R v Howe [1986] UKHL 4 (19 February 1986)

Regina

Clarkson (Appellant)
(On Appeal from the Court of Appeal (Criminal Division))

JUDGMENT

Die Jovis 19° Februarii 1987

Upon Report from the Appellate Committee to whom was
referred the Cause Regina against Clarkson, That the

Committee had heard Counsel on Monday the 17th, Tuesday the
18th, Wednesday the 19th and Thursday the 20th days of
November 1986, upon the Petition and Appeal of William George
Clarkson, currently detained at Her Majesty’s Prison,
Leicester Road, Market Harborough, Leicestershire, praying
that the matter of the Order set forth in the Schedule
thereto, namely an Order of Her Majesty’s Court of Appeal
(Criminal Division) of 27th January 1986, might be reviewed
before Her Majesty the Queen in Her Court of Parliament and
that the said Order might be reversed, varied or altered or
that the Petitioner might have such other relief in the
premises as to Her Majesty the Queen in Her Court of
Parliament might seem meet; and Counsel having been heard on
behalf of the Director of Public Prosecutions (on behalf of
Her Majesty), the respondent to the said appeal; and due
consideration had this day of what was offered on either side
in this Cause:

It is Ordered and Adjudged, by the Lords Spiritual and
Temporal in the Court of Parliament of Her Majesty the Queen
assembled, That the said Order of Her Majesty’s Court of
Appeal (Criminal Division) of 27th January 1986 complained of
in the said Appeal be, and the same is hereby, Affirmed; That
the first Certified Question be answered in the negative; That
the second and third Certified Questions be answered in the
affirmative; and That the said Petition and Appeal be, and
the same is hereby, dismissed this House.

Cler: Parliamentor:

Judgment: 19.2.87

HOUSE OF LORDS

REGINA
v.

BURKE
(APPELLANT)

(ON APPEAL FROM THE COURT OF APPEAL (CRIMINAL

DIVISION))

REGINA
v.

HOWE
(APPELLANT)

(ON APPEAL FROM THE COURT OF APPEAL (CRIMINAL

DIVISION))

REGINA
v.

BANNISTER
(APPELLANT)

(ON APPEAL FROM THE COURT OF APPEAL (CRIMINAL

DIVISION))

(CONJOINED APPEALS)

REGINA

CLARKSON
(APPELLANT)

(ON APPEAL FROM THE COURT OF APPEAL (CRIMINAL

DIVISION))

Lord Chancellor
Lord Bridge of Harwich
Lord Brandon of Oakbrook
Lord Griffiths
Lord Mackay of Clashfern

THE LORD HAILSHAM OF ST. MARYLEBONE

My Lords,

These appeals arise from two cases, one originating from a
trial in the Manchester Crown Court before Jupp J. and a jury,
(Howe and Bannister) and one in the Central Criminal Court before
the then Common Sergeant, Judge Tudor Price and a jury (Burke
and Clarkson).

– 1 –

Howe and Bannister were tried with two other defendants
(Murray and Bailey) both of whom during the trial changed their
plea to one of guilty and were appropriately sentenced.

The indictment in the case of Howe, Bannister, Murray and
Bailey accused the four men of two murders (Elgar and Pollitt)
and a conspiracy to murder (Redfern, an intended victim who
escaped in time). The three counts related to three successive
days, respectively 10, 11 and 12 October 1983.

Burke and Clarkson were charged with the murder of a
single victim (Botton) on 9 July 1983.

The four current appellants appealed against their
convictions to the Court of Appeal (Criminal Division) [1986] 1
Q.B. 626 who dismissed all four appeals in a judgment delivered on
21 January 1986. In giving leave to appeal to your Lordships’
House the Court of Appeal (Lord Lane C.J., Russell and Taylor
JJ.) certified three questions of law of general public importance
as involved in the decision. The three certified questions are:

“(1) Is duress available as a defence to a person charged
with murder as a principal in the first degree (the
actual killer)?

1. 1. Can one who incites or procures by duress another to
    kill or to be a party to a killing be convicted of
    murder if that other is acquitted by reason of duress?
  2. Does the defence of duress fail if the prosecution
    prove that a person of reasonable firmness sharing the
    characteristics of the defendant would not have given
    way to the threats as did the defendant?”

The first of these questions involves a reconsideration of the
much discussed decisions in Director of Public Prosecutions for
Northern Ireland v. Lynch [1975] AC 653 and Abbott v. The
Queen [1977] AC 755.

