It is seldom that the much-mythologised Magna Carta is actually engaged. Yet in A Reference by the Attorney General for Northern Ireland [2026] UKSC 16 (the Deprivation of Liberty Reference) the Supreme Court arguably permitted the violation of one of the oldest constitutional rights on our statute book: that no person shall be detained except by the law of the land. That guarantee first appeared as clause 39 of the 1215 charter, and survives today as chapter 29 of the Magna Carta confirmed in 1297, which remains in force. Eight hundred years later, the Supreme Court held that a person who appears content with their confinement is not, in law, detained at all, and so no legal authorisation of safeguarding is required. The right was not, exactly, violated. It was defined away through sleight of law, and for those least able to articulate objection.

The right and its remedy

A right is only as strong as the means of enforcing it, and the lasting achievement of the English tradition was to give the guarantee in chapter 29 a remedy with teeth. That remedy was the writ of habeas corpus. When Charles I imprisoned the knights who refused his forced loan, and the court declined to look behind the King’s command in Darnel’s Case (1627) 3 St Tr 1, the answer was the Petition of Right 1628, which reasserted that no free subject could be imprisoned without cause shown. Sir Edward Coke, glossing ‘law of the land’ in chapter 29 as ‘due process of law’ in the Second Part of his Institutes, treated habeas corpus as the very instrument by which that ancient promise was made good, and the Habeas Corpus Act 1679 then placed the procedure on a statutory footing. Substance and remedy had become a single constitutional idea: detention must rest on a legal reason that can be stated and tested.

That idea adapted, almost unchanged in its logic, into Article 5 of the European Convention on Human Rights. Article 5(1) permits detention only on a lawful basis falling within defined categories, Article 5(2) requires that the person be told the reasons, and Article 5(4), the modern descendant of the writ, guarantees the right to have the lawfulness of detention decided speedily by a court. The domestic courts have kept faith with the older common law too. In R (Jalloh) v Secretary of State for the Home Department [2020] UKSC 4 Lady Hale opened, at [1], by recalling that ‘the right to physical liberty was highly prized and protected by the common law long before the United Kingdom became party to the European Convention on Human Rights’, and that a person unlawfully imprisoned ‘could, and can, secure his release through the writ of habeas corpus’.

What unites these eight centuries is not a prohibition on confinement, which no legal system could sustain, but a discipline upon it. The state may detain, but only if it can give a legal reason and open that reason to challenge. Yet the protective apparatus presupposes something easily overlooked: a person able to set it in motion, to articulate their reasoned objections. The question the Deprivation of Liberty Reference forces upon us is what becomes of the guarantee when the detained person cannot object. It is to that question, and to the statutory scheme built precisely to answer it, that we now turn.

Habeas corpus by proxy

The Mental Capacity Act 2005 governs the decisions of those who cannot advocate for themselves. Where a person lacks capacity, a decision may be made for them in their best interests (ss 2-4). Most of its application is unremarkable. The gravest exercise of substitute power is confinement, and here the Act is at its most guarded. A person may not be deprived of their liberty unless it is formally authorised (s 4A), whether by the local government through the Deprivation of Liberty Safeguards in a care home or hospital (Schedule A1) or by order of the Court of Protection (s 16), with a right to review the deprivation in that court (s 21A).

These safeguards are best understood as habeas corpus exercised by proxy. The person who is supervised, medicated and prevented from leaving may be wholly unable to say ‘I object’, still less to find a court. The independent assessment, the advocate and the route to a judge under s 21A are the means by which the law demands a reason on that person’s behalf and tests it before an independent forum. They are the modern, statutory form of the very discipline the writ secured for the prisoner who could speak for himself, applied to those who cannot.

