Michaelmas Term [2016] UKSC 61 On appeal from: [2014] EWCA Crim 748

R v Golds (Appellant)
Lord Neuberger, President
Lady Hale, Deputy President
Lord Kerr
Lord Reed
Lord Hughes
Lord Toulson
Lord Thomas
30 November 2016
Heard on 14 June 2016
Appellant Respondent
David Etherington QC David Perry QC
Stephen Rose Tom Little
(Instructed by Taylor
Haldane & Barlex
(Instructed by Crown
Prosecution Service
Appeals and Review Unit
Page 2
LORD HUGHES: (with whom Lord Neuberger, Lady Hale, Lord Kerr, Lord
Reed, Lord Toulson and Lord Thomas agree)
1. The appellant Mark Golds was convicted by a jury of the murder of his
partner. He had admitted in court that he had killed her, and the sole issue at his trial
had been whether he had made out the partial defence of diminished responsibility,
and so fell to be convicted of manslaughter rather than of murder. The law to be
applied was section 2 of the Homicide Act 1957 after its recent revision by the
Coroners and Justice Act 2009. The issue is the correct approach to the statutory test
of whether his abilities were in specified respects “substantially impaired”: see
section 2(1)(b).
2. The appellant had attacked his partner with a knife at their home in front of
her young children after a running argument which had taken place on and off
throughout much of the day. He had inflicted some 22 knife wounds together with
blunt impact internal injuries. He had a history of mental disorder leading to
outpatient treatment and medication. Two consultant forensic psychiatrists gave
evidence that there was an abnormality of mental functioning arising from a
recognised medical condition, although they disagreed what that condition was.
There was no contradictory psychiatric evidence. The judge correctly identified the
questions which the jury needed to address (see para 8 below) and helpfully provided
a written summary of the ingredients of diminished responsibility. He also provided
a crystal clear written “route to verdict” document. On the issue of substantial
impairment of ability he told the jury:
“Mr Rose [counsel for the defence] did suggest to you in his
closing address that you would get some further help from me
when giving you directions in law as to what the word
substantially means, where it says substantially impaired his
ability to exercise those qualities. I am not going to give you
any help on the meaning of the word substantially, because
unless it creates real difficulty and you require further
elucidation, the general principle of English law is that where
an everyday word is used, don’t tell juries what it means. They
are bright enough and sensible enough to work it out for
themselves, so I am not going to paraphrase substantially.
Substantially is the word that is in the Act of Parliament and
that’s the word that you have to work with. If it becomes a
stumbling block in some way, well at the end of the day, you
can send me a note and in those circumstances, I am permitted
Page 3
to offer you a little more help, but not at this stage of
The jury did not ask for further help.
3. In the Court of Appeal (Criminal Division), amongst other grounds of appeal
which have not survived, the appellant contended (a) that the judge had been wrong
not to direct the jury as to what “substantially impaired” meant and (b) that the jury
might in the absence of such direction have applied a more stringent test than it
ought to have done. It was contended on his behalf that so long as the impairment
was more than merely trivial, the test of “substantially impaired” was met.
4. The Court of Appeal dismissed the appellant’s appeal ([2015] 1 WLR 1030)
but certified in relation to this ground that the following two questions of law of
general public importance were involved:
1. Where a defendant, being tried for murder, seeks to establish that he
is not guilty of murder by reason of diminished responsibility, is the Court
required to direct the jury as to the definition of the word “substantial” as in
the phrase “substantially impaired” found in section 2(1)(b) of the Homicide
Act 1957 as amended by section 52 of the Coroners and Justice Act 2009?
2. If the answer to the first question is in the affirmative, or if for some
other reason the judge chooses to direct the jury on the meaning of the word
“substantial”, is it to be defined as “something more than merely trivial”, or
alternatively in a way that connotes more than this, such as “something whilst
short of total impairment that is nevertheless significant and appreciable”?
The Court of Appeal’s answers to these questions were (1) that the judge was not,
on authority, required to give greater definition than he did and (2) that if he had
done so the appropriate formulation would have been that it was not enough that
there was some impairment; the jury had to ask if it was substantial. It would, the
court held, be wrong to direct the jury that it sufficed that the impairment was more
than merely trivial.
The statute
5. As now amended, section 2 Homicide Act 1957 provides a complete
definition of diminished responsibility. The material parts of it are as follows:
Page 4
“Persons suffering from diminished responsibility
2(1) A person (‘D’) who kills or is a party to the killing of
another is not to be convicted of murder if 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 –
(a) to understand the nature of D’s conduct;
(b) to form a rational judgment;
(c) to exercise self control.
(1B) For the purposes of subsection (1)(c), an abnormality of
mental functioning provides an explanation for D’s conduct if
it causes, or is a significant contributory factor in causing, D to
carry out that conduct.
(2) On a charge of murder, it shall be for the defence to
prove that the person charged is by virtue of this section not
liable to be convicted of murder.
(3) A person who but for this section would be liable,
whether as principal or as accessory, to be convicted of murder
shall be liable instead to be convicted of manslaughter.”
6. This differs from the previous formulation of the partial defence. As
originally enacted, section 2(1) provided:
Page 5
“(1) 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 (whether arising from a
condition of arrested or retarded development of mind or any
inherent causes or induced by disease or injury) as
substantially impaired his mental responsibility for his acts and
omissions in doing or being a party to the killing.”
7. It follows that the expression “substantially impaired” has been carried
forward from the old Act into its new form. But whereas previously it governed a
single question of “mental responsibility”, now it governs the ability to do one or
more of three specific things, to understand the nature of one’s acts, to form a
rational judgment and to exercise self-control. Those abilities were frequently the
focus of trials before the re-formulation of the law. But previously, the question for
the jury as to “mental responsibility” was a global one, partly a matter of capacity
and partly a matter of moral culpability, both including, additionally, consideration
of the extent of any causal link between the condition and the killing. Now, although
there is a single verdict, the process is more explicitly structured. The jury needs to
address successive specific questions about (1) impairment of particular abilities and
(2) cause of behaviour in killing. Both are of course relevant to moral culpability,
but the jury is not left the same general “mental responsibility” question that
previously it was. The word used to describe the level of impairment is, however,
the same.
8. The effect of the new statutory formulation is that the following four
questions will normally arise in a case where diminished responsibility is advanced.
(1) Did the accused suffer from an abnormality of mental functioning?
(2) If so, did it arise from a recognised medical condition?
(3) If yes to (1) and (2), did it substantially impair one or more of the
abilities listed in section 1A?
(4) If yes to (1), (2) and (3), did it cause or significantly contribute to his
killing the deceased?
Of course, in some cases one or more of these may be common ground. The function
of the judge is to focus the jury’s attention on what is at issue and to explain why
the issue(s) are relevant, as the judge did in the present case. It is not to read the jury
a general statement of the law.