In answering the second question, the Court of Appeal

invited us to reconsider the decision of the Court of Appeal in

Reg. v. Richards [1974] Q.B. 776, by which it considered itself
bound.

In the arguments presented before your Lordships many
other reported authorities and citations from established writers,
including Law Commission Report No. 83 on “defences of general
application” printed by order of the House of Commons, dated 27
July 1977.

Reference to other cases will be made as and where
appropriate. I take the facts of these truly horrible cases almost
verbatim from the judgment of the Lord Chief justice in the
instant appeal [1986] 1 Q.B. 626. First, as to the case of Howe
and Bannister, Murray and Bailey, the facts were as follows:

At the time of the offences Howe and Bailey were 19,
Bannister was 20 and Murray was 35. Howe had one minor
conviction for motoring offences. Bannister had convictions for

– 2 –

theft and burglary but none for violence. He was on probation.
Bailey had convictions for burglary and theft. Murray had
previous court appearances, including two convictions for assault
occasioning actual bodily harm, and in 1974 he had been convicted
of assault with intent to rob and robbery in respect of which he
had been sentenced to eight years’ imprisonment.

Bannister met Murray in Risley Remand Centre. Howe and
Bailey met in Stockport when Bailey was living in a hostel and
Howe happened to be living next door with his grandmother.
Murray came to visit Bailey when he was on six days’ home leave
from a sentence of two and a half years’ imprisonment. Bailey
introduced Howe to Murray. The Lord Chief Justice continued at
pp. 635-636:

“Count 1; murder of Elgar;

The first victim was a 17 year old youth called Elgar.
He was offered a job as a driver by Murray. On the
evening of 10 October 1983 all five men were driven by
Murray up into the hills between Stockport and Buxton,
eventually stopping at some public lavatories at a remote
spot called Goytsclough. Murray at some stage told both
appellants in effect that Elgar was a ‘grass,’ and that they
were going to kill him. Bannister was threatened with
violence if he did not give Elgar ‘a bit of a battering.’
From thenceforwards Elgar, who was naked, sobbing and
begging for mercy, was tortured, compelled to undergo
appalling sexual perversions and indignities, he was kicked
and punched. Bannister and Howe were doing the kicking
and punching. The coup de grace was executed by Bailey
who strangled Elgar with a headlock. It is unnecessary to
go into further details of the attack on Elgar which are
positively nauseating.

“In brief the two appellants asserted that they had

only acted as they did through fear of Murray, believing

that they would be treated in the same way as Elgar had

been treated if they did not comply with Murray’s directions.

“The prosecution were content to assent to the
proposition that death had been caused by Bailey strangling
the victim, although the kicks and punches would have
resulted in death moments later even in the absence of the
strangulation. The body was hidden by the appellants and
the other two men.

“On this basis the appellants were in the position of
what would have earlier been principals in the second degree
and duress was left to the jury as an issue on this count.

“Count 2; murder of Pollitt:

Very much the same course of conduct took place as
with Elgar. On 11 October 1983 the men picked up Pollitt,
a 19-year-old labourer, and took him to the same place
where all four men kicked and punched the youth. Murray
told Howe and Bannister to kill Pollitt, which they did by

– 3 –

strangling him with Bannister’s shoe lace. As the appellants
were in the position of principals in the first degree, the
judge did not leave duress to the jury on this count.

“Count 3; conspiracy to murder Redfern:

The third intended victim was a 21-year-old man.
The same procedure was followed, but Redfern suspected
that something was afoot and managed with some skill to
escape on his motorcycle from what would otherwise have
inevitably been another horrible murder. The judge left the
defence of duress to the jury on this charge of conspiracy
to murder.

“The grounds of appeal, which are the same in
respect of each of these appellants, are as follows: That the
judge erred in directing the jury; (1) in respect of count 2,
that the defence of duress was not available to a principal
in the first degree to the actual killing; (2) in respect of
counts 1 and 3, that the test as to whether the appellants
were acting under duress contains an ‘objective’ element;
that is to say, if the prosecution prove that a reasonable
man in the position of the defendant would not have felt
himself forced to comply with the threats, the defence
fails.”

So much for the facts relating to the appellants Howe
and Bannister.