P v Cheshire West and Chester Council [2014] UKSC 19 carried that discipline to its conclusion, and it did so partly by what it set aside. Strasbourg had identified three components of a deprivation of liberty: an objective component of confinement, a subjective component consisting in the absence of valid consent, and attribution to the state (Storck v Germany (2005) 43 EHRR 6, § 74). The Supreme Court treated the subjective component as inapplicable: a person who lacks the capacity to consent cannot give the valid consent it contemplates. Consent was, for this class of person, simply not among the possibilities, so the question turned on the objective component alone. On that question the Court held that a two-stage ‘acid test’ is requisite: a person is deprived of their liberty where subject to continuous supervision and control and not free to leave. As Baroness Hale put it, at [46], ‘what it means to be deprived of liberty must be the same for everyone, whether or not they have physical or mental disabilities’, for ‘a gilded cage is still a cage’. The comparator was not some notion of relative normality for the disabled, but the liberty of any other person, so that the protection of Article 5 was universal rather than scaled by impairment. Lord Kerr made the point in terms the Deprivation of Liberty Reference would later have to confront, at [76]: ‘Liberty means the state or condition of being free from external constraint. It is predominantly an objective state. It does not depend on one’s disposition to exploit one’s freedom. Nor is it diminished by one’s lack of capacity.’ The holding was deliberately protective. A benign setting and an absence of complaint were not to be mistaken for the absence of a deprivation of liberty.

This was itself an interpretation of Strasbourg jurisprudence. In HL v United Kingdom (2004) 40 EHRR 32 (Bournewood) a compliant, autistic patient was kept in hospital, the professionals having ‘exercised complete and effective control over his care and movements’ (§ 91), without formal authority or any means of challenge, and the want of those safeguards was held to violate Article 5. The lesson of Bournewood is the one the Deprivation of Liberty Reference now unsettles: that compliance is not consent, that the good faith of carers does not answer the absence of a check, and that those most in need of the writ’s protection are precisely those least able to ask for it.

The two innovations

The Deprivation of Liberty Reference made two moves, and they compound one another. The first was to dismantle the acid test. In place of Cheshire West’s two simple questions, the Court restored the older, contextual assessment associated with Guzzardi v Italy (1980) 3 EHRR 333, under which whether a person is deprived of their liberty turns on the type, duration, effects and manner of the measure, weighed in the round. Presented as a return to Strasbourg orthodoxy, this is less innocent than it looks. A multifactorial balance is, by design, less predictable than a bright line, and it hands the judgment largely to those who arrange the care. More significantly, it quietly reopens the comparator question that Cheshire West had closed. Once the assessment is contextual again, it becomes possible to ask whether confinement is merely ‘normal’ for a person with this or that impairment, which is the very relativism the universal comparator was meant to forbid.

The second move is the constitutionally significant one, and it reaches directly for the component Cheshire West had set aside. The Court held that valid consent, the subjective element of a deprivation, is to be determined independently of capacity under the Mental Capacity Act. Where Cheshire West had treated valid consent as beyond the reach of a person who lacks capacity, so that it could never negate their deprivation, the Deprivation of Liberty Reference makes it available to them and, indeed, decisive. The practical result is that a person who presents as content may be treated as having validly consented to their confinement, and so as not deprived of their liberty at all, even though that same person has been assessed as lacking the capacity to decide the very question.

Consent that nobody can give

The Court was straightforward about what, on its approach, valid consent comes to. A person who lacks capacity in a full legal sense may still, it reasoned, be able to register whether they are content, and that is enough. As the judgment puts it, at [135], such a person ‘may have impaired understanding of many things and may lack capacity in a full legal sense, but they may nevertheless have a fundamental understanding of whether they are broadly happy or unhappy’. On that footing, a person who manifests acceptance of their situation is treated as having validly consented, thus not deprived of their liberty at all.

Here the reasoning fails, and it fails on its own terms. A finding that a person lacks capacity in relation to their care is, precisely, a finding that they can neither give nor withhold a valid consent to it. To treat their apparent contentment as that consent is to keep the word while discarding the capacity that alone gives it meaning. It is to indulge the kind of unreality Lady Hale identified in Jalloh, where she dismissed as ‘completely unreal’ the suggestion that a person under constant constraint was ‘a free agent, able to come and go as he pleased’ ([27]). A consent read off a person’s surface contentment, severed from the relationships and conditions that shape what they express, is not the autonomous consent that Article 5 requires.

