The Hart-Fuller debate is the 1958 Harvard Law Review exchange over whether a rule must meet a moral standard before it counts as law. Hart said it need not; Fuller said a system that fails eight principles of legality has not made law at all.
In 2026, a nine-judge bench of the Supreme Court led by Chief Justice Surya Kant reserved judgment in the Sabarimala reference after sixteen days of argument. One of the seven questions referred to it asks what the word “morality” in Articles 25 and 26 of the Constitution covers, and whether it takes in constitutional morality at all. The Union has argued that it means public morality and nothing more. Strip the Indian names off that question and it is the one Hart and Fuller were arguing about seventy years ago: when a legal text uses a moral word, does the standard come from the community outside the law or from the legal order itself.
This article works through both positions and the case that started the argument, then follows the same question into Indian constitutional law.
The whole debate turns on what a judge does when a properly enacted rule produces an outcome the judge believes is wrong, so it is worth working through one fact pattern before the theory arrives. Suppose a state legislature passes an Act which is published in the official gazette on the very day it takes effect, and which makes it an offence to have criticised a named public authority at any time in the preceding two years.
A shopkeeper is prosecuted under it for a remark he made eighteen months ago, in a conversation with two customers. He did not know the Act was coming. Nobody did, because it was not in the gazette until the morning the police acted on it.
Hart’s answer is that the Act is law, and that saying so does not commit anyone to obeying it; the shopkeeper’s remedy lies outside the question of validity, and in India it lies in Article 20(1), which forbids conviction for an act that was not an offence when it was done. Fuller’s answer is that a rule published on the day it bites, reaching backwards two years into conduct nobody could have adjusted, fails so many of the conditions of rule-making that the legislature has not managed to make law at all. Both routes end in an acquittal here. The sections that follow are about why the reasoning still matters when the routes diverge.
Hart’s case in the Hart-Fuller debate for keeping law and morality apart
Hart’s case for keeping law and morality apart rests on a single distinction: a rule can be legally valid and morally indefensible at the same time, and calling such a rule invalid hides the choice a judge is actually making. He was not defending the content of any law. He was defending a way of describing the situation a lawyer is in when the law is bad.
The occasion matters, because most notes pages get it wrong. Hart delivered the Holmes Lecture at Harvard Law School in 1957, and the lecture ran as “Positivism and the Separation of Law and Morals”, 71 Harvard Law Review 593, in the February 1958 issue. Fuller’s reply began at page 630 of that same issue.
So this was not two scholars circling each other across a decade. It was a visiting Englishman telling an American faculty that their jurisprudence had gone soft, and a member of that faculty answering him in print within months.
Hart was defending a proposition he traced to Jeremy Bentham and John Austin, which is that the existence of law is one thing and its merit or demerit another. That inheritance is often used to tar him with everything Austin’s command theory contained, and the tarring is wrong. Hart rejected the command theory outright, and The Concept of Law is largely an argument that law is a union of primary and secondary rules rather than orders backed by threats. What he kept from the analytical school was the separation thesis alone, which is the load-bearing claim in every later version of legal positivism.
It helps to fix what the two men were and were not disagreeing about, because the article turns on this. Neither of them thought a wicked statute deserved obedience. Neither thought morality had no influence on law. The disagreement is narrower and harder: whether the question “is this law?” and the question “is this good?” are two questions or one, and it is a disagreement about the relationship between law and morality rather than about the merits of any rule.
The limits of Hart’s separation thesis
The separation thesis claims only that a moral defect does not by itself remove legal validity, and it stops well short of the four things it is usually accused of claiming. Those four words, “by itself”, carry the whole position. Drop them, as a great many summaries do, and Hart becomes someone who denies any connection between law and morality, which he never was.
Here is what the thesis does not say. It does not say that law and morality never overlap; most legal systems prohibit murder, and Hart thought that unsurprising rather than coincidental. It does not say that legislators are uninfluenced by moral conviction, or that an unjust law should be obeyed. And it does not say that judges never reach for moral standards, because Hart’s own account of hard cases says the opposite.
What it says is that none of that overlap is what makes a rule legally valid. Validity comes from the system’s own criteria of identification, from what Hart later called the rule of recognition, and a rule that satisfies those criteria is law however repellent its content. A judge who dislikes it is in the position of someone with a legal duty and a moral reason to breach it, which is an uncomfortable position and, in Hart’s view, an honest one.
