Dharma was not law and the equation has cost us

The first thing every law student in India is told about dharma is that it means law. The second thing is that the Indian Constitution stands in continuity with this ancient tradition; the third is that Yato dharmastato jayah, the motto of the Supreme Court, is the slogan of a continuous Indian commitment to the rule of law going back to the Brihadaranyaka Upanishad. The casual progression from dharma to law to the Constitution is performed every day in Indian classrooms and in Indian judgments. It is performed so often that it has stopped being noticed. It should be noticed, because it is wrong, and the wrong is now doing political work that the law schools do not seem to be tracking.

The position of this essay is direct. Dharma was not law. The translation of dharma as law is a habit inherited from a particular nineteenth-century colonial encounter; it is not a translation that survives careful reading of the Sanskrit texts that introduce the concept, and it does not survive the test of asking what kind of work the Indian Constitution is asking of its courts. The continued use of the equation has had consequences. It has lent civilisational authority to progressive readings of the Constitution that have nothing to do with the Constitution's actual textual grounding. It has lent equal civilisational authority to reactionary readings that treat the Constitution as a legible Hindu-traditional document. Both readings are made possible by the same translation error, and the error is no longer a matter of academic interest. It is a matter of which Constitution the country is going to read itself into in the decade ahead.

The shape of the casual equation

Begin with what is actually said. In an entry-level jurisprudence textbook, the standard line is that dharma means law and that the Bharatiya tradition placed dharma above the king, which is to say that the Indian tradition arrived at the rule of law a millennium and a half before Edward Coke. The textual warrant for the claim is the Brihadaranyaka Upanishad 1:4:14, which says that even kings are subordinate to dharma. The passage is real, and Radhakrishnan's gloss on it — 'Even kings are subordinate to Dharma, to the Rule of Law' — is the basis on which the equation has been built.³ The Mahabharata adds Yato dharmastato jayah. The Manusmriti adds dharma's four sources: Veda, Smriti, Sadachara, and the satisfaction of the dharmajna's inner conscience.² Justice M. Rama Jois's standard textbook builds the legal-history narrative around the proposition that ancient Indian dharma was the equivalent of modern law and that the king was its enforcement officer, not its source.⁴

The cumulative effect is a tidy story: dharma is law; ancient India had the rule of dharma; modern India has the rule of law; the Constitution is the modern medium of an old commitment. The Supreme Court has, in moments of expansiveness, signed up to this story. Ramaswamy J in Shri A.S. Narayana Deekshitulu v. State of Andhra Pradesh (1996) said in terms that the Constitution accords protection to dharma, identifying dharma with the 'core religion' and distinguishing it from religion proper.⁶ The motto on the Court's seal, Yato dharmastato jayah, is taken to be the seal of this continuity.

None of this would matter if the equation were doing only ceremonial work. The trouble is that it is doing substantive work, and the substantive work cuts in two opposite directions at once.

Why dharma is not law in any way the Constitution can recognise

Begin where P.V. Kane began, which is with the impossibility of the equation. Kane spent six decades on the History of Dharmashastra. The opening sentence of the first volume reads: 'Dharma is one of those Sanskrit words that defy all attempts at an exact rendering in English or any other tongue.'¹ Kane was not making a literary observation. He was making a structural one. Dharma in the Sanskrit corpus is a four-fold thing. It is srauta — the body of Vedic ritual prescription, the duties of sacrifice, the ordinances of caste and stage of life as derived from Vedic injunction. It is smaarta — the wider body of conduct derived from Smriti, governing the relations between persons and the obligations of station. It is rajadharma — the body of duties that bind kings. And it is vyavahara — the body of procedural and substantive rules by which disputes between persons are adjudicated.¹⁰

European law has a counterpart for one of these layers, vyavahara, and roughly for parts of a second, rajadharma. It has no counterpart at all for srauta. The first layer is not optional ornament that one can strip away to reach the legal core. The srauta layer is what holds the rest together. Manu's four sources of dharma are Veda first, then Smriti, then the conduct of the virtuous, then inner conscience — and the order is not interchangeable.² Where the Smriti conflicts with Sruti, Sruti wins. Where two Smritis conflict, custom decides. Where custom is absent, the king's discretion fills the gap, but the king is not a source of law; the king has no legislative competence.⁵ Katyayana puts it in terms that the textbook tradition usually omits: a king who acts according to his own fiat brings ruin to the people.