Page 6
Authority: “substantially impaired”
9. The concept of diminished responsibility was developed (with, at first,
varying terminology) by the common law in Scotland in the late 19th and early 20th
centuries as a means of mitigating, in an appropriate case, the mandatory sentence
of death attendant on murder: see Lord Justice-General Rodger’s helpful historical
survey in Galbraith v HM Advocate 2002 JC 1 (paras 23 to 27), together with the
report of the Scottish Law Commission SLC 195 (2004) at para 3.1. It operates by
reducing the offence of murder to that of culpable homicide. It was adopted by
English law via the Homicide Act 1957 for the same reason, and using the same
mechanism of partial defence, at a time when the abolition of capital punishment
was under debate but there was no Parliamentary majority for that greater step.
10. Soon after its introduction, the new partial defence was considered by the
Court of Criminal Appeal in R v Matheson [1958] 1 WLR 474, R v Spriggs [1958]
1 QB 270 and R v Byrne [1960] 2 QB 396. In the first case there was no occasion
for discussion of the meaning of “substantially impaired”; the defendant was agreed
to be certifiable. In Spriggs, however, the court considered the then conventional
formulations employed in Scotland in relation to the level of impairment, which
included (but were not confined to) references to the borderline of insanity (see HM
Advocate v Savage 1923 JC 49). The court (Lord Goddard CJ, Hilbery and Salmon
JJ) concluded that the correct course for the trial judge was not to attempt synonyms
or re-definition but simply to direct the jury in the terms of section 2.
11. In Byrne the defendant was a sexual psychopath who had strangled and
mutilated a young woman resident of the YWCA. The case on his behalf was that
he was unable to resist his impulse to gross and sadistic sexual violence. The judge’s
directions had amounted to excluding from abnormality of mind an inability to
control his urges, and this was held to have been wrong. The court further took the
view that on the medical evidence the defendant was so disturbed that there was no
room for doubt that diminished responsibility was made out. Giving the judgment
of the court, however, Lord Parker CJ addressed the question of substantial
impairment. He said this at 403-404:
“Assuming that the jury are satisfied on the balance of
probabilities that the accused was suffering from ‘abnormality
of mind’ from one of the causes specified in the parenthesis of
the subsection, the crucial question nevertheless arises: was the
abnormality such as substantially impaired his mental
responsibility for his acts in doing or being a party to the
killing? This is a question of degree and essentially one for the
jury. Medical evidence is, of course, relevant, but the question
involves a decision not merely as to whether there was some
Page 7
impairment of the mental responsibility of the accused for his
acts but whether such impairment can properly be called
‘substantial’, a matter upon which juries may quite legitimately
differ from doctors.

This court has repeatedly approved directions to the jury which
have followed directions given in Scots cases where the
doctrine of diminished responsibility forms part of the common
law. We need not repeat them. They are quoted in Reg v
Spriggs. They indicate that such abnormality as ‘substantially
impairs his mental responsibility’ involves a mental state which
in popular language (not that of the M’Naughten Rules) a jury
would regard as amounting to partial insanity or being on the
border-line of insanity.”
12. Both in England and in Scotland it has subsequently been held that it is not
usually helpful to direct juries in terms of the borderline of insanity. That is
demonstrated by considering the case where the mental impairment is depression,
to which (however severe) such a description is inapt. Such a formulation was later
disapproved in R v Seers (1984) 79 Cr App R 261 (a depression case) and is now
more often and wisely avoided even in a case of florid psychosis. Despite its use in
Byrne, it cannot have been the intention of the court in that case to require any such
direction, given the approval of Spriggs which had commended abstention from
elaboration of the words of the section. Giving the judgment in Seers Griffiths LJ
reached the same conclusion. At 264 he said this:
“It is to be remembered that in Byrne … all the doctors agreed
that Byrne could be described as partially insane; he was a
sexual psychopath who had hideously mutilated a young
woman he had killed. In such a case the evidence justifies
inviting a jury to determine the degree of impairment of mental
responsibility by a test of partial insanity. But it is not a
legitimate method of construing an Act of Parliament to
substitute for the words of the Act an entirely different phrase
and to say that it is to apply in all circumstances. We are sure
that this was not the intention of the court in Byrne …, and the
phrase was used as one way of assisting the jury to determine
the degree of impairment of mental responsibility in an
appropriate case, and no doubt to point out that Parliament by
the use of the word ‘substantial’ was indicating a serious degree
of impairment of mental responsibility.”
Page 8
But what is clear is that whilst the question whether the impairment was or was not
substantial was to be left to the jury in the unimproved words of the statute, the
underlying assumption was that “substantially” in this context meant impairment
which was of some importance or, as it was put in Seers, a serious degree of
impairment. The court cannot have contemplated in any of these cases that it was
sufficient that the impairment merely passed triviality.
13. R v Simcox The Times 25 February 1964; [1964] Crim LR 402 concerned a
man who had previously murdered his second wife and had now sought out his third
wife, with whom he was in dispute, taking with him a rifle with which he shot her
sister when it was her whom he encountered. Some four psychiatrists agreed that he
had an abnormality of mind, namely a paranoid personality. Each said that it
impaired his self-control, but none was prepared to say that the impairment was
substantial; they spoke of “moderate” impairment, or of his finding it “harder” than
others to control himself. The judge left the question to the jury in the terms of the
section, adding only that they should ask:
“do we think, looking at it broadly as commonsense people,
there was a substantial impairment of his mental responsibility
in what he did? If the answer to that is ‘yes’ then you find him
not guilty of murder but guilty of manslaughter. If the answer
to that is ‘no, there may be some impairment but we do not
think it was substantial. We do not think it was something
which really made any great difference although it may have
made it harder to control himself to refrain from crime’, then
you would find him guilty as charged.”
The Court of Appeal, whilst observing that the final sentence needed the previous
focus on the word “substantial” in order that it should not be thought that the absence
of self-control had to be total, approved this direction. It is to be seen that it was
essentially in accordance with Spriggs, since it repeated and emphasised, but did not
attempt to re-define, the statutory expression “substantially impaired”.
