IS the defence of insanity available to a defendant who, owing to a disease of the mind leading to a defect of reason, believed that his actions were being externally controlled? In Keal [2022] EWCA Crim 341, [2022] 4 W.L.R. 41, the Court of Appeal answered, “No”, holding that any other answer would require an inappropriate departure from the M'Naghten rules (1843) 8 E.R. 718. As will be seen below, the court's conservative view of its own powers is understandable. Yet there is little binding authority on the question arising in Keal. It was not addressed when the House of Lords endorsed M'Naghten in Sullivan [1984] A.C. 156. Moreover, even if a trip to the Supreme Court be required, there is much to be said for expanding the insanity defence to cover at least some cases of “command delusion”.
Robert Keal attempted to kill his father, mother and grandmother with weapons including knives, scissors and a cricket bat, during a sustained attack at their shared home. There were moments during the attack when Keal apologised to his victims and stated that he was unable to stop himself. To his mother, at one point he said, “I'm sorry, this isn't me, it's the devil”. Keal had a long history of severe mental illness and, on the day before the attack, had attempted suicide and been hospitalised. When found by authorities after the attack, he was in a disturbed state, in a country road wearing only his underpants.
At trial, the psychiatric experts agreed that Keal was seriously unwell and in a psychotic state at the time of the offence. He satisfied the first two requirements in M'Naghten that a defendant who claims insanity must prove that he was, at the time of the alleged offence, (1) suffering from a “disease of the mind”, (2) leading to a “defect of reason”. However, the experts also agreed that Keal understood the “nature and quality” of his actions, ruling out a defence under limb (3a) of the rules. In order to qualify as insane, therefore, Keal had to prove the alternative limb, (3b) that he “did not know that what he was doing was wrong”; a test that was itself paraphrased in M'Naghten in terms of the accused's not being “conscious that the act was one which he ought not to do”. Experts disagreed on this last issue, and Keal was convicted.
On appeal, defence counsel rehearsed an argument, rejected by the trial judge, that Keal's belief that he lacked control over his actions was relevant to whether he satisfied limb (3b). It was argued that the jury should have been directed accordingly. Counsel proposed a two-part interpretation of limb (3b). Keal would fail to satisfy the wrongness requirement, it was submitted, if “i) at the time he knew he was doing something he should not do; and ii) he chose nonetheless to do it”. The jury might be directed in such terms, and that “If on the balance of probabilities you conclude that his delusion was such that he believed he had no choice, then the defence of insanity is made out” (at [27], emphases added).
In support of this proposal, counsel offered two kinds of arguments, a “narrow” and a “broad” approach. On the narrow approach, it was argued that recognising command delusions is compatible with the text of M‘Naghten itself. For Keal to know that his “act was one which he ought not to do”, he would have to know that it was his act. Keal's belief that he lacked agency (because it was Satan who was directing his bodily movements) precluded such knowledge. This imaginative argument was supported by reference to a frequently overlooked passage in M'Naghten where it is said that defendants labouring under a partial delusion, but who are not insane, must “be considered in the same situation as to responsibility as if the facts with respect to which the delusion exists were real”. Keal's delusion, the argument ran, meant that his lack of agency should be treated as “real” in this sense.
Alternatively, counsel argued for a broad approach: more sweeping reform of the M'Naghten rules by introducing an additional (3c) “loss of capacity to control” limb to supplement the existing two cognitive limbs, in line with the Law Commission's Discussion Paper proposal (Insanity and Automatism, 2013). The Commission's proposal is familiar, and the focus here will be on counsel's narrower, more novel approach. Like the broad approach, it too was robustly rejected by Lord Burnett L.C.J., who observed (at [47]) that it was confusing to define wrongness in subpara. i):
by reference to knowledge of wrongdoing; whilst in sub-paragraph ii) it is defined by reference to “choice” …. The fact that the proposed direction is difficult to understand is indicative of the underlying conceptual difficulty in the appellant's case of fitting the element of “choice” into the concept of “wrongdoing” and knowledge of wrongdoing.
