Hohfeldian Rights and the Sloppy Indian Use of "Right"

Wesley Newcomb Hohfeld published "Some Fundamental Legal Conceptions as Applied in Judicial Reasoning" in 1913. His point was small in print and large in implication. The word "right," as used in everyday legal speech, is a homonym. It does at least four different things, and it is responsible for a recurring class of confusions in judicial reasoning. The four things are a claim (the correlative of a duty in someone else), a liberty (the absence of a duty in oneself), a power (an ability to alter another's legal position), and an immunity (the inability of another to alter one's own legal position).¹ Hohfeld's prescription was modest. If lawyers are going to use the word "right" to mean all four, they should at least pause long enough to ask which of the four they mean. The Indian Supreme Court has, in fifty years of building a rights jurisprudence, very rarely paused.

The four ideas Hohfeld kept apart

Dias, in his standard exposition of the scheme, gives the cleanest entry point. Claim, liberty, power, and immunity each answer a different question.² A claim in X is a sign that Y ought to do something — its correlative is a duty in Y. A liberty in X means that X has no duty to refrain from doing something — its correlative is the absence of any claim against X. A power in X is the ability to change Y's legal position — its correlative is a liability in Y to be so changed. An immunity in X is the absence of any power in Y to change X's legal position — its correlative is a disability in Y.

The four are conceptually independent. To say I have a liberty to wear a bowler hat is not to say that anyone has a duty not to interfere; the liberty and the claim are separable. As Dias observes, X may by contract extinguish his claim against Y not to be interfered with, while keeping the liberty itself intact. If X then succeeds in wearing the hat, he has exercised his liberty. If Y prevents him, X cannot complain — the claim is gone, even though the liberty is not.¹⁰ Or take Bradford Corporation v Pickles: a landowner has the liberty to extract subterranean water, but no claim against another extracting it before it reaches him.⁸ Or Musgrove v Chun Teeong Toy: an alien at common law has the liberty to enter British territory, but no claim not to be prevented.⁴ Liberty and claim are different things. A jurisprudence that does not separate them collapses a question about what X may do into a question about what Y must not do, which is a separate question and frequently has a different answer.

Powers and immunities raise distinct issues again. The power to make a will is an ability to alter legal positions on death. The immunity from criminal prosecution that a sitting head of state enjoys in some systems is the absence of any power in the courts to disturb that position. Neither of these is a claim or a liberty in the Hohfeldian sense. To call them "rights" without distinction is to lose the analytical traction Hohfeld was offering.

Chhatrapati Singh, writing in the Journal of the Indian Law Institute in 1985, pressed a further point. Basic legal relations are typically triadic, not dyadic. The state is the third term in most legal relationships, and the dyadic Hohfeldian scheme tends to mask that fact.⁶ A policeman has a duty to the state to arrest, and a power against the suspect to do so. The constraint and the enabling element are not the same relationship, but they get confounded once both are called "the right to arrest." Indian constitutional doctrine, as we shall see, has a deep version of the triadic problem.

The Indian default: everything is a "right"

Pick up any major Indian constitutional law decision of the last forty years and you will find the word "right" doing at least three of Hohfeld's four jobs simultaneously, without flagging the shifts. The Court speaks of a right to privacy, a right to free speech, a right to dignity, a right to die, a right to be forgotten, a right to internet access, a right to sleep. Each of these is treated as comparable to every other, and the analysis proceeds on the assumption that what is true of one will be true of the others.

This is not pedantry. The reason Hohfeld's distinctions matter is that they sort the questions the court must answer. A claim-right requires the identification of the correlative duty-bearer; if there is no duty-bearer, there is no claim-right, only a liberty. A liberty-right requires us to identify whether the surrounding duties have actually been removed or whether they remain in force. A power-right requires us to identify what the legal position is that the holder can alter, and the limits within which the alteration counts as valid. An immunity-right requires us to identify the actor who is being disabled. The Court frequently fuses these, with the result that questions about correlative duties, about scope, about who is restrained, get conflated into a single rhetorical move: the recognition of the "right."

Three lines of Indian doctrine illustrate the cost. They are not the only three; they are three where the consequences are most visible.

