Retaking the English Castle: A Legal Critique of Restore Britain's Self-Defence Paper

Restore Britain's policy paper Retaking the English Castle: Self-Defence in an Era of Anarcho-Tyranny arrived in 2025 with the kind of rhetorical confidence that tends to divide legal opinion before it has even been read. Authored by Harrison Pitt and Rupert Lowe MP, it makes five specific legislative proposals aimed at restoring what it describes as the English tradition of ordered freedom, specifically the right to defend yourself, your family, and your home without being treated by the state as a greater threat to public order than the person who broke in through your window.

Stripped of its more combative framing, the paper raises legal questions that are both serious and overdue. This article provides an analytical critique of each proposal, identifies where the paper succeeds and where it falls short, and directs readers to a full cost-benefit analysis I have prepared as a companion document, available here.

My own position, set out plainly: self-defence within the home should be an absolute right. Where a householder kills an intruder and can demonstrate that the force used was responsive to a genuine threat, that death should be legally justified. The current framework does not go that far, and neither does this paper. Both fall short of where the law ought to be.

The Legal Baseline: What the Law Currently Says

Before assessing the proposals, it is worth stating clearly what English law currently permits and where the gaps lie.

The foundational provision is section 76 of the Criminal Justice and Immigration Act 2008, which codified the common law self-defence standard. A person acts lawfully in self-defence if they use such force as is reasonable in the circumstances as they believed them to be. The subjective element is significant: a householder acting on a genuine but mistaken belief that they were under threat is not necessarily stripped of the defence.

The Crime and Courts Act 2013, specifically section 43, amended this for householder cases. It introduced a higher threshold for disentitling a householder from the defence: force must be grossly disproportionate, rather than merely disproportionate, to fall outside the protection. This was intended to give homeowners more room to act decisively. In practice, the distinction was substantially narrowed by the High Court in R (Collins) v Secretary of State for Justice [2016] EWHC 33 (Admin), which held that the ultimate test for juries remains whether the force used was reasonable in the circumstances as the householder believed them to be. Grossly disproportionate force is never reasonable; disproportionate force might or might not be, depending on those circumstances.

The practical result is a standard that looks clearer than it is. Householders still face prosecution. Juries still have to navigate the gap between what felt necessary in a moment of genuine terror and what appears proportionate to a court months later. The law gives with one hand and qualifies with the other.

Proposal I: Legalising Pepper Spray

The paper proposes repealing section 5(b) of the Firearms Act 1968 and section 6 of the Offensive Weapons Act 2019 to permit law-abiding adults to carry pepper spray for self-defence.

The case is well-founded. The Ministry of Justice data cited in the paper is primary and credible: nearly 4,638 convictions over approximately nine years for possession of a non-lethal defensive substance, with an average custodial sentence of nine months in 13% of cases. Those are real people with real criminal records for carrying something that most of continental Europe permits without notable adverse consequences.

The argument for empowering vulnerable people is also the strongest in the paper. A formal right to self-defence is functionally worthless to a woman alone at night if she lacks any practical means of exercising it. Legalising pepper spray does not arm society; it closes a gap between what the law permits in theory and what it makes possible in practice.

The paper's principal weakness here is the absence of any proposed regulatory framework. Unqualified repeal is unusual even among jurisdictions that permit civilian carry. A minimum age restriction, a maximum concentration limit, and licensed point-of-sale controls would significantly reduce the misuse risk without undermining the proposal's purpose. As drafted, the proposal is correct in direction but incomplete in detail.

Proposal II: The Stanley Amendment

This is the strongest proposal in the paper, and the one most likely to command cross-party support if separated from the wider ideological context.

The Occupiers' Liability Act 1984 imposes a duty of care on occupiers towards non-visitors, including criminal trespassers. The Court of Appeal applied this in Revill v Newbery [1995] 2 WLR 239, holding that Ted Newbery, who wounded a burglar while defending his property, was civilly liable to that burglar under the 1984 Act despite being acquitted of criminal charges. Newbery paid £4,000 to the man who had come to rob him.

The rejected Stanley Amendment would have inserted a single clause: no duty of care is owed to a person entering premises having committed or intending to commit a criminal offence there. Lord Hailsham dismissed it as unnecessary. The Newbery case, and cases like it, demonstrated otherwise.

The paper is right to call for its passage. The common law principle of ex turpi causa non oritur actio (no claim can arise from a criminal act) should bar a burglar from suing a homeowner in these circumstances. It does not, because the 1984 Act effectively displaced it in the civil context. The Stanley Amendment restores it.

One drafting adjustment I would recommend: preserve residual civil liability where force used would satisfy the criminal standard for gross disproportionality. This addresses the Article 2 ECHR tension without undermining the Amendment's core purpose.

Proposal III: Repealing Scotland's Hate Crime Act

The substantive case against the Hate Crime and Public Order (Scotland) Act 2021 is genuinely strong. The Act removed the dwelling defence, enabling prosecution for speech within a private home. Section 5(1) authorises search warrants for domestic premises on suspicion of stirring up hatred. The thresholds are vague. The chilling effect on private expression is real.