14. Three years later the Court of Criminal Appeal considered the case of R v
Lloyd [1967] 1 QB 175, which would appear to be the indirect origin of the
submission made in the present case that “substantially impaired” means any
impairment greater than the merely trivial. The defendant had killed his wife. There
was evidence that from time to time he had suffered recurrent episodes of reactive
depression. Two psychiatrists gave evidence that this was a mental abnormality
which to some extent impaired his mental responsibility. Neither was prepared to
say that the impairment was substantial. The first said that the depression impaired
his responsibility “to some extent”. The second said that there was some effect; he
could not say to what degree, but although it was not as low as minimal it was not
Page 9
substantial. The medical evidence was thus to similar effect as in Simcox. At trial,
Ashworth J had directed the jury in the terms of the statute, but he had then added:
“Fourthly, this word ‘substantial’, members of the jury. I am
not going to try to find a parallel for the word ‘substantial’. You
are the judges, but your own common sense will tell you what
it means. This far I will go. Substantial does not mean total, that
is to say, the mental responsibility need not be totally impaired,
so to speak destroyed altogether. At the other end of the scale
substantial does not mean trivial or minimal. It is something in
between and Parliament has left it to you and other juries to say
on the evidence, was the mental responsibility impaired, and, if
so, was it substantially impaired?” (p 178)
15. Counsel for the defendant, on appeal, contended that the judge had erred in
not directing the jury that “substantially” meant “really present” or “not trivial”.
That was a submission that it meant no more than that there was some operating
impairment, and thus that any such sufficed, so long as it was not trivial, and was
exactly the same submission which is now made in the present case. Since the
doctors had agreed that the depression was not trivial in its effect, the defendant was,
it was submitted, entitled to be acquitted of murder. That contention was firmly
rejected by the court. Edmund Davies J, giving the judgment of the court, said this
at 180B
“This court is wholly unable to accept that submission. The
word ‘substantially’ obviously is inserted in the Act with a
view to carrying some meaning. It does carry a meaning. This
court is quite unable to see that the direction given to the jury
on the meaning of this word, can validly be criticised, and finds
itself in a difficulty of saying that any distinction can be validly
drawn between the direction given in the instant case and that
approved of by this court in Reg v Simcox.”
16. It is the decision of the Court of Appeal which is the authority. But it is
equally clear that Ashworth J, in saying what he did, had no intention of telling the
jury that any impairment beyond the trivial sufficed. Firstly, if that had been his
intention, it would have followed that the evidence in the case satisfied the test and
a verdict of diminished responsibility ought to have followed unless the jury
disagreed; this the judge would surely have told the jury. Secondly, such an intention
is inconsistent with the judge telling the jury that he was not going to find a synonym
for the word “substantially”. Thirdly, the judge’s summing up makes clear that he
had before him Bryne, with its references to the borderline of insanity, although
(anticipating Seers) he sensibly did not adopt that expression in a case concerning
Page 10
depression. In referring to the spectrum of impairment as he did, he may have had
in mind the warning in Simcox (see para 13 above) that it should be made clear that
the impairment did not need to be total. What he was clearly saying was that before
an impairment could be substantial it must of course be greater than the merely
trivial, but that, beyond that, what amounted to substantial impairment was a matter
of degree for the jury.
17. Over the years since, a reference of this kind to the extremities of possible
impairment has sometimes been thought not simply to be helpful to juries but also
to provide a possible definition of the meaning of “substantially”. R v Egan [1992]
4 All ER 470 concerned the case where there is both abnormality of mind and
voluntary intoxication. Its principal decision largely anticipated the test for such a
case which was later adumbrated by the House of Lords in R v Dietschmann [2003]
UKHL 10; [2003] 1 AC 1209, but the court was held by the House to have erred in
its treatment of other prior decisions. No real issue arose in relation to the meaning
of “substantially impaired” except as to how drink was to be accommodated within
it. But one of those prior decisions on drink, R v Gittens [1984] QB 698, 703, had
contained the conclusion of Lord Lane CJ that the jury should ignore the effect of
drink, as later held to be the law in Dietschmann. Lord Lane had pointed out that
voluntary intoxication could not constitute a mental abnormality arising from
disease or inherent cause, so the jury should ignore it and then go on to
“consider whether the combined effect of the other matters
which do fall within the section amounted to such abnormality
of mind as substantially impaired the defendant’s mental
responsibility within the meaning of ‘substantial’ set out in R v
In Egan, having cited that passage, Watkins LJ added in passing:
“In R v Lloyd … directions as to the word ‘substantial’, to the
effect that (1) the jury should approach the word in a broad
commonsense way or (2) the word meant ‘more than some
trivial degree of impairment which does not make any
appreciable difference to a person’s ability to control himself,
but it means less than total impairment’ were both approved.”
There was no occasion for analysis of Lloyd in Egan. But although it was correct
that Ashworth J’s direction had been approved, it would be quite inaccurate to imply
that the effect of the case was that “substantially” meant the same as “more than
some trivial degree of impairment”. It may well be that Watkins LJ meant to say no
such thing, rather than simply to refer to Ashworth J’s formulation as convenient,
Page 11
but if he did, it was a misreading of Lloyd. The decision in Lloyd, to which no doubt
Lord Lane CJ was referring in Gittens, was precisely the opposite, viz: that
“substantially” was not the same as “more than trivial” – see para 13 above.
18. The difficulty for later readers was compounded by the closing words of the
judgment in Egan at 480h:
“Finally, for the avoidance of doubt, we advise judges that
guidance as to the meaning of ‘substantial’ should be explicitly
provided for the jury by using one or other of the two meanings
in R v Lloyd.”
This proposition that Lloyd authorised two meanings of “substantially” may have
achieved some currency since. If it has, it too is based on a misunderstanding. The
most that Lloyd ever said was that two methods of summing up were unexceptional:
the first to tell the jury simply to use its common sense without further elaboration
and the second to allude to the spectrum between just beyond trivial impairment and
total impairment. The decision of the court was explicitly that impairment beyond
“more than merely trivial” is required; it follows that if the second approach,
referring to the spectrum, is adopted in summing up, this must be made clear. But
the court in Lloyd was not attempting in its (extempore) judgment to ordain a
template for future summings-up. It was dealing with the submission that the
defendant in that case was entitled to have his conviction for murder set aside
because any impairment beyond the merely trivial sufficed, and this submission it
rejected. All that mattered in that case, as in most cases before an appellate criminal
court, was whether the judge had misdirected the jury to the disadvantage of the
19. With or without any implication of two meanings, Ashworth J’s additional
“spectrum” illustration has gained currency. It has figured in successive Crown
Court Benchbooks. For example, the first (2010) edition, published before the new
statutory formula came into operation, carefully avoided dictating the terms of
summing up to judges. However, it cited at p 340 what Ashworth J had said, and
added that the direction was approved by the Court of Criminal Appeal. A little later
it gave one illustration of the kind of summing up which might be employed. It did
so in the context of the more difficult case where diminished responsibility is
complicated by drink and/or by alcohol dependence, but the example was equally
relevant also to non-alcohol cases. One suggested form of words (at p 347) was:
“This requires you to consider to what extent the defendant’s
state of mind differed from that of the ordinary person. Was it
so abnormal that the defendant’s mental responsibility was
Page 12
substantially reduced? ‘Substantially’ is an ordinary English
word to which you will bring your own experience. It means
less than total and more than trivial. Where you draw the line
is for your own good judgment.”