Line one: privacy, Puttaswamy, and the missing correlative

The nine-judge bench in Justice K.S. Puttaswamy v Union of India declared privacy a fundamental right under Article 21.³ The judgment is widely admired, and on its central holding it is correct. But the structure of the right it recognised is not analysed with Hohfeldian care, and the result is that the doctrine has been hard to apply.

Read carefully, Puttaswamy moves between three of Hohfeld's four relations within the same paragraph. The right to privacy is presented as a claim against the state — the state ought not to surveil, ought not to collect personal data without authorisation, ought not to intrude into the home. That is a claim in the Hohfeldian sense; the correlative is a duty in the state. But Puttaswamy also speaks of privacy as a liberty — an individual is free to make intimate choices about sexuality, reproduction, association. The correlative there is the absence of a claim in anyone else (including the state) to interfere. And the judgment further speaks of privacy as an immunity — the state cannot, by ordinary legislation, abridge core elements of privacy without satisfying proportionality. That is a Hohfeldian immunity; the correlative is a disability in the state to legislate certain content.

Three different relations, three different correlatives, three different questions about what the doctrine of privacy authorises a court to do. In Puttaswamy itself the conflation does not bite, because the case is a declaratory one. It bites in the cases that have followed. When privacy is invoked against a private actor — a media organisation, an employer, a service provider — the Court has had to choose between treating privacy as a claim with a private duty-bearer, a liberty defensible against private interference, or an immunity from regulatory action. The case law oscillates. Section 8 of the Right to Information Act creates a privacy ground for exemption from disclosure; the High Courts and Information Commissions read this sometimes as a claim against the state to withhold, sometimes as a liberty of the third party to keep information hidden, sometimes as a third-party immunity. The opinions read as if they are all about the same thing. They are not.

The deeper version of the problem is this. Privacy as a claim against the state is straightforward — the duty is in the state, the correlative is clear. Privacy as a liberty against private actors is harder, because the liberty by itself does not generate any duty in those actors not to interfere. Dias's bowler-hat example is exactly the point. To establish a claim against private actors, the Court would need to find or create a duty, either through tort doctrine or through statutory grounding. Puttaswamy did not do this; it treated the recognition of the right as if it did the work of the correlative duty automatically. Cases like Aadhaar leaks, telecom interception complaints, and platform-data complaints have been litigated on the assumption that the existence of the "right to privacy" generates all the necessary duties at once. It does not. The liberty is one thing; the claims that would support it against various actors are a series of further questions, each of which Puttaswamy left open.

Line two: free speech and the liberty/claim conflation

Article 19(1)(a) is one of the most carefully drafted provisions in the Constitution: a freedom of speech and expression subject to the reasonable restrictions in 19(2). The drafting is, on its face, Hohfeldian — "freedom" suggests liberty rather than claim. The Court's treatment of Article 19 has nevertheless moved steadily from liberty toward claim, and the doctrinal consequences have been awkward.

The clearest illustration is Shreya Singhal v Union of India, which struck down Section 66A of the IT Act.⁵ The decision is rightly celebrated. But notice the move. Section 66A criminalised certain online speech; the Court held that the criminalisation violated Article 19(1)(a). On a liberty reading, the Court was saying: there is no duty in the speaker to refrain from this speech, because Section 66A is not a valid 19(2) restriction. On a claim reading, the Court was saying: the speaker has a claim against the state to be left alone, with a correlative duty in the state. The judgment, in different paragraphs, says both.

The conflation has consequences when Article 19 is invoked against private platforms. If 19(1)(a) is a liberty against the state, it generates no duty on Twitter, X, YouTube, or Facebook not to remove content. If it is a claim, the duty-bearer must be identified. The horizontal-rights question, which has bedevilled the Court for two decades, is in significant part a Hohfeldian question that the Court has not framed in Hohfeldian terms. Recent judgments have ruled, sometimes, that platforms have obligations of natural justice when removing content. Where do those obligations come from? On a liberty reading, they cannot come from Article 19 itself; they must come from elsewhere. On a claim reading, they come from Article 19, but at the cost of treating private platforms as having state-like duties — which raises a different problem about why they should be so treated.