Why fight so hard for a distinction this narrow? Because Hart thought the alternative disarms the critic. If a wicked rule is not law at all, there is nothing there to condemn, only something to ignore, and the citizen who obeyed it did not fail a moral test because no test was ever set. Keeping the two questions apart is what preserves the sentence Hart wanted every lawyer to be able to say, that this is law and it is too evil to be obeyed.
The minimum content of natural law
The minimum content of natural law is the small set of rules Hart says every viable legal system must contain, and he derives it from five facts about human beings rather than from any moral order. It sits in chapter IX of The Concept of Law, published in 1961, and it is the part of Hart that Indian answer scripts most often leave out.
The five truisms are human vulnerability, approximate equality, limited altruism, limited resources, and limited understanding and strength of will. People can be hurt, and can hurt each other. No one is so much stronger than everyone else that he can dominate them indefinitely without cooperation.
People are neither angels nor devils; they are capable of both help and harm. Goods are scarce, so some system of holding and transferring them is needed. And people do not always see or do what is in their own long-term interest.
Given those five facts, Hart says, any legal system that is going to last must contain rules restricting the free use of violence, some form of property rule, and some rule requiring promises to be kept. A system without them is not immoral so much as unworkable; nobody would have reason to belong to it.
Note carefully what the argument is and is not. It is a claim about survival, not about validity, so a system can contain the minimum content and still be as vicious as the Nazi legal order was. That is precisely why Hart thought he could concede it without conceding anything to Fuller.
The core and the penumbra of a legal rule
The core of a legal rule is the set of cases its words plainly cover, and the penumbra is the set of cases where the words neither plainly cover nor plainly exclude the facts. Hart’s illustration has outlived almost everything else in the 1958 essay. A local ordinance says no vehicles in the park. A motor car is in the core, while a bicycle, a pair of roller skates and a child’s toy motor car sit somewhere in the penumbra.
He was aiming at two targets at once, which is why the example gets used for opposite purposes. Against the American legal realists, who said rules never really decide anything and judges dress up their preferences afterwards, Hart said there is a core and the core does decide. Against the formalists, who said a rule mechanically settles every case within its subject matter, he said the penumbra exists and the judge has to choose. In the penumbra the judge exercises discretion, and the discretion is guided by what the rule ought to be, given its purpose.
That is a substantial concession and Hart made it openly. So does it give the game away? He did not think so, and the reason is worth stating precisely, because it is where Fuller aimed his attack.
Deciding what a rule ought to cover in a penumbral case is a judgment about what the law should be; it does not turn the law into whatever it ought to be, and it leaves the core untouched. The line held for Hart. Whether it holds at all is the subject of Fuller’s reply.
Hart on the grudge informer case and the Radbruch formula
Hart used the grudge informer case to show what a court gives up when it invalidates a statute on moral grounds instead of saying plainly that the statute was law and should not have been obeyed. This is the passage everyone remembers from the 1958 essay, and it is also the passage where the facts turn out not to be what he thought they were.
The story as Hart had it runs like this. In 1944 a German woman denounced her husband to the authorities for insulting remarks about Hitler that he had made while home on leave from the army. She wanted to be rid of him because she was involved with another man. He was tried, sentenced to death, and in the event sent to the front rather than executed.
In 1949 she was prosecuted in a West German court for unlawfully depriving him of his liberty, an offence under the German Criminal Code of 1871, which had remained in force throughout the Nazi period. Her defence was that she had done nothing the law of the time forbade.
Hart understood the court to have convicted her by holding that the Nazi statutes she relied on were, as the judgment put it, “contrary to the sound conscience and sense of justice of all decent human beings”. He was not pleased. Calling that result a triumph of natural law over positivism struck him as, in his word, “hysteria”.
Two honest routes were available instead: leave her unpunished, or pass a frankly retrospective statute and punish her under it “with a full consciousness of what was sacrificed” in doing so. Retrospective criminal legislation is an evil, he accepted. But choosing it openly at least tells the truth about the choice, and the sentence he wanted the court to be able to say is the one that gives the essay its point: laws may be laws but too evil to be obeyed.
The grudge informer case as the Bamberg court decided it
The Bamberg court of appeal convicted the wife of unlawfully depriving her husband of liberty, and it did so without holding that the Nazi statutes were void for conflict with natural law. The reasoning is more technical than the legend, and once you have it, a lot of the standard exam answer on this case falls apart.