Moreover, hopes for change were said to founder on “established authority” in R. v Kopsch (1927) 19 Cr App R 50, which dismissed lack of control defences as based on “the fantastic theory of irresistible impulse”. It was also noted that Parliament had failed to implement the Law Commission's recommendation to introduce a lack of control defence, despite legislating in cognate areas, and that a single appeal did not allow for proper reconsideration of the law.
Doubtless the proposed direction could have been improved. Counsel's narrow approach nonetheless has considerable merit. Analysing cases of command delusion as primarily a cognitive matter, such that the defendant's belief in his lack of agency means he cannot be said to know he is doing anything (and, a fortiori, anything wrong), fits within an interpretation of the knowledge of wrongdoing limb that does not require the court to “find” an entirely new defence based on volitional considerations. It does not seem odd for D's belief about his involuntariness to qualify as insanity via limb (3b) of M'Naghten, provided it results from a relevant disease of the mind. The remarks in M'Naghten about partial delusions only buttress that interpretation.
Kopsch, the sole contrary authority cited, is distinguishable. In contrast to Keal's unambiguous illness, Kopsch's criminal urge was merely “suggested” to come from his “subconscious mind”. Indeed, venerable as it is, Kopsch is not compelling authority against the proposed change. The judgment is terse, lacking analysis, and is focused on rejecting lack of control as a separate, volitional limb.
Some may question the practical value of this reinterpretation. Either way, Keal goes into hospital. But rejecting the narrow approach meant that Keal became stigmatised not as somebody unwell but as a convicted criminal, with implications for his employment, moral standing and ability to seek civil redress for any mismanagement of his case by mental health providers (Henderson v Dorset Healthcare [2020] UKSC 43, [2021] A.C. 563). Were there clear policy reasons for this, such as the need to symbolise the defendant's wrongdoing through conviction, that might be a valuable outcome. Yet the trial judge's decision to issue a hospital order in lieu of a prison sentence is tantamount to admitting that criminal justice has no place here, whether in exacting retribution or protecting public safety. Counsel was right to question whether the case can fit with notions of culpability. Keal was seriously ill, not criminally culpable. Trial courts and juries already recognise the problem: they frequently ignore the present law to achieve justice for the unwell (R. Mackay “Righting the Wrong? Some Observations on the Second Limb of the M'Naghten Rules” [2009] Crim. L.R. 80). The appellate courts remain firmly in the nineteenth century.
To be sure, the Court of Appeal had reason for caution. It is the penultimate forum of appeal and has ruled on knowledge of wrongdoing several times over the past century. The same cannot be said of the Supreme Court. Except for Sullivan [1984] A.C. 156, which decides nothing in relation to matters of control or knowledge of wrongdoing, there is no modern apex court authority on this fundamental doctrine. Admittedly, there are guidelines on judicial expansion of defences in cases like C v DPP [1995] 2 All E.R. 43, at [36], which look ominous for the proposed change:
(1) if the solution is doubtful, the judges should beware of imposing their own remedy. (2) Caution should prevail if Parliament has rejected opportunities of clearing up a known difficulty or has legislated, while leaving the difficulty untouched. (3) Disputed matters of social policy are less suitable areas for judicial intervention than purely legal problems. (4) Fundamental legal doctrines should not be lightly set aside. (5) Judges should not make a change unless they can achieve finality and certainty.
Ultimately, though, these principles admit the need for some judicial creativity. And the courts’ reticence in relation other defences need not carry over to the instant case, which is suited to boldness for at least three reasons. First, the special verdict and the introduction of hospital orders for the guilty mean that a finding of not guilty by reason of insanity poses no additional risk to public safety. Second, the proposed development is limited. It remains consistent with a “cognitive” theory of insanity in M'Naghten. Third, Parliament's legislative history cuts both ways. It has willingly left in place rules that are confusing, open-textured and frequently ignored, something that could be read as giving wide latitude to courts rather than committing to the status quo. While comprehensive parliamentary reform of the doctrine would be preferable, this should not preclude the Supreme Court, at least, from mending the law when cases like Keal provide the opportunity.
