bano removed one practice; it did not establish a general principle that personal law is subject to Part III, because two of the three majority judges reasoned from the 1937 Act rather than from personal law as such. Uttarakhand's Uniform Civil Code came into force on 27 January 2025, the first in a State, and its live in registration provisions are under challenge on privacy grounds following Puttaswamy.

The deeper criticism. The framework is overwhelmingly protective rather than enabling: it responds to violence after it happens and restricts what women may do in the name of safety, rather than removing the structural barriers to their economic independence. Protection without economic autonomy leaves the woman dependent on the household she needs protection from, which is why the residence right in the 2005 Act and the coparcenary right in Vineeta Sharma are worth more in practice than several of the criminal provisions.

The failure is not of law but of delivery: conviction rates around thirty per cent for rape, an Internal Committee regime that barely touches the ninety per cent of women in the unorganised sector, the lowest female workforce participation of any comparable economy, and a personal law regime still organised by religion. The 106th Amendment illustrates the whole problem: it is in force since April 2026 and still cannot be applied, because Parliament tied it to a census and a delimitation that have not happened and then defeated the Bill that would have untied it. Rights on paper have never been India's difficulty in this field; the difficulty is that the institutions meant to deliver them are the same institutions in which the discrimination originates.

Krishna Iyer J. held that a prisoner does not become a non person or a slave of the State; that Part III does not stop at the prison gate; that solitary confinement of a prisoner under sentence of death before the sentence becomes final and executable is illegal; and that bar fetters may be imposed only for the shortest period and with reasons recorded and subject to review. The Court appointed District Judges as visitors and directed a grievance mechanism. The reasoning that carries the law is that imprisonment authorises the deprivation of locomotion, and nothing more: every other right survives unless its curtailment is a necessary consequence of incarceration.

What follows from that principle. Charles Sobraj v. Superintendent, Central Jail, Tihar, (1978) 4 SCC 104, held that a prisoner retains all rights except those necessarily lost by incarceration and that the Court will intervene where prison conditions offend Article 21. Prem Shankar Shukla v. Delhi Administration, (1980) 3 SCC 526, held that routine handcuffing is unconstitutional: it may be resorted to only where there is clear and present danger of escape, the reasons must be recorded, and the escorting officer must justify it to the court, because handcuffing is prima facie inhuman and violates Article 21 read with Article 14.

This established public law compensation for violation of fundamental rights, later developed in Nilabati Behera v. State of Orissa, (1993) 2 SCC 746, where the Court held that sovereign immunity is no defence to a claim for compensation for a custodial death under the writ jurisdiction, and in D.K. Basu v. State of West Bengal, (1997) 1 SCC 416, which laid down eleven mandatory requirements for arrest and detention, including a memo of arrest attested by a witness, information to a friend or relative, a medical examination every forty eight hours and entry in a diary, since incorporated into the Code and now into sections 47 to 58 of the BNSS.

The most recent development, and the most important. Sukanya Shantha v. Union of India, decided 3 October 2024, arose from a journalist's petition following her article showing that prison manuals across India assigned labour by caste, segregated barracks by caste, and contained provisions discriminating against denotified tribes and persons classed as "habitual offenders".

The framework and the gap. The Model Prison Manual, 2016 and the Model Prisons and Correctional Services Act, 2023, circulated to replace the Prisons Act, 1894, are advisory: prisons are a State subject in Entry 4 of List II, so the Union can only model and persuade. The record shows the gap. Prison Statistics India 2023 records about 1,332 prisons holding roughly 5.3 lakh inmates against a capacity for far fewer, an occupancy rate of 120.8 per cent, and 73.5 per cent undertrials. Section 479 of the BNSS is the significant reform: a first time offender who has served one third of the maximum sentence must be released on bond, and the duty to apply is placed on the jail superintendent rather than on the prisoner, and the Supreme Court has held the provision applies retrospectively. Whether it works depends on implementation, which is the recurring theme of this note.

(b) Civil liberties under pre-independence law

The framework was one of statutory permissions, not rights. Before 1950 there was no bill of rights. The Government of India Acts of 1919 and 1935 distributed legislative and executive power and said nothing about the liberties of the subject. What a person in British India had was whatever the ordinary law left him after a large body of security legislation had taken its share, and no court could hold an Act of the legislature void for infringing a liberty.