Start with the trial court, which acquitted her. Its reasoning was that her report and her husband’s subsequent detention came about because he had broken a valid law, and because a properly constituted court-martial had then convicted him through a properly conducted process. On that view she had set nothing unlawful in motion.
The court of appeal reversed, and its ground was that the trial court had wrongly inferred the legality of her report from the legality of the court-martial. She had used the court-martial as an instrument, which made her an indirect perpetrator of the deprivation of liberty. She knew that a report of this kind led inexorably to a range of outcomes running from a year’s imprisonment to death.
The distinction the court drew is the one that decided the case, and it is a distinction between two legal duties rather than between law and morality. She was under no legal duty to inform. Once her report reached the court-martial, that tribunal was under a duty to decide as it did, and it did no more than apply the positive law as it was meant to be applied.
So the judges could be left unincriminated while the informer was convicted, which is visibly the result the court was steering towards, and the reasoning is contrived for that purpose. It is still, at every step, legal reasoning.
The court did enter one reservation, and the reservation is Radbruch’s. It accepted that a statute could in principle lack the force of law for conflict with natural law, but held that these statutes did not reach that pitch, because they commanded only an omission, that is, making no public remarks of a certain kind.
Gustav Radbruch, who had held a broadly positivist view of law before 1933, had concluded in short articles published in 1945 and 1946 that extreme injustice is not law and that such statutes should be left out of account in working out anyone’s legal position. Hart’s answer to Radbruch was blunt. He called the position “extraordinary naivety”, and said Radbruch had only half digested the liberal message he was trying to deliver to the German legal profession.
The correction Hart accepted in a footnote
Hart accepted, in a footnote added to the reprinted essay, that H.O. Pappe had shown his account of the case to be wrong in several respects. This is the single most useful thing an Indian student can carry into an exam on this topic, because almost every notes page still repeats Hart’s version as though it were the record.
What Pappe showed, shortly after the 1958 exchange, is that the German courts dealing with the grudge informer cases did not adopt a higher-law argument at all. They worked through the interpretation of the 1871 Code, concentrating on the absence of any duty to inform and on the private setting in which the remarks had been made, and reached the conclusion that there had been an unlawful deprivation of liberty on those grounds.
Hart’s response was partly generous and partly a shrug. He summarised the corrections in the footnote, noting that the court had accepted the theoretical possibility that statutes might be invalid for conflict with natural law but had held that these ones were not, and that the accused was guilty because she had no duty to inform and reported for purely personal reasons. He said Pappe’s careful analysis should be studied. Then he added that the case as he had understood it could be treated as a hypothetical one, and carried on.
Does the correction sink the argument? Not by itself, and this is where care is needed. The dilemma Hart described, of a court asked to apply a rule that is valid and vile, is real whether or not the Bamberg bench faced it, and an Indian bench faced a version of it in 1976.
But the correction does two things worth carrying. It shows that a court with a natural-law escape route available preferred to decide by interpretation, which is closer to Fuller’s method than to the legend. And it means that anyone who writes “the German court held the Nazi statutes void under natural law” is writing something that did not happen.
Fuller’s answer in the Hart-Fuller debate on the inner morality of law
Fuller answered that law is not a set of rules a sovereign happens to issue but a purposive enterprise, and that the enterprise carries eight standards a rule-maker has to meet before there is any law to obey. That reframing is the whole of his reply. Once law is a thing people are trying to do rather than a thing that simply exists, it becomes possible for an attempt to fail.
His argument about the Nazi legal order is therefore not the one usually attributed to him. He did not say Nazi law was wicked and therefore not law; that is the argument Hart attacked, and it belongs to Radbruch rather than to Fuller.
What Fuller said is that the Nazi regime had stopped running a rule system in any recognisable sense. It governed through secret enactments, through retrospective decrees passed to cure illegalities already committed, and through a settled official practice of departing from the published text whenever the text was inconvenient. A citizen could not orient his conduct by it, which is the one thing a system of rules is for.
There is a detail here that almost no summary records, and it changes the shape of the disagreement. On the grudge informer, Fuller agreed with Hart that the retrospective statute was the better course. Where he differed was on what a retrospective statute would be doing: for Hart it would be an open sacrifice of a principle of legality to punish a wicked woman, while for Fuller it would be a way of marking that the regime’s decrees had never been law, which is a restoration of legality rather than a departure from it. That agreement on the practical answer, with total disagreement on its description, is what makes this a debate about the natural law tradition rather than about the German cases.