The Court has not addressed this cleanly. Instead, it has spoken of free speech as a "value" that informs the interpretation of all law, which is true but unhelpful. The reason it is unhelpful is that the same value is being asked to do two different jobs — liberty against the state, claim against private actors — and the analytical work that distinguishes the two has been folded into the rhetoric of "value." Hohfeld's distinction is the analytical work, and it has not been done.

Line three: dignity, and the disappearance of correlatives

Of the three lines, the most acute is dignity. Since the late 1990s the Court has treated dignity as a foundational element of Article 21, and over the last decade dignity has become an engine of doctrine in its own right. Privacy, autonomy, sexual orientation, end-of-life choices, religious practice — each has been derived from dignity. Common Cause in 2018 rested the doctrine of advance medical directives on dignity.⁷ Joseph Shine in the same year rested the decriminalisation of adultery on dignity.⁹ Indian Young Lawyers Association rested temple-entry rights on dignity.¹¹

Dignity is the hardest of the three to fit into Hohfeld's scheme, and that is not an accident. The judgments treat dignity sometimes as a claim against the state (the state ought not to humiliate), sometimes as a liberty (the individual is free to make choices that respect her own dignity), sometimes as a power (the individual can alter her own legal position by, for instance, advance directives), and sometimes as an immunity (no law can intrude on a core of dignity-bearing autonomy). All four Hohfeldian relations are present in the dignity jurisprudence, and they are present without distinction. Correlatives are not identified. Duty-bearers are not specified. Liabilities are not stated.

This is not because dignity is a richer concept than Hohfeld can capture. It is because the Court has used dignity as a placeholder that absorbs each of the four relations at once, depending on what the case needs. In Common Cause, the advance-directive holding implies that the individual has a power to alter her own legal position prospectively, and that the state has a corresponding liability to give effect to that alteration. The judgment does not say this; it says that dignity requires it. In Joseph Shine, the decriminalisation holding implies that there is no longer a duty in either spouse to refrain from adultery, and that adultery has moved from the criminal-prohibition column to the liberty column. The judgment again does not say this; it says that dignity requires it. In Indian Young Lawyers Association, the temple-entry holding implies a claim in the excluded group against the temple authorities, with a correlative duty in those authorities. Or it implies a liberty of the excluded group to enter, with the question of whether the authorities have a duty not to interfere left open. The judgment, again, does not distinguish.

The cost of this is that dignity, as it now functions in Indian constitutional reasoning, is unfalsifiable. It can produce any of the four Hohfeldian relations on demand, but there is no internal test for which relation it is producing. A doctrine that can move freely among claim, liberty, power, and immunity, depending on the result the bench wants, is not a doctrine in the analytical sense; it is a license. This is not a criticism of any particular outcome — most of the outcomes are defensible on their own terms — but a criticism of the reasoning. The reasoning does not survive translation into Hohfeld's vocabulary because once translated the question becomes: which of the four is in play, and what is the correlative? The answer differs by case, and the Court has not, in fifty years, organised the answer.

Two replies, and what they leave standing

Two replies to this analysis deserve attention.

The first reply is that Hohfeld's scheme is itself inadequate. Chhatrapati Singh's 1985 critique made the point that the scheme is dyadic and therefore misses the triadic character of most legal relations. Lyons's puzzle — that one can have both a right and a duty to vote — is a puzzle only because the Hohfeldian scheme presents the relations as dyadic; once the state is introduced as a third term, the puzzle dissolves.⁶ Other critics have pressed similar objections. So perhaps, the reply runs, the Court is right to operate at a higher level of generality than Hohfeld provides; the four relations cannot capture what Indian constitutional doctrine is doing, and the failure to fit is a problem for Hohfeld, not for the Court.

There is some force here. The dyadic limitation is real, and Indian constitutional rights, especially the ones that involve the state as both duty-bearer and norm-creator, do not always reduce cleanly to Hohfeld's eight elements. But notice what the reply asks the Court to do. It asks the Court to operate above Hohfeld's level — to articulate the rights jurisprudence in some richer vocabulary that captures triadic relations. The Court does not do this either. It does not operate above Hohfeld; it operates below him, in the homonymous use of "right" that Hohfeld was specifically trying to repair. To say that Hohfeld is inadequate as a corrective is one thing; to use the inadequacy as cover for the indistinction Hohfeld was diagnosing is another.