The security legislation. The Regulation III of 1818 (the Bengal State Prisoners Regulation) permitted detention without trial for reasons of state and was still being used in the twentieth century. The Indian Penal Code, 1860 contained sedition in section 124A, inserted in 1870, under which Tilak and later Gandhi were convicted; in his 1922 trial Gandhi described it as the prince among the political sections of the Code designed to suppress the liberty of the citizen.

The judicial position, and the one case that matters. The courts applied the statutes and could not question them. The characteristic decision is Emperor v. Sibnath Banerjee, AIR 1945 PC 156. Detention orders under Rule 26 of the Defence of India Rules were challenged; the Federal Court had held the Rule ultra vires in part and had scrutinised whether the Governor had personally applied his mind. The Privy Council reversed, holding that Rule 26 was intra vires the Defence of India Act, and that the requirement of satisfaction was satisfied by the ordinary rules of business under which a Secretary could act on behalf of the Government. The consequence was that judicial control over preventive detention in British India was reduced to checking whether the order was formally within the Rule. The contrast with Makhan Singh and with modern Article 22 jurisprudence is the point of citing it.

The measure of the change. Romesh Thappar v. State of Madras, AIR 1950 SC 124, decided within months of the Constitution, struck down a ban on the circulation of a journal imposed under the Madras Maintenance of Public Order Act, 1949, holding that a restriction on the ground of "public safety" or "public order" was not covered by the then wording of Article 19(2), which spoke of security of the State. This is the decision that led to the First Amendment, 1951, adding "public order" and "friendly relations with foreign States" to Article 19(2) and inserting the word "reasonable". Whatever one thinks of that amendment, the sequence shows what changed in 1950: a court struck down an executive order under a valid statute because a written guarantee of free speech now bound the legislature itself, which was impossible under any pre independence law.

(c) Security of state

The concept and its constitutional place. "Security of the State" appears in Article 19(2) as one of the grounds on which reasonable restrictions may be imposed on the freedom of speech and expression; in Article 22(3)(b) and the entries on preventive detention; in Entry 9 of List I, which gives Parliament exclusive power over preventive detention for reasons connected with defence, foreign affairs or the security of India; and in Entry 3 of List III, which gives the Union and the States concurrent power over preventive detention for reasons connected with the security of a State, the maintenance of public order or the maintenance of supplies and services essential to the community. That legislative division is itself a definition: security of India is a Union subject, security of a State and public order are concurrent.

Why the distinction has practical bite. It decides three things. One, legislative competence: which List a preventive detention law falls under. Two, the validity of an individual order: a detention order reciting only ordinary criminal conduct is made under the wrong head, and this remains the most common ground on which the Supreme Court quashes detention orders, most recently in cases where a detenu's activities amounted to individual offences however grave. Three, the reach of restrictions on speech: a restriction justified as protecting the security of the State must be aimed at something graver than public inconvenience.

The speech cases. Romesh Thappar v. State of Madras, AIR 1950 SC 124, held that the security of the State in the unamended Article 19(2) meant something far graver than ordinary breaches of public order, and that a law authorising restriction for the wider purpose could not be saved for the narrower one. Kedar Nath Singh v. State of Bihar, AIR 1962 SC 955, upheld section 124A of the Penal Code but read it down so that only words having a tendency to incite violence or to create public disorder amount to sedition, and mere criticism of the Government, however strong, does not; the security of the State ground cannot be stretched to cover disaffection without incitement.

Conclusion. Security of the State is the narrowest and gravest of the three expressions the Constitution and the detention statutes use, and Ram Manohar Lohia fixed its meaning with the image of three concentric circles: law and order at the circumference, public order within it, and security of the State at the centre. The distinction is not academic.

It allocates legislative power between Entry 9 of List I and Entry 3 of List III; it is the commonest reason a preventive detention order is quashed, because an order founded on ordinary crime has been made under the wrong head; and it limits how far speech may be restricted, since Romesh Thappar, Kedar Nath Singh and Shreya Singhal together confine restriction to incitement and refuse to allow the gravest constitutional ground to be used for ordinary disorder. Whether section 152 of the Bharatiya Nyaya Sanhita respects that boundary is the question now before the Supreme Court.