Law as a purposive enterprise
Fuller defined law as the enterprise of subjecting human conduct to the governance of rules, and the word enterprise carries the argument, because an enterprise can fail and can fail by degrees. Compare that with the positivist picture, in which a rule either satisfies the criteria of validity or does not, and the system either exists or does not. Fuller’s law is something a legislature is always in the middle of achieving, more or less well.
That is why he can say that a legal system is a matter of degree without saying anything mystical. A statute that is slightly obscure is worse law than a clear one, in the way that a badly built bridge is worse engineering than a good one, and nobody needs a higher law to say so. The standard is internal to the activity. It comes from what the activity is trying to do.
Fidelity to law is the second move, and it is aimed straight at the gap in Hart’s account. Hart can explain why a citizen is subject to law, but Fuller thought he could not explain why a citizen owes it anything, why the reasonable response to a legal order is loyalty rather than mere prudence. His answer is that a legal order which keeps to the eight principles has already given the citizen something, namely rules he can plan by and a state that binds itself. Fidelity is owed to that, and it is not owed to a regime that supplies none of it.
He then split the standards two ways, and the split is the answer to the obvious objection. A morality of duty sets the minimum a person must meet or be condemned; a morality of aspiration sets an excellence to be pursued. Most of the eight principles belong on the aspiration side, so a system with some obscure statutes and some inconsistent enforcement is still law and is simply worse at being law. Only total failure on a principle, secrecy rather than imperfect publicity, destroys the enterprise.
Fuller’s eight principles of legality
Fuller’s eight principles of legality are generality, promulgation, non-retroactivity, clarity, absence of contradiction, absence of impossible demands, constancy through time, and congruence between the declared rule and official action. They appear in The Morality of Law, published in 1964 and drawn from the Storrs Lectures he gave at Yale in 1963, which is six years after the exchange with Hart began.
He introduces them through a parable rather than a list, and the parable is doing real work. King Rex sets out to reform his kingdom’s law and finds eight distinct routes to disaster.
He begins by deciding every dispute ad hoc, so there are no rules at all. He then writes a code and keeps it secret. He publishes a code that governs only the past. He publishes a prospective code nobody can understand.
He clarifies it into a set of contradictions. He removes the contradictions by commanding things nobody can do. He relaxes the impossible demands by amending the code daily.
And when at last he has a clear, stable, published, possible code, he decides cases without reference to it. Fuller’s verdict on him is the point of the whole chapter: Rex never even succeeded in creating any law at all, good or bad.
Take the first three together, because they are the ones a citizen notices first. Generality means there must be rules rather than a series of individual commands, so that like cases are governed alike; a statute aimed at one named person is the limiting case of a failure here. Promulgation means the rules must be published before they are used, which is why an unpublished notification that binds from a back-date is a legality problem and not merely an administrative lapse. Non-retroactivity means a rule must speak to conduct that has not yet happened, since a rule addressed to yesterday cannot guide anybody.
The middle three concern whether the rule can be followed at all. Clarity means the rule must be intelligible to those who have to comply with it, and Fuller is careful that this is a demand on the drafter rather than an excuse for the reader. Absence of contradiction means the system must not require and forbid the same act, because a citizen faced with two incompatible commands has effectively been given none. Absence of impossible demands means the law must not require what cannot be done, which is the principle behind every rule that measures a duty by what was reasonably practicable.
The last two are about the rule-maker’s own conduct, and they are where Fuller’s argument stops looking like a drafting checklist. Constancy through time means the rules must not change so often that nobody can rely on them; a compliance regime rewritten every quarter fails this even if each version is perfectly clear.
Congruence, which Fuller treated as the most complex of the eight, means the rule as administered must match the rule as declared. It is the principle breached by selective enforcement, by an official practice that quietly departs from the published circular, and by a licensing authority that applies an unwritten criterion. Fuller thought congruence was where most legal systems actually decay, because it can fail without a single word of the statute book changing.
Set the eight against Indian constitutional practice and two of them are already there in the text. Article 20(1) forbids conviction for an act that was not an offence when it was committed, and it forbids a penalty greater than the one in force at that time. That is Fuller’s third principle, non-retroactivity, written into Part III and made enforceable. It is not a coincidence and it is not borrowing; it is the same insight arrived at independently, which is a large part of why Fuller thought the eight were discoveries about law rather than preferences about drafting.
There is one more use for the list, and it is the practical one. The eight give a lawyer a vocabulary for complaints that otherwise sound like grumbling. A retrospective tax demand, a circular applied before it is issued, a rule that is announced in one form and enforced in another: each of these is a specific failure with a name, and naming it is the first step in an argument about arbitrariness under Article 14.