The autonomous consent doctrine sits uneasily with a principle the Strasbourg court has affirmed since the Vagrancy Cases. In De Wilde, Ooms and Versyp v Belgium (1979-80) 1 EHRR 373 three destitute men had presented themselves to the police and asked to be detained, and the Court rejected the suggestion that men who had reported voluntarily could not have been deprived of their liberty, holding that ‘the right to liberty is too important in a “democratic society” within the meaning of the Convention for a person to lose the benefit of the protection of the Convention for the single reason that he gives himself up to be taken into detention’ (§ 65). Fully capacitous submission, in other words, is not waiver.

The Deprivation of Liberty Reference does not overlook that principle: it cites it, at [122], in the form restated by the Grand Chamber in Stanev v Bulgaria (2012) 55 EHRR 22, that the right to liberty is too important to be lost ‘for the single reason that [a person] may have given himself up to be taken into detention’, and ‘especially when he is legally incapable of consenting to, or disagreeing with, the proposed action’. But having stated the principle, the Court at once turns it aside, reasoning that a lack of legal capacity does not prevent a person from understanding and accepting their situation, and building from there to the conclusion that manifested contentment is itself valid consent. The difficulty is that the emphasised clause, ‘especially when he is legally incapable of consenting to, or disagreeing with’, treats incapacity as a reason for greater vigilance, not as the doorway to a thinner kind of consent. The Court has kept the authority and reversed its force: a line that once held submission insufficient to defeat Article 5 is made to support the proposition that an incapable person’s mere contentment removes its protection altogether. The protection is withdrawn from the most dependent, the most heavily supervised, and the least able to register an objection; from exactly the people for whom it was devised.

The discipline of reasons

Lon Fuller, in The Morality of Law, located the morality internal to law in features such as its generality, publicity and constancy, all of which presuppose that official action can be articulated and defended rather than merely imposed surreptitiously, and he warned that every departure from law’s inner morality is ‘an affront to man’s dignity as a responsible agent’ (p 162). Habeas corpus and Article 5(4) are these requirements made operational in the field of detention. They are demands for reasons, backed by a forum in which the reasons can be tested.

The constitutional cost of the Deprivation of Liberty Reference is that by making the classification of a deprivation turn on apparent contentment, the Court removes the trigger that sets the whole apparatus in motion. Where there is no deprivation there is no duty to justify, and where there is no duty to justify there is no occasion for challenge. A person who is supervised, medicated and prevented from leaving, but who is classified as a consenting non-detainee, occupies exactly the condition the tradition exists to prevent.  It is no answer that the old law had become unworkable. The administrative burden was real, but a constitutional right marks the limit of what the state may do. It is not a variable to be adjusted until it fits the resources the state is willing to commit. To let the burden of compliance set the scope of the right to liberty is to invert the relationship between power and its constraints, and to do so, here, at the expense of those least able to register the loss.

The state may interfere with liberty. The whole achievement of the tradition that runs from chapter 29 through the writ to Article 5 is to insist that when it does, it must do so in ways that can be tested before a court. Habeas corpus secured that demand for the prisoner who could invoke it; the Deprivation of Liberty safeguards extended it to those who could not. The Deprivation of Liberty Reference quietly relaxes the demand for the second group, by treating their want of an articulated objection, or their contentment, as consent, and by allowing the scope of their right to be set by those who hold power over them. Instead of a formal legal procedure, detention can be based on an inference by the state that an incapacitous person is content. It is the arbitrary power that constitutional rights, since Magna Carta, were built to restrain.

Craig Wells is a lecturer in constitutional law, legal theory and social work law at the University of Kent.

(Suggested citation: C. Wells, ‘The Discipline of Reasons: Liberty, Vulnerability and Arbitrary Power in A Reference by the Attorney General for Northern Ireland [2026] UKSC 16’, U.K. Const. L. Blog (15th June 2026) (available at https://ukconstitutionallaw.org/))