Where the paper goes wrong is the mechanism. It proposes using section 35 of the Scotland Act 1998 to repeal the Act from Westminster. Section 35 has never been used. Its terms relate to blocking a Bill before Royal Assent, not repealing enacted legislation. The paper acknowledges this requires an amendment to section 35 itself, but does not assess the constitutional implications of that move. It would represent the most significant Westminster intervention in devolved legislation since the Scotland Act came into force, and would face immediate judicial review.

The problem is real. The mechanism is legally uncertain and carries a political cost disproportionate to the outcome sought. A UK-wide statutory dwelling defence to speech offences, enacted under reserved Article 10 ECHR obligations, would achieve the same result without dismantling the architecture of the devolution settlement.

Proposal IV: Preserving Trial by Jury

The paper's defence of the jury system is constitutionally sound and practically important. David Lammy's proposal to replace juries in the majority of criminal cases with single judges would, as the paper correctly observes, save approximately 0.2% of the Ministry of Justice budget at the cost of a constitutional safeguard with centuries of accumulated legitimacy.

The specific self-defence argument is the weakest evidentially. The paper contends that defendants pleading self-defence fare better before juries than before judges trained in diversity, equity, and inclusion orthodoxy. The intuitive case has force: juries bring lived experience of fear and threat to the assessment of what a reasonable person does in a moment of genuine danger. But the paper provides no data on acquittal rates in bench versus jury trials for self-defence pleas, and the claim about judicial DEI training influencing verdicts is asserted rather than evidenced.

The constitutional case stands regardless. The Jamie Michael acquittal, reached by a jury in 17 minutes, is a useful illustration. The paper is also right that the Crown Court backlog is not primarily caused by juries, even if it does not adequately engage with how that backlog would be addressed under its proposals.

The proposal to require the personal consent of the Attorney General before any prosecution may proceed for alleged excessive force in self-defence is technically the most detailed in the paper, and the statutory analysis is largely accurate. The model drawn from section 27 of the Public Order Act 1986 is a legitimate and workable precedent.

The structural weakness is exposed by the paper's own evidence. Lord Hermer, as Attorney General, used his section 27 consent power to authorise the prosecution of Lucy Connolly for online speech following the Southport atrocity. Attorney General consent is not a structural safeguard; it is a political one. Its effectiveness depends entirely on the political character of the incumbent.

A supplementary mechanism would address this: reversing the evidential burden in self-defence prosecutions, requiring the prosecution to disprove reasonable belief in the necessity of force to the criminal standard. That would provide a protection independent of whoever holds the office of Attorney General at any given time.

Where the Paper Falls Short of Where the Law Should Be

My own position goes further than any of these proposals. Self-defence in the home should be absolute. Where a householder kills an intruder and can demonstrate to the criminal standard that they were responding to a genuine threat, that killing should be lawful. Full stop.

The current law does not provide this. The 2013 Act introduced a higher threshold but retained the reasonableness requirement. The Collins decision narrowed even that. A householder who kills a burglar in a moment of genuine, adrenaline-driven terror may still face prosecution if a court later concludes the force was not reasonable in the circumstances. That is an intolerable position.

The paper gestures toward this view without committing to it. The Stanley Amendment addresses the civil dimension; the AG consent requirement addresses prosecutorial gatekeeping. Neither addresses the fundamental point: the right to defend your home should not require justification to a standard calibrated in the calm of a courtroom.

Absolute householder defence would require the United Kingdom to address its position under Article 2 of the ECHR, which imposes a positive obligation to protect life. That is not an insurmountable obstacle. The ECHR permits lethal force in defence of persons from unlawful violence. Providing that the force was genuinely defensive and not grossly excessive, that framework is compatible with a strong householder protection. The political will to legislate for it, rather than the legal architecture, is what has always been absent.

Conclusion

Retaking the English Castle is a document that rewards careful legal reading. Three of its five proposals, the Stanley Amendment, pepper spray legalisation, and the AG consent requirement, are targeted, low-cost, and legally defensible. They address real injustices with real precedent behind them. Separated from the paper's broader rhetorical framework, they are capable of serious parliamentary advocacy.

The Scotland proposal is constitutionally overreaching. The jury trial proposal is constitutionally sound but evidentially incomplete. Both are fixable.

What the paper cannot fix, because it does not attempt to, is the deeper problem: that English law still treats the householder as someone who must justify themselves to the state for defending their home. Until that changes, the castle remains breached.

I have prepared a full cost-benefit analysis of all five proposals, assessing each against direct cost, evidential quality, legislative feasibility, and ECHR compatibility. It is available as a downloadable document here.

Jason Dillon LLB LLM is an AI law specialist and the writer behind Professor Nomad, a publication covering law, technology, and life on the road. Nothing in this article constitutes legal advice. If you are involved in a self-defence matter or a dispute arising from any of the issues discussed, seek advice from a qualified solicitor.