Subsequent editions, before and after the 2009 Act amendments, contained similar
passages until the decision of the Court of Appeal in the present case. It will be seen
that this formulation does not tell the jury that any impairment beyond the merely
trivial suffices, but with hindsight it is possible that if one does not go back to the
decision in Lloyd, it might be taken by some to carry that implication.
20. In R v Ramchurn [2010] EWCA Crim 194; [2010] 2 Cr App R 18 (an
unamended 1957 Act case) the trial judge had understandably adopted these
suggestions. His written direction to the jury was:
“‘Substantially impaired’ means just that. You must conclude
that his abnormality of mind was a real cause of the defendant’s
conduct. The defendant need not prove that his condition was
the sole cause of it, but he must show that it was more than a
merely trivial one which did not make any real or appreciable
difference to his ability to control himself.”
In retirement, the jury asked a specific question: what was the difference between
“trivial” and “substantial”? The judge responded with the Ashworth formula. He
told them:
“The following direction has been approved at a senior level
and it is this; the direction on the words ‘substantially
impaired’. Your own common sense will tell you what it
means. ‘Substantial’ does not mean ‘total’. That is to say the
mental responsibility need not be totally impaired, so to speak,
destroyed altogether. The other end of the scale, ‘substantial’
does not mean ‘trivial’ or ‘minimal’. It is something in between
and Parliament has left it to you to say on the evidence was the
mental responsibility impaired and if so, was it substantially
21. The defendant in Ramchurn had planned and executed the killing of his
wife’s lover, a cousin to whom he had originally given a home. He had threatened
previously that he would kill him, and had made a number of preparations to do so,
such as trying to get keys to gain access to the victim’s home, and when that failed
Page 13
arranging a meeting to carry out his plan, equipping himself with a rope ligature for
the purpose. He disposed of the body some distance away and set up a false alibi.
The evidence was that he was depressed. One doctor described his state as “an
emotional turmoil” and a “tortured frame of mind”, and expressed the opinion that
“in the tumultuous final moments which resulted in the death” the impairment of
mental responsibility would have been substantial. The other agreed that there was
an element of depression, and accepted that it had played some part in the killing.
Carefully cross-examined, he agreed that the impact of the depression on the
defendant’s mental responsibility was more than trivial, but he disagreed that it was
substantial. The jury convicted of murder.
22. The argument for the defendant on appeal in Ramchurn was that there were
two inconsistent meanings of “substantially” to be derived from Lloyd, that the judge
had in consequence failed to give the jury a clear direction and moreover that the
law was in too uncertain a state to satisfy the requirements of article 7 of the ECHR.
Accordingly, it was contended, the conviction for murder was unsafe. The Court of
Appeal rejected those arguments. At para 23, Lord Judge CJ addressed specifically
the “two meanings” argument, founded then as now on a combination of Lloyd with
Egan. The argument was rejected:
“It is, however, clear on analysis that in Lloyd the court rejected
the submission that there were two meanings for the word
‘substantially’. In the judgment in Lloyd the word
‘substantially’ carried ‘some’ meaning or ‘a’ meaning. It was
accepted in Lloyd that there were different ways of illustrating
the same concept and, if necessary, explaining its relevance to
the jury. If the court in Egan had intended to convey that the
words ‘substantially impaired’ embraced two different
concepts or levels of impairment, it would have said so not by
citing Lloyd as authority in support, but by distinguishing
Lloyd. In the result, just as the court in Lloyd could see no
effective difference between the directions in Simcox and
Lloyd, the Court of Appeal in Egan could see no difficulty in
the deployment of either of the two methods of explanation
found in Lloyd.”
23. The court recorded that section 2 had been in force for 50 years and applied
in countless murder trials, and observed that in its experience the test of substantial
impairment was probably, in practice, the least difficult aspect of what can be a
difficult defence to convey to a jury. It went on specifically to endorse the general
starting point that the test was in ordinary English and should be left to the judgment
of the jury. In so doing, it said this at para 15:
Page 14
“‘Substantially’ is an ordinary English word which appears in
the context of a statutory provision creating a special defence
which, to reflect reduced mental responsibility for what
otherwise would be murderous actions, reduces the crime from
murder to manslaughter. Its presence in the statute is deliberate.
It is designed to ensure that the murderous activity of a
defendant should not result in a conviction for manslaughter
rather than murder on account of any impairment of mental
responsibility, however trivial and insignificant; but equally
that the defence should be available without the defendant
having to show that his mental responsibility for his actions
was so grossly impaired as to be extinguished. That is the
purpose of this defence and this language. The Concise Oxford
Dictionary offers ‘of real importance’ and ‘having substance’
as suggested meanings for ‘substantially’. But, in reality, even
the Concise Oxford Dictionary tells us very little more about
the ordinary meaning and understanding to be attached to the
word ‘substantially’. The jury must decide for itself whether
the defendant’s mental responsibility for his actions was
impaired and, assuming that they find that it was, whether the
impairment was substantial.”
24. Thus the appeal failed in Ramchurn. The court was plainly not adopting the
submission that “substantially” means any impairment beyond the merely trivial, for
if it had done so, the evidence of both psychiatrists would have met the test. It is
right to remember that the focus on the meaning of that word in the half dozen cases
here reviewed, and in the present case, does not mean that it is often the occasion of
difficulty. But the fact that the present submission is now made for the third time,
despite its failure in both Lloyd and Ramchurn, does demonstrate that the use of the
Ashworth “spectrum” formula may encourage semantic debate, at least in some
cases. Moreover it is known that in at least one case which reached the Court of
Appeal on sentence, the trial judge had directed the jury that the test of “substantially
impaired” was met by an impairment which was more than minimal: R v Brown
(Robert) [2011] EWCA Crim 2796; [2012] 2 Cr App R(S) 156. Since the appeal
was limited to sentence in that case, the correctness of that direction did not call for
adjudication. But that case is a further illustration of difficulty. When the defendant
was, on that direction, convicted of manslaughter, the judge concluded when it came
to sentence that in fact his responsibility had nevertheless been “substantial”, and
the Court of Appeal decided that he was indeed entitled so to do, and to impose a
very long determinate sentence (24 years) in consequence.
Page 15
25. The rejection in the foregoing cases of the contention that any impairment
beyond the merely trivial will suffice is consistent with the way in which the law of
diminished responsibility has evolved in Scotland, where it originated. The law was
reviewed in some depth by a specially convened court of five in Galbraith v HM
Advocate 2002 JC 1. The court held that the partial defence was not confined to
mental illness, strictly so called, and that other mental abnormalities might also be
capable of diminishing the responsibility of the accused, including in that case a
combination of learned helplessness and post-traumatic stress disorder following
alleged persistent abuse. The decision anticipated the new English section 2(1)(a)
by requiring that there be “some recognised mental abnormality” (paras 53 and 54).
