IN R. v Golds [2014] EWCA Crim 748, the Court of Appeal was asked to clarify the meaning of “substantially impaired” in the partial defence of diminished responsibility in murder cases. By virtue of s. 2(1) of the Homicide Act 1957, as amended by the Coroners and Justice Act 2009, s. 52, diminished responsibility is made out where:
(1) … D was suffering from an abnormality of mental functioning which –
(a) arose from a recognised medical condition,
(b) substantially impaired D's ability to do one or more of the things mentioned in subsection (1A), and
(c) provides an explanation for D's acts and omissions in doing or being a party to the killing.
(1A) Those things are –
to understand the nature of D's conduct;
to form a rational judgment;
to exercise self-control. …
The facts in Golds were tragically redolent of many such cases. Golds suffered from schizophrenia, and was said to live in fear of “everything” and to hear voices criticising and tormenting him. His partner (Claire) had registered to become his carer. He stopped taking his medication and his condition got steadily worse. On the fateful day, according to other witnesses, there was a difficult meeting with Claire and her family, and Golds later assaulted Claire at their home. Claire told Golds to leave, and they argued further over a bank card. Golds himself did not remember what happened next, but he fetched a knife and said that he would kill Claire, which he did (22 separate knife wounds were found on her body). When the police arrived, he said that Claire “had Satan in her eyes”.
According to all the medical experts who examined him, by the time of the killing, Golds had probably entered into a state of dissociation in response to a high level of anxiety and stress. None of the experts doubted that he suffered from the delusion that Claire had become a demon. However, the trial judge left the vital question of whether his schizophrenia “substantially impaired” Golds's “ability to form a rational judgement” to the jury without any explanation of the word “substantially”. The jury convicted Golds of murder.
Golds argued on appeal that the judge should have offered a definition of the word “substantially”, to the effect that it should mean only “by more than a minimal degree”. But the Court of Appeal held that the word “substantially” means that something more is required, as one might mean in the phrase “substantially rewarded”. It was thus decided that there had been no misdirection and Golds's conviction for murder was upheld. It does not seem to have been argued that the trial judge had still prejudiced Golds's chances by not giving the jury any definition at all. We know that the word “substantially” at least does not signify a degree of impairment which is so high that the defendant is barely aware of what he is doing at all (R. v Byrne [1960] 2 Q.B. 396 and, more recently, R. v Ramchurn [2010] EWCA Crim 194, [2010] 2 Cr. App. R. 3) – but would an unguided jury necessarily understand that?
The court reached its conclusion on the meaning of “substantially impaired” purely by reviewing previous Court of Appeal authorities; and all those authorities concerned the phrase as it was used in the older version of diminished responsibility. It was assumed in R. v Brown [2011] EWCA Crim 2797 (where the jury accepted the partial defence and the defendant appealed only against his sentence) that the meaning of “substantially impaired” under the amended version remains the same. But this assumption can surely be doubted.
Under the old version of s. 2(1) of the Homicide Act 1957, which provided merely that “Where a person kills or is a party to the killing of another, he shall not be convicted of murder if he was suffering from such abnormality of mind … as substantially impaired his mental responsibility for his acts and omissions in doing or being a party to the killing”, there was (oddly) no requirement that the defendant's abnormality should actually have played a role in the killing. It was possible that a defendant whose abnormality substantially impaired him in some way might raise the defence even though he might have acted in the same way regardless of the abnormality. That is, there might be correlation between the abnormality and the killing, but no causation. Small wonder then that courts might have wished to reduce this risk when interpreting the word “substantially” – the more serious the impairing effects of D's abnormality, the less likely that it would have been merely a correlating factor.
However, the amended s. 2(1)(c) of the Homicide Act 1957 (above) introduced separately the element that the defendant's abnormality should have “provided an explanation” for what he did. Further, this element, as with the others, must be proved by the defendant. Surely then, in most cases where the defendant's abnormality is thought to have impaired him more than minimally, but to no greater degree than that, he will often fail the test of causation, and the outcome of the case will thus be unaffected by the definition of “substantially”.
That might still leave cases where the degree of impairment was relatively mild, and yet the abnormality can be proved to have had a causal link with the killing. Here, some may wish to blame the defendant for not fighting harder against the effects of his illness, and then everything might still hang on the proper interpretation of “substantially”. But it must be very difficult to ascertain how a mentally ill person could have combated the effects of his own abnormality, if indeed he is even aware of it. In many cases, it will be part and parcel of the medical condition that the sufferer has difficulty in accepting it or in following medical advice. It is submitted that if Parliament had wanted courts to embark on this exercise when it used the word “substantially”, it should have said so when the partial defence was revised, and the courts should not supply this meaning in the absence of any clear intention.
The suggestion so far is that the phrase “substantially impaired” should be interpreted in the context of the rest of the partial defence, and that the amendments in 2009 point towards a less demanding interpretation. But there is a further, instrumental reason to prefer this course. To declare that the word “substantially” has some higher, unspecified meaning carries the danger that frequently no one else in the criminal justice system will feel able to apply its meaning. Even when the experts agree that there was “substantial impairment”, a prosecution for murder might proceed on the basis that the jury might apply a more stringent meaning to the phrase; and a judge might feel impelled to leave this possibility to the jury.
This situation might seem consistent with the jury being the ultimate fact-finder. But, subsequent to Golds, another constitution of the Court of Appeal in R. v Brennan [2014] EWCA Crim 2387 sought to limit the ability of the jury to override the experts. Brennan had also perpetrated a brutal killing, in his case as a result of schizotypal disorder. Again, all the experts testified for the defendant and one of them warned that, although the killing was planned, “core rationality is still retained by people with severe disorders”. However, the trial judge directed the jury that they did not have to “buy into” the experts’ opinions. The jury duly convicted Brennan of murder. Brennan complained about the judge's direction and the Court of Appeal quashed the conviction. The court took note of the recent statutory amendments to the partial defence. It found that “most, if not all of the aspects of the new provisions relate entirely to psychiatric matters” and thought that it was appropriate for expert psychiatrists in some cases to offer an opinion even on the “ultimate issue” (at [51]). It concluded that “Where there simply is no rational or proper basis for departing from uncontradicted and unchallenged expert evidence, then juries may not do so” (at [44]).
The cases of Golds and Brennan sit uneasily together. To allow a jury to apply its own meaning to the word “substantially”, per Golds, would often allow that jury indirectly to go behind medical evidence supporting the partial defence, as discouraged in Brennan. Yet the cases of Golds and Brennan both suggest that the danger of a prosecution of murder being brought – and successfully – on such a basis seems very real. Golds was arguably even on the borderline of legal insanity. It is hard to see how a jury could properly have found that his ability to form a rational judgment was not “substantially impaired”, whatever degree of impairment it might have thought to be appropriate.
The Supreme Court is due to hear a further appeal in Golds. It is submitted that it should allow the appeal and prefer the less demanding interpretation of “substantially”. It should also take the opportunity to approve Brennan. The effect should be that fewer prosecutions for murder will be brought in the teeth of the evidence supporting diminished responsibility; and jury verdicts should only be required where the experts differ, or where there are rational reasons to find their conclusions to be in error, or to find that the killing would likely have occurred in any event.