A Constitution that takes Veda as the first source of obligation is not the Constitution of India. A Constitution that places custom above legislative competence is not the Constitution of India. A Constitution in which the legislature has no authority to alter the Smriti is not the Constitution of India. The Constitution is, on Hartian or Kelsenian or any other modern positivist account, the supreme criterion of validity in the Indian legal system; the legislature has plenary competence under the document; custom is a residual source. The grammar is incompatible with the dharmashastric grammar at its foundations. To say that the Constitution stands in continuity with the dharma tradition is to say something that is not, in the technical sense, true.

How the translation happened

The translation error has a history, and the history is not innocent. The British in India needed to govern through some version of indigenous personal law. They needed pandits and maulvis at the colonial bench to advise on what the Hindu and Mohammedan laws required. They needed translations of the Manusmriti, the Mitakshara, the Dayabhaga, the Vyavahara Mayukha. The translators were not Sanskrit scholars working for the sake of Sanskrit; they were colonial functionaries producing usable manuals. The translators rendered dharma as law because law was the European concept that could be put to administrative use. The srauta layer fell away in translation because it was not useful for adjudication. The Sadachara layer was reduced to 'usage'. The vyavahara layer was retained as 'civil law'. Rajadharma was rendered as 'public law' or quietly omitted.

Dr Raghu Vira said it plainly in the Constituent Assembly: 'The fact is that Dharma never meant and can never mean religion. … But the Englishmen made a deliberate use of this for their own ulterior purposes.'⁷ The point Raghu Vira was making — that dharma had been deformed in colonial translation — is true. The conclusion the Constituent Assembly drew from it, that dharma must be rescued by identifying it with what the Constitution does, was a separate move, and a more questionable one. The translation that needed to be undone was being redone in a new register.

The progressive misuse

The progressive misuse of the equation is the easier one to see, because progressive constitutional interpretation in India has openly courted dharma in the last twenty years. The phrase 'constitutional morality', which has done the bulk of the heavy doctrinal lifting in cases from Naz Foundation to Navtej Singh Johar to Puttaswamy, is increasingly glossed in Indian classrooms and Indian commentary as a constitutional restatement of dharma. The gloss is offered as a defence of a particular reading of the document — that the Constitution carries within it a moral order that goes beyond its text and that the Court is entitled to read out of it. The argument has Western roots in Ronald Dworkin and George Grote, but the rhetorical force in India comes from the dharma connection. Constitutional morality, the gloss says, is dharma in modern dress. To say that the Constitution requires constitutional morality is to say that it requires dharma — and dharma is supreme; the legislature must yield; the king is subordinate to dharma.⁹

The progressive use of dharma in this way has done real work. It read down Section 377. It supplied the philosophical grounding for the privacy judgment in Puttaswamy.¹² It supplied the rhetorical framework for the Sabarimala majority. It has been used to justify the Court's readiness to override democratic majorities in defence of minority dignity. I do not, on the merits, disagree with most of these outcomes. The argument I am making is about the warrant for them, not their content. The warrant offered — that the Constitution embodies dharma and that dharma supplies the standard of constitutional morality — is doing too much work for what it is.

If dharma were a determinate moral category, the warrant might be defensible. Dharma is not a determinate moral category. Its content depends on srauta source, on Smriti source, on the conduct of the dharmajna, on the inner conscience of the cultured. Manu's Manusmriti 2:1 names the dharmajna's heart as a source, but it does so in terms that should give the progressive interpreter pause: the dharma sanctioned by the heart of 'the learned and the good, who are free from love and hate'.¹¹ The dharmajna is not the median citizen; the dharmajna is the dvija Brahmana versed in the Veda. The conscience that ratifies dharma is not constitutional conscience as we understand it. To borrow dharma's authority while ignoring its source-structure is to take its prestige without taking its content. The borrowing has worked, in the cases I named, because the Court has reached results most readers found defensible on independent grounds. The borrowing will keep working only as long as the results keep being defensible on independent grounds. The day they are not, the dharma rhetoric will be available to the other side too.

The reactionary misuse

The reactionary misuse is the one in the room. The argument from the Hindu Rashtra direction — call it the civilisational argument — runs as follows: India is a civilisation older than its 1950 Constitution; the civilisation has always been organised around dharma; the Constitution is the modern instrument of that civilisational order; the Court's motto Yato dharmastato jayah is the open acknowledgement of this fact. From these premises the move to a more substantive conclusion is short. If the Constitution is dharma in modern form, and dharma has historically privileged Hindu sources of normativity, then the Constitution should be read in continuity with those sources. The argument does not say the Constitution is a Hindu document; it says the Constitution is a dharmic document, and that dharmic means Hindu-traditional in a sense that includes the Veda, the Smriti, and the conduct of the cultured majority.