As to the level of impairment, the court held, for reasons essentially the same as had
been given by the English court in Seers, that previous references to the borderline
of insanity were simply examples of what would plainly qualify rather than a test
for inclusion. Lord Justice-General Rodger summarised the rule in this way at para
“In every case, in colloquial terms, there must, unfortunately,
have been something far wrong with the accused, which
affected the way he acted … While the plea of diminished
responsibility will be available only where the accused’s
abnormality of mind had substantial effects in relation to his
act, there is no requirement that his state of mind should have
bordered on insanity.

In essence, the jury should be told that they must be satisfied
that, by reason of the abnormality of mind in question, the
ability of the accused, as compared with a normal person, to
determine or control his actings was substantially impaired.”
Thus “substantially impaired” was adopted as the test, and used in the sense of
something “far wrong” with the accused.
26. There was, then, one difference between Scottish and English law, because
in Scotland Galbraith held that psychopathic personality disorder was not capable
of being a basis for diminished responsibility in the same way as in both jurisdictions
voluntary intoxication cannot by itself found the plea: see Galbraith at para 54 and,
in England, R v Dowds [2012] EWCA (Crim) 281; [2012] 1 WLR 2576. Now,
Page 16
however, that distinction has gone. Following scrutiny by the Scottish Law
Commission the law has been put into statutory form by section 51B of the Criminal
Procedure (Scotland) Act 1995, inserted by section 168 of the Criminal Justice and
Licensing (Scotland) Act 2010, (asp 13). Provision is made by subsection (3) to
exclude voluntary intoxication but, on the Commission’s recommendation, not for
a similar exclusion for psychopathic personality disorder. The new Scottish
definition of diminished responsibility in subsection (1) provides:
“A person who would otherwise be convicted of murder is
instead to be convicted of culpable homicide on grounds of
diminished responsibility if the person’s ability to determine or
control conduct for which the person would otherwise be
convicted of murder was, at the time of the conduct,
substantially impaired by reason of abnormality of mind.”
Thus, the Scottish law now expresses, like the English, the essential feature of
abnormality of mind such as impairs the ability to determine or control conduct, and,
like English law, adopts as the test for the level of impairment the same expression,
namely “substantially”. Plainly in Scotland this expression was used in the
knowledge of the meaning authoritatively given to it by Galbraith, which the
Scottish Law Commission had endorsed: SLC 195, July 2004, paras 3.15-3.17.
Usage of language
27. The admirably concise submissions of Mr Etherington QC for the appellant
correctly point out that as a matter simply of dictionary definition, “substantial” is
capable of meaning either (1) “present rather than illusory or fanciful, thus having
some substance” or (2) “important or weighty”, as in “a substantial meal” or “a
substantial salary”. The first meaning could fairly be paraphrased as “having any
effect more than the merely trivial”, whereas the second meaning cannot. It is also
clear that either sense may be used in law making. In the context of disability
discrimination, the Equality Act 2010 defines disability in section 6 as an
impairment which has a substantial and long-term effect on day to day activities,
and by the interpretation section, section 212, provides that “‘Substantial’ means
more than minor or trivial.” It thus uses the word in the first sense. Conversely, the
expression “significant and substantial” when used to identify which breaches by
the police of the Codes of Practice under the Police and Criminal Evidence Act 1984
will lead to the exclusion of evidence (see for example R v Absolam (1988) 88 Cr
App R 332 and R v Keenan [1990] 2 QB 54) is undoubtedly used in the second
sense. It is to be accepted that the word may take its meaning from its context. It is
not surprising that in the context of triggering a duty to make reasonable adjustments
to assist the disabled, the first sense should be used by the Equality Act; the extent
of adjustments required varies with the level of disability and a wide spectrum of
Page 17
both is to be expected. Mr Etherington additionally submits that this usage shows
that the first sense does not entirely strip the word “substantially” of meaning.
Conclusions: “substantially”
28. The foregoing review of the authorities clearly shows that in the context of
diminished responsibility the expression “substantially” has always been held, when
the issue has been confronted, to be used in the second of the senses identified above.
29. True it is that in Lloyd Edmund-Davies J observed that that word had been
put into the 1957 Homicide Act with a view to it carrying some meaning. If by that
he meant that it could have no purpose at all unless it was used in the second sense
above, the Equality Act usage may suggest otherwise, although even without the
word “substantially” it is perhaps open to doubt that a merely trivial effect would be
taken to be included either in “impairment” or in “disability”. But this does not alter
the central thrust of the decision in Lloyd, which was that in the context of
diminished responsibility an impairment of consequence or weight is what is
required to reduce murder to manslaughter, and not any impairment which is greater
than merely trivial.
30. There is no basis for thinking that when the same expression was carried
forward into the new formulation of diminished responsibility any change of sense
was intended. The adverb “substantially” is applied now, as before, to the verb
“impaired”. In the absence of any indication to the contrary, Parliament is to be taken
to have adopted the established sense in which this word has been used for 50 years.
31. The reformulation of the law followed the recommendation of the Law
Commission, except to the irrelevant extent that it did not incorporate
“developmental immaturity” as an extension beyond recognised medical conditions.
The Commission had addressed diminished responsibility in two reports, each
preceded by a detailed consultation paper: Partial Defences to Murder Law Com
290 (2004) and Murder, Manslaughter and Infanticide Law Com 304 (2006). Prior
to the earlier report, it had consulted upon a number of possible formulations of the
test for diminished responsibility – see Partial Defences at 5.52 et seq. Most
employed the adverb “substantially”.
32. The Commission was concerned to ensure that a requirement for causation
was explicitly incorporated into the proposed statutory test, as it now has been, and
had consulted on the question whether this test would suffice without any threshold
of substantial impairment – see possible version (6) at 5.52. It is no doubt true that
in many cases the question whether the impairment is sufficient to establish the
Page 18
partial defence will march alongside the question whether it was “a significant
contributory factor” in causing the killing. But this will not always be so. Where, for
example, the recognised medical condition is an emotionally unstable personality
disorder leading to histrionic and impulsive behaviour, or where it is depression
leading to distorted thinking, the medical evidence may make it clear that it has had
some impact on behaviour and thus was a significant cause. The jury may be
satisfied that if the defendant’s personality had been different, or if there had not
been some depression, he would not have killed as he did. The real question thus
may very well be whether the condition passes the threshold of substantial
impairment, or does not.
33. An illustration is afforded by the facts of R v Brown. The defendant’s
marriage had broken down. He was living elsewhere with his girlfriend. There were
acrimonious negotiations over the division of property between himself and his wife.