Fuller’s reply on interpretation and purpose
Fuller called Hart’s core-and-penumbra account of interpretation untenable, because interpretation rarely turns on the meaning of an individual word. Hart’s picture assumes a judge looks at “vehicle”, finds a settled centre of meaning and a fuzzy edge, and works out which zone the case falls in. Fuller’s answer is that this is not what anybody does.
His counter-example is the one that has kept the argument alive for seventy years. Suppose some local patriots want to mount a working World War II truck on a pedestal in the park as a war memorial, and others object, citing the ordinance. The truck is a vehicle by any dictionary, and it is in perfect working order, so on the word alone it is squarely in the core.
Everyone nonetheless knows the ordinance does not touch it. Where did that knowledge come from, if not from the word?
It came from the purpose, and here is the part that is usually mis-stated. Fuller’s claim is not the modest one that hard cases need purposive reasoning; it is that purpose is doing the work in the easy cases too. We exclude a noisy car without thinking about it, as he put it, because we can see clearly enough what the rule is aiming at. If purpose is what makes the core a core, then the boundary Hart drew between finding the law and deciding what it should be does not sit where he put it, and that is the thread Dworkin picked up a decade later.
Hart’s 1965 review and Fuller’s reply to his critics
Hart reviewed The Morality of Law in 1965 and argued that Fuller’s eight principles are conditions of an efficient rule system rather than moral standards, and Fuller answered him in the revised edition four years later. This second round is skipped in almost every summary of the debate, which is unfortunate, because it is where the disagreement finally becomes a single testable question.
The 1958 exchange had left both men partly talking past each other. They were arguing about the German cases, about what a post-war court should have done, and about a set of facts that turned out to be wrong. By 1965 the German material has receded and what is left is a clean question: are the eight principles a morality, or are they simply what you have to do if you want your rules to work.
Hart’s charge that legality is efficacy, not morality
Hart’s charge was that a rule-maker pursuing wicked ends has exactly the same reason to follow the eight principles as a just one, which is why he refused to call them a morality. The review appeared at 78 Harvard Law Review 1281 in 1965, and it did more to fix Fuller’s reputation than anything Fuller wrote.
The argument is short and hard to shake off. Every purposive activity has internal standards of success, and following those standards makes you better at the activity without making you better. A rule-maker who publishes clearly, generalises properly and applies his own rules consistently will get more compliance than one who does not, and that is true whether he is running a welfare state or a system of racial exclusion. The principles are, in Hart’s terms, not valued for their own sake by those who follow them.
He then pushed it to the point that people remember. If the internal standards of an activity amount to a morality, then the principles of the poisoner’s art amount to a morality of poisoning, which is plainly absurd. Calling efficacy a morality blurs the distinction between being effective at a purpose and being good, and that distinction is the one Hart had spent his career defending.
Fuller’s reply came in the revised 1969 edition of The Morality of Law, in a fifth chapter written to answer his critics, and he told Hart in correspondence that this was the point he regarded as central to the whole controversy. His answer is that the eight are not neutral tools, because they constrain what can be done through law. A ruler who genuinely publishes his rules, keeps them general, makes them prospective and abides by them in his own courts has given up the instruments that make systematic oppression cheap; he cannot single out an individual, cannot punish yesterday’s conduct, and cannot say one thing and do another. The principles are other-regarding, and that is what makes them more than efficacy.
Neither man abandoned his position, and the honest report is that both propositions can be true at once. It is true that a tyrant has instrumental reason to publish clear rules, and it is also true that publishing them costs him something he would rather keep. Most positivists writing today concede that the eight are genuine constraints on what a legal system can be used for, while continuing to deny that the constraint is a moral one. That is a narrower disagreement than the one they started with, which is what a good debate is supposed to produce.
The Hart-Dworkin debate that followed
The Hart-Dworkin debate replaced this one as the central argument in English-language jurisprudence, and it moved the ground from the morality of law-making to the content of legal reasoning. It is a different debate, and treating it as a continuation of this one is a common error in exam answers.
Dworkin’s objection begins where Fuller’s attack on the penumbra left off. Legal systems, he said, contain principles as well as rules; principles have weight and are argued about rather than applied all or nothing; and a rule of recognition, which identifies law by its source rather than its content, cannot pick them out. If that is right, the penumbra is not a space where the judge has discretion to legislate. It is a space where there is still a legal answer, reached by argument rather than by choice.