This argument is doing work in Indian public life in 2025. It is doing work in the framing of the new criminal codes — the Bharatiya Nyaya Sanhita, the Bharatiya Nagarik Suraksha Sanhita, the Bharatiya Sakshya Adhiniyam — whose Hindi-Sanskrit nomenclature is not innocent and whose preambular framing reaches for civilisational continuity. It is doing work in the recasting of constitutional history as a story of Bharatiya jurisprudence whose Western additions are corrigible. It is doing work in the Court itself, in the increasing willingness of certain benches to read dharmashastric authority into constitutional interpretation. The recent textbook by Seema Singh and Vinayak Sharma, from which much of the source material for this essay is drawn, identifies the golden triangle of Articles 14, 19 and 21 — the Maneka Gandhi triangle — with the trivarga of dharma, artha and kama. The identification is offered with the best of intentions. It is also exactly the move the civilisational reading wants made.

The reactionary use of the equation does not steal anything that the progressive use has not first authorised. Both rely on the same translation. Both insist that the Constitution embodies dharma. Both invoke Yato dharmastato jayah. The reading of dharma changes with the user. When the user is a Court protecting transgender rights, dharma means equality and dignity. When the user is a movement promoting Hindu Rashtra, dharma means civilisational majoritarianism. Both readings claim the same warrant. Both readings can — because the source-text is multi-layered and indeterminate — defend themselves by citing different layers. The translation has handed equal authority to incompatible programmes.

The strongest case for keeping the equation, addressed honestly

I owe the reader the strongest version of the defence of the equation before pushing on. The defence is this. Indian constitutional theory needs a vocabulary that connects the Constitution to the moral resources of the country it governs. The Western vocabularies — natural law, constitutional morality, dignity — are foreign, abstract, and have no traction with the median citizen. Dharma has traction. To strip the Constitution of dharma is to deprive it of cultural authority and to leave it floating on its own textual self-assertion. That is a real risk in a country where the textual self-assertion of the Constitution is contested. A Constitution that has only Hartian validity going for it will lose, over time, against a tradition with deeper roots. The argument was made eloquently by Granville Austin in his work on the Indian Constitution and it has been made in different registers by Upendra Baxi, by Marc Galanter, by Rajeev Dhavan, by every serious commentator on Indian constitutional culture.

The defence is honest, and it is not stupid. There is a real question about the cultural authority of the Constitution and a real question about whether the document can be defended only on positivist grounds. The constitutional patriotism that Pratap Bhanu Mehta and Sudipta Kaviraj have argued for needs more than text and procedure. It needs some claim about the values the document embodies and the reasons those values matter. The dharma vocabulary offers a way of making that claim in a register that does not feel borrowed from Locke and Kant. The Constituent Assembly itself, as the debates show, was not blind to this need; the framers reached for dharma in moments when they wanted to anchor the document in something older than the British procedural inheritance they had just thrown off.

What the defence cannot establish is that dharma is the right resource for this work. Dharma's traction is undeniable, but the traction is exactly what makes it dangerous. A concept that carries weight in the median citizen's mind is a concept whose meaning the median citizen will fill in. The median citizen in 2025 is not filling dharma in with constitutional-morality content. The median citizen is filling dharma in with content drawn from television serials, from political speeches, from temple discourses, from social-media reels. The content is increasingly indistinguishable from the civilisational-majoritarian content of the reactionary use. Borrowing dharma's authority for constitutional purposes is borrowing from an account whose other branches are not in friendly hands. The defence's premise — that the Constitution needs a culturally rooted vocabulary — is correct. The defence's conclusion — that dharma is that vocabulary — concedes the war the progressive position cannot afford to concede.

There is a further problem with the defence that deserves naming. The very feature of dharma that gives it cultural traction — its multi-layered, station-dependent, source-hierarchical structure — is the feature that makes it useless as a candidate for the supreme criterion of a modern republic. Dharma's content varies with the dharma-knower. The dharma of the king is not the dharma of the householder; the dharma of the Brahmana is not the dharma of the Shudra; the dharma of the renunciate is not the dharma of the warrior. A Constitution that took dharma seriously as a source would have to take this stratification seriously, and a Constitution that took the stratification seriously could not be the Constitution that Articles 14 and 17 actually committed the country to. The choice between dharma and the Constitution is, at the level of detailed normative content, a choice. The framers made that choice; they did not pretend otherwise; only the post-1980 academy has retconned the choice into a continuity.