He felt that she was dishonestly concealing her assets and cheating him, and that she
had unfairly manipulated him into signing what he saw as a disadvantageous prenuptial agreement. He planned to kill her. He prepared a grave in Windsor Great
Park and, when returning the children to her after a weekend, took with him a
hammer hidden in his daughter’s bag and beat her to death, before dismantling the
CCTV equipment which would have recorded his movements, and disposing of the
body in the grave. There was psychiatric evidence that he had developed an
adjustment disorder, a recognised medical condition, arising from the severe stress
of life events. The jury must have accepted the diagnosis, and that the adjustment
disorder was a significant cause of his killing his wife. On the judge’s direction, that
impairment beyond the merely trivial sufficed, the conviction for manslaughter
followed. Whether or not the jury would have concluded, but for that direction, that
the impairment was substantial, can never be known. But it is clear that such a
conclusion would not follow necessarily from the finding of significant causation.
34. After consultation, the Commission’s final conclusion, in the second report
at 1.17, was that although there were some infelicities in the wording it was not
persuaded that any of the alternative formulations canvassed would sufficiently
improve the law to justify interfering with a workable form of words. It had pointed
out in the earlier report at 7.91 that the approach to the concept was essentially
pragmatic, that the leading authority remained Byrne and that this partial defence
had, unlike provocation, troubled the House of Lords only once in 50 years. The
formula now incorporated into the statute was recommended. The specific
requirement for causation was added, but the threshold of substantial impairment
was maintained.
35. It follows that there is nothing in the change of the formulation of the test for
diminished responsibility to cause a different view to be taken now of the sense in
which the word “substantially” is used in conjunction with “impairment”.
Page 19
36. This use of the expression accords with principle. Diminished responsibility
effects a radical alteration in the offence of which a defendant is convicted. The
context is a homicide. By definition, before any question of diminished
responsibility can arise, the homicide must have been done with murderous intent,
to kill or to do grievous bodily harm, and without either provocation or self-defence.
Whilst it is true that at one end of the scale of responsibility the sentence in a case
of diminished responsibility may be severe, or indeed an indefinite life sentence
owing to the risk which the defendant presents to the public, the difference between
a conviction for murder and a conviction for manslaughter is of considerable
importance both for the public and for those connected with the deceased. It is just
that where a substantial impairment is demonstrated, the defendant is convicted of
the lesser offence and not of murder. But it is appropriate, as it always has been, for
the reduction to the lesser offence to be occasioned where there is a weighty reason
for it and not merely a reason which just passes the trivial.
Directing juries: good practice
37. As Mr Perry QC for the Crown rightly submitted, there are many examples
of ordinary English words incorporating questions of degree, which are left to juries
to apply without attempts at further definition. No-one attempts to define
“reasonable” in the many contexts in which it appears. Nor should there be any
further sophistication applied to the standard of proof required, that the jury be
“sure”, at least beyond the comparable expression “leaving no reasonable doubt”.
The same principle of leaving an ordinary word alone was applied by the House of
Lords in Brutus v Cozens [1973] AC 854 to the expression “insulting”, and would
apply equally, no doubt, to its sister expressions “abusive” and “threatening”. In all
these cases the understandable itch of the lawyer to re-define needs to be resisted.
Any attempt to find synonyms for such ordinary English expressions, although they
involve questions of degree, simply complicates the jury’s exercise, and leads to
further semantic debate about the boundaries of meaning of the synonym.
38. Where, however, as here, there are two identifiable and different senses in
which the expression in question may be used, the potential for inconsistent usage
may need to be reduced. The existence of the two senses of the word “substantially”
identified above means that the law should, in relation to diminished responsibility,
be clear which sense is being employed. If it is not, there is, first, a risk of trials
being distracted into semantic arguments between the two. Secondly, there is a risk
that different juries may apply different senses. Thirdly, medical evidence (nearly
always forensic psychiatric evidence) has always been a practical necessity where
the issue is diminished responsibility. If anything, the 2009 changes to the law have
emphasised this necessity by tying the partial defence more clearly to a recognised
medical condition, although in practice this was always required. Although it is for
the jury, and not for the doctors, to determine whether the partial defence is made
out, and this important difference of function is well recognised by responsible
Page 20
forensic psychiatrists, it is inevitable that they may express an opinion as to whether
the impairment was or was not substantial, and if they do not do so in their reports,
as commonly many do, they may be asked about it in oral evidence. It is therefore
important that if they use the expression, they do so in the sense in which it is used
by the courts. If there is doubt about the sense in which they have used it, their
reports may be misunderstood and decisions made upon them falsified, and much
time at trials is likely to be taken up unnecessarily by cross examination on the
semantic question. The experience of R v Brown (supra at paras 24 and 33)
underlines the need for clarification.
39. The sense in which “substantially impaired” is used in relation to diminished
responsibility is, for the reasons set out above, the second of the two senses. It is not
synonymous with “anything more than merely trivial impairment”.
40. It does not follow that it is either necessary or wise to attempt a re-definition
of “substantially” for the jury. First, in many cases the debate here addressed will
simply not arise. There will be many cases where the suggested condition is such
that, if the defendant was affected by it at the time, the impairment could only be
substantial, and the issue is whether he was or was not so affected. Second, if the
occasion for elucidation does arise, the judge’s first task is to convey to the jury, by
whatever form of words suits the case before it, that the statute uses an ordinary
English word and that they must avoid substituting a different one for it. Third,
however, various phrases have been used in the cases to convey the sense in which
“substantially” is understood in this context. The words used by the Court of Appeal
in the second certified question in the present case (“significant and appreciable”)
are one way of putting it, providing that the word “appreciable” is treated not as
being synonymous with merely recognisable but rather with the connotation of
being considerable. Other phrases used have been “a serious degree of impairment”
(Seers), “not total impairment but substantial” (Ramchurn) or “something far
wrong” (Galbraith). These are acceptable ways of elucidating the sense of the
statutory requirement but it is neither necessary nor appropriate for this court to
mandate a particular form of words in substitution for the language used by
Parliament. The jury must understand that “substantially” involves a matter of
degree, and that it is for it to use the collective good sense of its members to say
whether the condition in the case it is trying reaches that level or not.