The reason it matters here is that Dworkin’s target is Hart’s account of adjudication, not his account of validity, and the two are separable. A reader who has followed the Fuller exchange already has the essential move, which is the claim that purpose and principle are inside legal reasoning rather than outside it. Where the arguments come apart is on what follows: Fuller drew a conclusion about how rules must be made, and Dworkin drew one about how they must be read.
Indian courts have taken both sides of the Hart-Fuller debate
Indian courts have taken both sides of the Hart-Fuller debate, and the switch is datable: the Supreme Court read Article 21 positivistically until 1978, and it has read morality into constitutional adjudication ever since. That is not an analogy imported from an English seminar. The Constituent Assembly made the choice deliberately, in the text, and the Court then lived with the consequences for nearly thirty years.
The choice was between “due process of law” and “procedure established by law” in Article 21. The framers took the second, which imports the Japanese formulation and drops the American one, and the effect was to tell courts to ask whether a procedure had been enacted rather than whether it was fair. Put in the vocabulary of this article, the Assembly wrote the separation thesis into the liberty guarantee, and it did so on purpose after being warned what it would cost.
That is why Kesavananda Bharati v. State of Kerala, (1973) 4 SCC 225 is such an awkward case for the positivist reading, and why it belongs in this discussion rather than in a separate one about amendments. The Court held that Parliament’s power to amend the Constitution stops short of destroying its basic structure, which is a limit on a validly enacted law that no provision states.
But the Court did not find that limit in a higher moral order. It found it in the Constitution’s own scheme, in what the document was trying to be, which is a Fullerian move rather than a natural-law one. The basic structure doctrine is an internal standard, and the difference matters when you are asked to place it.
The positivist reading of Article 21 in Gopalan and ADM Jabalpur
Article 21 was read positivistically from 1950 to 1976, which meant that any procedure laid down by a validly enacted law satisfied it, however unfair that procedure was. The reading has two landmarks, and the second is the one that ended it.
A.K. Gopalan v. State of Madras, AIR 1950 SC 27, decided in the Court’s first year, came out of a habeas corpus petition against detention under the Preventive Detention Act, 1950. The Court held that “law” in Article 21 meant enacted law, that the procedure it established did not have to satisfy any independent standard of fairness, and that Articles 19, 21 and 22 occupied separate compartments. A person detained under a valid statute had no answer under Article 19 and no answer under any doctrine of due process, because the framers had refused to put one in.
Additional District Magistrate, Jabalpur v. Shivkant Shukla, (1976) 2 SCC 521 took the reading to its end point. During the Emergency the President had suspended the right to move any court for enforcement of Article 21, and the question was whether a High Court could still entertain a habeas corpus petition from a detenu who said his detention was not authorised by the detention law at all. The majority of four held that it could not: with the article suspended there was no locus to complain, whatever the legality of the detention. The judgment is the most criticised in the Court’s history, and Justice H.R. Khanna, who dissented and said that the State had no power to deprive a person of life or liberty without the authority of law, was passed over for the office of Chief Justice.
Read the majority against the grudge informer discussion and the parallel is exact, which is worth stating carefully, because the judges were not thinking in these terms and nothing in the judgment says they were. The majority gave the positivist answer to a question about a wicked law: the suspension was validly ordered, the consequence followed, and the Court’s job stopped there. Hart’s own objection to that posture is not that it is wrong about validity but that it is not the whole of a judge’s position. His prescription, that a court should say the law is law and refuse to lend itself to it rather than pretend the question does not arise, is close to what Khanna J did and to what the majority declined to do.
Maneka Gandhi and the turn to internal standards
Maneka Gandhi v. Union of India, (1978) 1 SCC 248 ended the positivist reading by holding that procedure under Article 21 must be fair, just and reasonable, and the Court found that standard inside the Constitution rather than above it. The case reached it on unglamorous facts. The petitioner’s passport had been impounded and the government declined to give reasons, relying on the public interest exception in the Passports Act, 1967.
The Court held that Articles 14, 19 and 21 are not watertight compartments but a single scheme, so a law that takes away personal liberty must also survive the tests of the other two. A procedure that is arbitrary or unfair is therefore no procedure at all for the purposes of Article 21. The Maneka Gandhi judgment dissolved Gopalan’s compartments, and with them the proposition that enactment alone is enough.