The second reply is that legal doctrine does not have to be analytically clean to be useful. The basic-structure doctrine is not analytically clean, and it has done significant work in Indian constitutional law for fifty years. The same may be true of the rights jurisprudence: a homonymous "right" that absorbs claims, liberties, powers, and immunities depending on context may be a feature rather than a bug, because it allows the Court to fashion remedies that a more analytically precise vocabulary would foreclose.

This is the more substantive reply, and it deserves more than a quick dismissal. There is a real argument that the open-ended use of "right" has allowed the Court to do justice in cases where a careful Hohfeldian analysis would have produced no remedy, or would have transferred the question to the legislature for additional structure. Privacy against private actors, free-speech protection against private platforms, dignity against social practices — each of these has produced outcomes that ordinary citizens benefit from, and each was reached through reasoning that, in Hohfeldian terms, was unsound.

The reply has force, and I will not pretend it has none. But the price of the open-ended use is not zero. It is paid in three forms. First, in inconsistency: the same "right" can support directly opposing outcomes in two cases because the relation being invoked is, in one case, a claim and, in another, a liberty. Second, in unpredictability: lawyers cannot tell their clients what the right does, because the right does different things in different cases. Third, and most seriously, in legitimacy: a court that derives outcomes from a doctrine whose internal structure shifts case-to-case looks to outside observers like a court that is reaching results first and reasoning afterwards. The Indian Supreme Court has, in the last decade, attracted exactly this charge in respect of its rights jurisprudence. The Hohfeldian critique is not the whole of the charge, but it is a substantial part of it.

What a Hohfeld-aware court would do differently

A court alert to Hohfeld's distinctions would not abandon the language of rights. It would clarify, in each major judgment, which of the four relations the right at issue is. It would identify the correlative in the same paragraph in which it identifies the right. It would tell the reader, when it derives a right from a higher constitutional value such as dignity, whether the derived right is a claim with a state duty-bearer, a liberty with the surrounding duties removed, a power with a specified scope of alteration, or an immunity with a defined disability in the regulator.

None of this is hard. It is, in fact, what well-reasoned constitutional judgments in other common-law jurisdictions routinely do — sometimes self-consciously, sometimes by long inheritance of the analytical tradition. The Indian Court is capable of doing it; in several judgments — Khanna J's opinion in Kesavananda, Chandrachud J's dissent in ADM Jabalpur, the structural reasoning in I.R. Coelho — it has shown that it can. The point is not that the Court lacks the tools but that it does not use them with the consistency that the rights jurisprudence requires.

The cost of the indistinction, finally, is felt outside the courtroom. Lawyers advising clients on what the right to privacy, or the right to free speech, or the right to dignity protects them against, must hedge in proportion to the doctrine's hedging. Trial courts and High Courts must guess at which of the four relations the Supreme Court has in mind when it pronounces a new right, and they often guess wrong. Legislative drafters, designing statutes to comply with the constitutional rights jurisprudence, draft to the homonym rather than to the underlying relations, with the result that the statutes too are unstable when challenged.

Where this leaves the homonym

Hohfeld's contribution to legal thought was, in the end, modest. He did not propose a new theory of law. He proposed a vocabulary. The vocabulary kept four ideas apart that ordinary speech welded together. Lawyers who used the vocabulary made fewer mistakes; lawyers who did not, made more. The Indian Supreme Court has, on the whole, not used the vocabulary, and the rights jurisprudence carries the cost of the omission. Privacy, free speech, and dignity are the three most prominent places where the cost is visible, but they are not the only three. The question of what an Indian "right" is — claim, liberty, power, immunity, or something that switches among the four — is a question the Court has, decade after decade, declined to answer at the level of doctrine. The reasoning would survive translation if the answer were given. Until it is given, the reasoning does not.

Get the week's biggest decisions, distilled.

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