41. It seems likely that the Ashworth “spectrum” illustration will have been of
assistance to juries in some cases, for it helps to explain (a) that the impairment need
not be total to suffice and (b) that “substantially” is a question of degree. But, as the
experience of Lloyd, Ramchurn and the present case teaches, if it is to be used it
needs to be combined with making it clear that it is not the law that any impairment
beyond the merely trivial will suffice. The impairment must of course pass the
merely trivial to be considered, just as it need not reach the total, but whether, when
it has passed the trivial, it can properly be regarded as substantial, is a matter for the
Page 21
jury in the individual case, aided as it will be by the experts’ exposition of the kind
of impairment which the condition under consideration may have generated in the
accused. Unless the spectrum illustration has been used by someone in the case, it
is preferable for the judge not to introduce it. If it has been used, or if, on mature
consideration the judge considers that it may help the jury in the particular case on
trial, it needs to be coupled with a clear statement that it is not enough that the
impairment be merely more than trivial; it must be such as is judged by the jury to
be substantial. For the same reason, if an expert witness, or indeed counsel, should
introduce into the case the expression “more than merely trivial”, the same clear
statement should be made to assist the jury.
42. Once this usage is understood by all concerned with the trial, there ought to
be no occasion for the jury to be distracted by debate about the meaning of the word.
What matters is what kind of effect the medical condition was likely to have had on
the three relevant capacities of the accused. So long as the experts understand the
sense in which “substantially” is used in the statute (which should henceforth be
clear), and that the decision whether the threshold is met is for the jury rather than
for them, it is a matter for individual judgment whether they offer their own opinion
on whether the impairment will have been substantial or confine themselves to the
kind of practical effect it would have had. If they do the former, they will be
understood to be using the word in the second sense set out in para 27 above.
43. It follows that the questions certified by the Court of Appeal should be
answered as follows:
(1) Ordinarily in a murder trial where diminished responsibility is in issue
the judge need not direct the jury beyond the terms of the statute and should
not attempt to define the meaning of “substantially”. Experience has shown
that the issue of its correct interpretation is unlikely to arise in many cases.
The jury should normally be given to understand that the expression is an
ordinary English word, that it imports a question of degree, and that whether
in the case before it the impairment can properly be described as substantial
is for it to resolve.
(2) If, however, the jury has been introduced to the question of whether
any impairment beyond the merely trivial will suffice, or if it has been
introduced to the concept of a spectrum between the greater than trivial and
the total, the judge should explain that whilst the impairment must indeed
pass the merely trivial before it need be considered, it is not the law that any
impairment beyond the trivial will suffice. The judge should likewise make
this clear if a risk arises that the jury might misunderstand the import of the
expression; whether this risk arises or not is a judgment to be arrived at by
the trial judge who is charged with overseeing the dynamics of the trial.
Page 22
Diminished responsibility involves an impairment of one or more of the
abilities listed in the statute to an extent which the jury judges to be
substantial, and which it is satisfied significantly contributed to his
committing the offence. Illustrative expressions of the sense of the word may
be employed so long as the jury is given clearly to understand that no single
synonym is to be substituted for the statutory word: see para 40 above.
R v Brennan
44. Counsel drew attention to the Court of Appeal decision in R v Brennan [2014]
EWCA Crim 2387; [2015] 1 WLR 2060, decided after both trial and appeal in the
present case.
45. The defendant in that case (aged 22 at the time of the offence) had a nineyear history of disturbed childhood, sexual abuse and outpatient mental health
treatment together with one instance when he was sectioned following a suicide
attempt. On the undisputed psychiatric evidence he suffered from a schizotypal
disorder as well as an emotionally unstable personality disorder. He was obsessed
with witchcraft and Satanist killings. He was also depressed. He had planned and
executed the ritualistic killing of a client whom he had served as a male prostitute.
He left notes of what he planned to do, and after killing the man with one or more
knives, had scored his back and painted or written on the walls symbols such as a
pentagram and references to Satan and to Krishna, before cleaning himself up and
going to the police station to report what he had done. He was treated by the police
as needing an appropriate adult to attend his interviews, and told that person that he
had been having thoughts of killing somebody (apparently anybody) for several
weeks. At trial the only issue was diminished responsibility.
46. The Court of Appeal held that in that case there was only one possible
outcome. There was simply no basis for a verdict of murder and moreover this was
so clear that the judge ought not to have left it open to the jury. The court regarded
that decision as a straightforward application of R v Galbraith [1981] 1 WLR 1039;
73 Cr App R 124. It went on to offer some general observations about the
circumstances in which a judge ought to withdraw murder from the jury where the
issue is diminished responsibility and uncontradicted psychiatric evidence supports
the defence case on that topic.
47. The report suggests that Brennan was a case in which the Crown expressly
did not challenge the diagnosis of the single consultant psychiatrist called and barely
challenged her opinion that the defendant’s condition substantially impaired his
ability to form rational judgments. (There was perhaps greater challenge to the
opinion that his ability to control himself was also substantially impaired). That was
Page 23
a reasoned decision. The Crown had a second psychiatric report, disclosed in
ordinary course to the defence, which agreed those conclusions. Counsel for the
Crown had then, legitimately, tested the evidence of the psychiatrist, in particular
by drawing attention to the defendant’s consumption of drink and drugs, and to the
clear evidence of pre-planning. As to the first, the psychiatrist’s answer had,
however, been that the underlying mental condition effected sufficient impairment
independently of any additional disinhibition attributable to intoxication. As to the
second, she had said that a disordered and impaired mind may well be no less
capable of premeditation and detailed planning than a rational one, and that that was
what had happened. Those answers had not been challenged, presumably because
they were not, on the facts, capable of dispute.
48. It is an important part of the Crown’s function, where the charge is murder
and a case of diminished responsibility is advanced, to assess the expert evidence –
almost invariably obtained on both sides – and its relationship to any dispute of fact.
If it is clear that the defendant was indeed suffering from a recognised medical
condition which substantially impaired him in one of the material respects, and that
this condition was a significant cause of the killing, the Crown is entitled to, and
conventionally frequently does, accept that the correct verdict is guilty of
manslaughter on the grounds of diminished responsibility and no trial need ensue.
In practice quite a large proportion of verdicts of manslaughter on this ground arise
from the Crown taking this responsible course: see the research undertaken for the
Law Commission by Professor Mackay cited in Partial Defences to Murder Law
Com 290 (2004) at Appendix B, especially paras 6, 20 and 21. Acceptance of a plea
to manslaughter may properly be given either before trial, thus making it
unnecessary, or after testing the evidence if that is required.