What the Constitution actually has, and how dharma-talk crowds it out

The Constitution does not need dharma. It has its own resources. They are less culturally resonant; they are more textually grounded; they are not as easily wielded against an opposing political mobilisation. The resources are Article 14's guarantee of equality, Article 19's seven freedoms, Article 21's right to life and personal liberty as interpreted from Maneka Gandhi onward, Article 25's protection of conscience, Article 51A's catalogue of duties, the basic-structure doctrine, and the body of case law that has accumulated around them.⁸ These are real resources. The interpretive shift Chintan Chandrachud documents from textualism through structuralism to a more eclectic methodology has, even at its weakest, kept Indian constitutional jurisprudence rooted in the document and its scheme. The shift did not require dharma to do its work, and the cases that did the most important work — Kesavananda, Maneka Gandhi, Indira Nehru Gandhi, IR Coelho — did not call on dharma to ground their reasoning. They called on the structure of the document, the scheme of Part III, the limited nature of the amending power, the doctrine of basic structure that the Court had developed.

The cases that have called on dharma — Narayana Deekshitulu, parts of the Sabarimala majority, scattered observations in the Ayodhya judgment — have produced reasoning that is hard to defend on doctrinal grounds and easy to use against the result they wanted. Narayana Deekshitulu's identification of dharma with the core religion the Constitution protects is the doctrine that, slightly turned, becomes the framework in which Hindu Rashtra reads the Constitution as its own. The reasoning is the same. The reading is the same. The vocabulary is the same. The conclusions diverge only because the judges have, so far, kept the meaning of dharma in line with constitutional-morality content. The judges cannot keep doing that indefinitely. The pressure from the other direction is growing, and the rhetorical tools are common property.

There is a quieter cost too, less obvious than the political one. Every time a judge or a commentator reaches for dharma to do work that Article 14 or Article 21 could have done, the article does not get the doctrinal development it needs. The dharma move is a substitute for a more careful argument from the text and the scheme. The substitution is comfortable in the moment and corrosive over time. Indian constitutional law has, over the last twenty years, produced fewer rigorous doctrinal frameworks in the equality jurisprudence than the equality clause deserves. Part of the reason is that the heavy lifting has been done by dharma-adjacent vocabulary — constitutional morality, transformative constitutionalism, dignity — that signals a moral commitment without doing the doctrinal work of specifying what equality permits and forbids in concrete situations. The vocabulary has obscured the gaps. The gaps are still there. The argument is not that this vocabulary should be banished; the argument is that it should not be allowed to do the work that proper doctrinal analysis of the text should be doing.

The argument is not, then, that dharma should be banished from Indian constitutional discourse. It is older than the discourse; banishing it is neither possible nor desirable. The argument is narrower. Dharma should stop being translated as law. It should stop being identified with the Constitution. It should stop being invoked as the civilisational warrant for whatever the speaker wants the Constitution to mean.

What dharma is — a many-layered category of moral, ritual, customary and adjudicative obligation, varying with station, with stage of life, with caste, with regional usage, and grounded ultimately in Vedic source — is worth studying in its own terms, in Sanskrit, in the Smritis, in the commentarial tradition, in Kane and Rama Jois and Lingat. The study is a serious intellectual enterprise. It is not the same enterprise as Indian constitutional law. The Constitution does its own work. The dharma tradition does its own work. The two intersect at the level of cultural background but not at the level of doctrine, and treating them as overlapping doctrinally has produced the situation we are now in.

The situation we are now in is one in which the progressive reading of the Constitution and the reactionary reading of the Constitution both rest on the same translation error. The progressive reading uses dharma to ground a liberal-egalitarian content. The reactionary reading uses dharma to ground a Hindu-civilisational content. Both readings are entitled to the move because the move is sitting in every textbook, in every Supreme Court motto, in every Constituent Assembly speech that translated dharma as law. The move was made for honourable reasons. It has run its course. The reasons it was honourable in 1948 — to claim continuity with the country's intellectual heritage at a moment of severance from colonial authority — are not the reasons it is being used now. Now it is being used to settle which kind of India the Constitution constitutes, and the answer it produces depends on who is asking.

A Constitution whose meaning depends on who is asking is a Constitution that has not been theorised by its own scholarly community as rigorously as its situation now demands. The work of theorising it without leaning on the dharma translation has not been seriously begun. The legal academy in India has, for several generations, treated the equation as a settled background condition of constitutional interpretation. It is not settled, and it has not been a benign background. The cost of leaving it unexamined is being paid in the present in courts and in legislatures and in textbooks. Indian constitutional theory has owed itself a reckoning with dharma for forty years. The reckoning is overdue and the price of further delay is rising.

Get the week's biggest decisions, distilled.

A weekly digest for lawyers, students and the legally curious. Every Friday morning. Free.