Notice what the Court did not do, because this is the point at which Indian jurisprudence usually gets described loosely. It did not discover a natural law standing above the Constitution and measure the statute against it. It read fairness out of the guarantee of equality in Article 14 and the reasonableness requirements in Article 19, which are provisions of the document itself. The standard is internal to the legal order, which is Fuller’s structure rather than Radbruch’s.
K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1 closed the loop forty years later. Nine judges held that privacy is a fundamental right, expressly overruled ADM Jabalpur, and Justice Kaul recorded his agreement with the overruling of what he called an aberration in the Court’s constitutional jurisprudence. The proposition that did the work is the one that goes furthest towards natural law of anything an Indian bench has said: that the Constitution is not the sole repository of rights, and that certain rights are recognised by it rather than created by it. Even there the Court grounded the holding in Part III, so the conclusion arrives inside the document rather than from outside it.
Constitutional morality and the Sabarimala reference
Constitutional morality is the idea that the morality relevant to constitutional adjudication comes from the Constitution’s own commitments rather than from prevailing social opinion. It is the doctrine that carries the Fullerian structure furthest in Indian law, and it is the doctrine now in front of nine judges.
The phrase entered Indian constitutional vocabulary through Dr B.R. Ambedkar, in his speech to the Constituent Assembly on 4 November 1948, where he quoted the classical historian George Grote at length while defending the decision to write administrative detail into the Constitution. His warning was that constitutional morality is not a natural sentiment, that it has to be cultivated, and that the Indian people had yet to learn it.
Navtej Singh Johar v. Union of India, (2018) 10 SCC 1 made the doctrine load-bearing. Reading down Section 377 of the Indian Penal Code, 1860 so far as it criminalised consensual sexual acts between adults, the Court held that popular or social morality cannot be a ground for restricting a fundamental right, and that the morality which counts is the constitutional one. The judgment is the clearest Indian statement that a legal order supplies its own standards, and it is very close to Fuller’s claim that the enterprise of law carries commitments the rule-maker cannot discard at will.
The same move was made twice more within a year. Indian Young Lawyers Association v. State of Kerala, (2019) 11 SCC 1 held the exclusion of women of menstruating age from the Sabarimala temple unconstitutional, and Joseph Shine v. Union of India, (2019) 3 SCC 39 struck down the adultery offence in Section 497 of the Indian Penal Code, 1860. In each, a practice with deep social sanction was measured against a standard drawn from the Constitution and found wanting.
That is the doctrine now under challenge. The Sabarimala judgment went into review, and in Kantaru Rajeevaru v. Indian Young Lawyers Association the Court referred seven questions to a bench of nine. Arguments began on 7 April 2026 before a bench led by Chief Justice Surya Kant, and judgment was reserved on 14 May 2026 after sixteen days.
The third referred question asks, in terms, what the scope and extent of the word “morality” is under Articles 25 and 26 and whether it is meant to include constitutional morality. The Union has urged that it means public morality; one amicus submitted that constitutional morality may help a court gauge the depth of a violation but cannot on its own be a basis for invalidating legislation.
The parallel is worth stating as a parallel and nothing more, because the Court has not framed it this way and no judgment should be predicted from it. If the bench holds that “morality” in Articles 25 and 26 means the morality of the community, it will have located the standard outside the legal order, which is where Hart always said moral standards sit. If it holds that the Constitution supplies its own morality for its own provisions, the standard is internal to the legal order, which is the structure Fuller was arguing for in 1958.
The question the nine judges have reserved is, on its face, about temple entry and the reach of Articles 25 and 26. What it settles underneath is where a court takes its moral standard from, and that is the question Hart and Fuller left open between 1958 and 1969.
Frequently asked questions
Who won the Hart-Fuller debate?
Neither position was abandoned, and the honest answer names what each side gave up. Hart conceded that judges decide penumbral cases by reference to purpose, and that every viable legal system contains a minimum content of rules protecting persons, property and promises. Fuller conceded that most of his eight principles are aspirations rather than duties, so a legal system can fall short of several of them and still be law. What survives is a narrower disagreement about whether the eight principles are a morality or the conditions of an efficient rule system.
What is the grudge informer case in the Hart-Fuller debate?
It is a 1949 West German prosecution of a woman who, in 1944, denounced her husband for remarks about Hitler in order to be rid of him, and who was convicted of unlawfully depriving him of liberty under the German Criminal Code of 1871. Hart used it to argue that a court should say plainly that a wicked statute was law rather than declare it void on moral grounds. The important qualification is that H.O. Pappe later showed the case had been misreported: the Bamberg court of appeal did not hold the Nazi statutes void under natural law, and it convicted her on the footing that she was under no legal duty to inform. Hart acknowledged the corrections in a footnote and treated his own version as a hypothetical.