49. Given the answers of the psychiatrist in Brennan and the state of the
evidence, it is clear that the Crown could not properly ask the jury to convict of
murder unless it was to reject one or more parts of the expert evidence. Certainly a
jury is not bound by the expert. In some cases, pre-planning, especially involving
meticulous preparations, may indicate self-control which gives grounds for rejecting
an opinion that self-control was substantially impaired. In others, there may be
legitimate grounds for asking the jury to disagree about the level of impairment. In
yet further cases, it may be perfectly proper to ask the jury to conclude that it was
the drink or drugs which led to the killing, whilst the underlying mental condition
was in the background. That is not by any means an exhaustive catalogue of
questions which a jury may properly be invited to decide. However, as the Court of
Appeal rightly held, if the jury is to be invited to reject the expert opinion, some
rational basis for doing so must at least be suggested, and none had been at trial nor
was on appeal. It is not open to the Crown in this kind of situation simply to invite
the jury to convict of murder without suggesting why the expert evidence ought not
to be accepted. In particular, it would not have been a proper basis for rejecting
diminished responsibility that the circumstances of the killing had been particularly
Page 24
violent or sadistic. It is a well-known factor in such cases that such brutality may
(understandably) be taken by a jury to point away from the partial defence;
sometimes it may truly do so, but not infrequently it is the product of the mental
50. It may be agreed that the ordinary principles of R v Galbraith are capable of
being applied in a trial where the sole issue is diminished responsibility. A court
ought, however, to be cautious about doing so, and for several reasons. First, a
murder trial is a particularly sensitive event. If the issue is diminished responsibility,
a killing with murderous intent must, ex hypothesi, have been carried out. If a trial
is contested, it is of considerable importance that the verdict be that of the jury.
Second, the onus of proof in relation to diminished responsibility lies on the
defendant, albeit on the balance of probabilities rather than to the ordinary criminal
standard. The Galbraith process is generally a conclusion that no jury, properly
directed, could be satisfied that the Crown has proved the relevant offence so that it
is sure. In the context of diminished responsibility, murder can only be withdrawn
from the jury if the judge is satisfied that no jury could fail to find that the defendant
has proved it. Thirdly, a finding of diminished responsibility is not a single-issue
matter; it requires the defendant to prove that the answer to each of the four questions
set out in para 8 above is “yes”. Whilst the effect of the changes in the law has
certainly been to emphasise the importance of medical evidence, causation (question
4) is essentially a jury question. So, for the reasons explained above, is question 3:
whether the impairment of relevant ability(ies) was substantial. That the judge may
entertain little doubt about what he thinks the right verdict ought to be is not
sufficient reason in this context, any more than in any other, for withdrawing from
the jury issues which are properly theirs to decide.
51. Where, however, in a diminished responsibility trial the medical evidence
supports the plea and is uncontradicted, the judge needs to ensure that the Crown
explains the basis on which it is inviting the jury to reject that evidence. He needs to
ensure that the basis advanced is one which the jury can properly adopt. If the facts
of the case give rise to it, he needs to warn the jury that brutal killings may be the
product of disordered minds and that planning, whilst it may be relevant to selfcontrol, may well be consistent with disordered thinking. While he needs to make it
clear to the jury that, if there is a proper basis for rejecting the expert evidence, the
decision is theirs – that trial is by jury and not by expert – it will also ordinarily be
wise to advise the jury against attempting to make themselves amateur psychiatrists,
and that if there is undisputed expert evidence the jury will probably wish to accept
it, unless there is some identified reason for not doing so. To this extent, the
approach of the court in Brennan is to be endorsed.
Page 25
The present case
52. In the present case the appellant and the deceased had lived together for
around three years before she was killed on a Sunday in July 2012. On that day she
and he, and her two sons aged 13 and eight, had been to a family barbecue. The
couple had rowed at the party, in part because she said that he had hit her in the past,
in part because he demanded that she give him a bank card which she refused to do,
and in part because he wanted to go home and she did not. After they had returned
home, separately, and after her mother had visited the house, the argument was
renewed later in the evening. Outside the house, the appellant seized the deceased
by her face, held her by her hair and slapped her across the cheek. She insisted that
he leave the home. He packed a bag but refused to leave. Some time later that
evening he attacked her. By then the deceased had a large lump on her face. The
several stages of this attack were witnessed by one or both of her two sons. The older
son intervened in the argument. He stood between them and said that he would not
leave them alone. The appellant then fetched a knife from the kitchen, but the older
son took it from his pocket. The boy told his mother about the knife and the appellant
said “It’s self defence”. She went and sat on the bed but the appellant went after her
and punched her in the head, whereupon she hit him back. He had a small cut on his
eyebrow which the boys said he squeezed to increase the blood flow. Then he
attacked the deceased with a second knife which he produced, kneeling on her arms
as he did so and shouting that he was going to kill her. She was afterwards found to
have some 22 knife wounds, plus internal bleeding injuries to her abdomen and liver,
apparently from a kick or similar blow(s) or contact with a hard object, which latter
injuries were the fatal ones.
53. When the police arrived the appellant became extremely violent. He was
described as snarling like an animal and appearing as if deranged. At some stage he
said to the police that “She is evil … The demon’s gone … She had Satan in her
54. The appellant was 46 years old. Since he was about 23 he had been referred
by his GP for out-patient psychiatric consultations from time to time. He had never
been admitted to hospital but had complained of depression, paranoid fears and, at
times, of hearing voices in his head. He had been prescribed anti-depressant and
anti-psychotic drugs and was still under such prescription at the time of the offence,
although he had told the doctors that he was not taking his medicine. One consultant
psychiatrist diagnosed his condition as a mixed personality disorder with paranoid,
emotionally unstable, anxious and dependent traits. On the basis largely of what he
had said to the police, the doctor concluded that at the time of the killing he was
additionally in the grip of an acute psychotic episode and was driven by persecutory
beliefs. The second psychiatrist disagreed that there was a personality disorder, but
concluded that the appellant was at the time of the offence suffering from a paranoid
psychotic illness, most likely schizophrenia. Both expressed the opinion that the
Page 26
different conditions they identified substantially impaired the relevant statutory
abilities, although they were not at one as to which. The first psychiatrist thought
that the ability to form a rational judgment and to exercise self-control were
impaired, but that the defendant knew what he was doing; the second agreed on the
first two counts but additionally thought that the ability to understand the nature of
his conduct was impaired. The Crown case was that he was simply very angry with
his partner, and had been on and off all day, for unremarkable domestic reasons.
There was some evidence of an ability to control himself on previous occasions
when there had been assaults on her which had not been uncontrolled. The
truthfulness of his assertion that he had seen “Satan” was in issue, and may or may
not have been consistent with asserting self-defence at the time. The renewal of the
attack despite the warning presence of the children and the removal of the first knife
might perhaps be some indicator of self-control and give some support to the
contention that the cause was simple anger rather than distorted thinking.
55. That being the state of the evidence, the debate between the two possible
meanings of the expression “substantially” barely arose. If the appellant was indeed
in the grip of a psychotic episode involving persecutory delusions when he killed
his partner, that would, by any ordinary standard, involve substantial impairment of
one or more of the statutory abilities. The real question appears to have been
whether, on the balance of probabilities, he had been. The judge left the issues
squarely to the jury, correctly reminding them more than once that the doctors were
agreed that there was a medical condition substantially impairing his abilities.
56. It follows that for the several reasons set out above, this appeal must be