What is the difference between Hart and Fuller on law and morality?
Hart held that legal validity and moral merit are separate questions, so a rule that satisfies a system’s criteria of identification is law however unjust it is. Fuller held that law is a purposive enterprise which carries eight internal standards of legality, and that a system failing them altogether has not produced law at all. The difference is about where the standard sits: for Hart it is outside the legal order and available for criticising it, and for Fuller it is inside the activity of making rules.
How is the Hart-Fuller debate relevant in India?
The Constituent Assembly chose “procedure established by law” over “due process” in Article 21, which put the separation thesis into the text, and the Supreme Court applied that reading through A.K. Gopalan and to its end point in ADM Jabalpur during the Emergency. Maneka Gandhi reversed the position in 1978 by requiring procedure to be fair, just and reasonable, and it drew that standard from the Constitution rather than from any higher law. The question is live again in the Sabarimala reference, where nine judges have reserved judgment on whether “morality” in Articles 25 and 26 includes constitutional morality.
References
- H.L.A. Hart, “Positivism and the Separation of Law and Morals”, 71 Harvard Law Review 593 (February 1958), delivered as the Holmes Lecture at Harvard Law School in 1957.
- Lon L. Fuller, “Positivism and Fidelity to Law: A Reply to Professor Hart”, 71 Harvard Law Review 630 (1958).
- H.L.A. Hart, The Concept of Law (Oxford University Press, 1961), chapter IX (Laws and Morals), on the minimum content of natural law.
- Lon L. Fuller, The Morality of Law (Yale University Press, 1964), chapter II (the King Rex parable and the eight principles of legality); revised edition, 1969, chapter 5 (A Reply to Critics).
- H.L.A. Hart, review of The Morality of Law, 78 Harvard Law Review 1281 (April 1965).
- David Dyzenhaus, “The Grudge Informer Case Revisited” (New York University School of Law), on the Bamberg Court of Appeal’s actual reasoning, H.O. Pappe’s correction and the footnote Hart added to the reprinted essay.
- Jeremy Waldron, “Positivism and Legality: Hart’s Equivocal Response to Fuller” (New York University School of Law), on the efficacy objection and Fuller’s 1969 reply.
- Frederick Schauer, “A Critical Guide to Vehicles in the Park”, 83 New York University Law Review 1109 (2008), on the core, the penumbra and Fuller’s war memorial truck.
- A.K. Gopalan v. State of Madras, AIR 1950 SC 27 (Supreme Court of India, 19 May 1950).
- Kesavananda Bharati Sripadagalvaru v. State of Kerala, (1973) 4 SCC 225 (Supreme Court of India, 24 April 1973).
- Additional District Magistrate, Jabalpur v. Shivkant Shukla, (1976) 2 SCC 521 (Supreme Court of India, 28 April 1976), including the dissent of Khanna J.
- Maneka Gandhi v. Union of India, (1978) 1 SCC 248 (Supreme Court of India, 25 January 1978).
- K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1 (Supreme Court of India, nine judges, 24 August 2017), expressly overruling ADM Jabalpur.
- Navtej Singh Johar v. Union of India, (2018) 10 SCC 1 (Supreme Court of India, 6 September 2018).
- Indian Young Lawyers Association v. State of Kerala, (2019) 11 SCC 1 (Supreme Court of India, 28 September 2018).
- Joseph Shine v. Union of India, (2019) 3 SCC 39 (Supreme Court of India, 27 September 2018).
- Kantaru Rajeevaru v. Indian Young Lawyers Association (Supreme Court of India, 14 November 2019), referring seven questions to a bench of nine judges.
- Supreme Court Observer, Sabarimala Reference, Day 16, recording that judgment was reserved on 14 May 2026 by a nine-judge bench led by Chief Justice Surya Kant after sixteen days of hearing, and setting out referred question 3 on the meaning of “morality” under Articles 25 and 26.
- Constituent Assembly of India, Debates of 4 November 1948, in which Dr B.R. Ambedkar introduced the expression “constitutional morality”, quoting George Grote.
This article is for informational and educational purposes only and does not constitute legal advice. Readers should consult a qualified advocate before acting on any statute, judgment or doctrine discussed above.