The Lord Chief Justice then turned to the case of
Burke and Clarkson. In the case of Howe and Bannister the
defence of duress was left to the jury by Jupp J. on the
third count of conspiracy to murder and rejected by the
jury. However, Jupp J. had directed the jury in relation to
duress. In this case Burke dressed as a policeman had killed
Botton with a sawn off shotgun at the entrance of Botton’s
house. The Lord Chief Justice continued at p. 637:

“The prosecution’s case was that Burke had done this
at the request of Clarkson, who was anxious to prevent
Botton from giving evidence against him.”

“Clarkson’s defence was that he had nothing to do
with the shooting at all. Burke’s defence was that he had
agreed to shoot Botton because of his fear that Clarkson
would kill him if he did not, but when it came to the
event, the gun went off accidentally and the killing
therefore was unintentional and amounted to no more than
manslaughter.

“This defence was evidently rejected by the judge, but
his defence in respect of manslaughter was left to the jury.

A further submission on behalf of Burke was not argued
before us. On the third certified question Jupp J.’s direction to
the jury on the nature of duress, although not identical with that
of Judge Tudor Price equally raised the question of the objective
element in the threats required raised by the third question
certified by the Court of Appeal. The only ground of appeal

– 4 –

separately argued before your Lordships on the part of Clarkson
depends on the answer to be given to the second certified question
to which I will return later.

On the third certified question Mr. Self on behalf of the
appellants other than Clarkson addressed to us an impressive
argument that the test applied is not objective, i.e. not:

“. . . whether the threat was of such gravity that it might
well have caused a reasonable man placed in the same
situation as the defendants to act as the defendant did”

nor . . .

“would a sober person of reasonable firmness sharing the
defendant’s characteristics have responded to the threats by
taking part in the killing?”

Mr. Self’s submission was to the effect that where the
defence of duress is available to an accused, the test of duress
does not contain either of these objective elements but is purely
subjective to the accused and depends solely on the effect which
the actual threat had on the mind of the particular accused.

From the above, it will be seen that each of the appellants
other than Clarkson were actual participants, whether as principal
in the first or second degree, in the perpetration of the actual
killing of the respective victims in respect of the counts of

murder. Clarkson’s defence, which was that he had nothing
whatever to do with the murder of Botton was rejected by the
jury and nothing turns on this at all except that the jury’s verdict
means that he was the real villain of the piece, since, on any
view of the facts, Burke acted at his suggestion and was very
much under his influence. The success of Clarkson’s appeal
depends solely on the answer to the second certified question, and
only arises in the event that Burke’s appeal on the first or third
question succeeds or succeeds to the extent that a conviction of
Burke for manslaughter should be substituted in the case for the
verdict of murder. Counsel for Clarkson very properly conceded
that the decision of the second question was free from authority
binding on this House and open to your Lordships, both on principle
and authority, since Reg. v. Richards [1974] Q.B. 776, by which
the Court of Appeal regarded itself as bound, is not binding on
your Lordships’ House. It was conceded that, on the facts as they
must be assumed to be for the purposes of the second question,
apart from authority the point was totally devoid of merit, and,
despite the advocacy of counsel, their Lordships did not think fit
to trouble counsel for the respondent on this question. The Court
of Appeal does not appear to have referred directly (although it
was mentioned in argument), to the horrible case of Reg. v. Cogan
[1976] QB 217, which, if the answer had not been otherwise
obvious, seems to me to dispose of the matter. In the event, we
also did not call upon the respondent to argue the third certified
question since, in my opinion, and, I believe, that of my noble and
learned friends, the definition of duress, whether applicable to
murder or not, was correctly stated by both trial judges to contain
an objective element on the lines of their respective directions and
this must involve a threat of such a degree of violence that “a
person of reasonable firmness” with the characteristics and in the

– 5 –

situation of the defendant could not have been expected to resist.
No doubt there are subjective elements as well, but, unless the
test is purely subjective to the defendant which, in my view, it is
not, the answer to the third certified question, like that to the
second, must be “yes.”

This leaves us free to discuss the first, and principal issue
in the appeal which is the answer to be given to the first of the
three certified questions. In my opinion, this must be decided on
principle and authority, and the answer must in the end demand a
reconsideration of the two authorities of Director of Public
Prosecutions for Northern Ireland v.Lynch [1975] AC 653 and
Abbott v. The Queen [1977] AC 755. Having been myself a party
to Abbott, I feel I owe it to the two noble and learned friends
then with me in the majority to say that we were very conscious
of the fact that our decision would only be of persuasive authority
in the English jurisdiction whilst the decision in Lynch, though a
Northern Irish case, which distinguished for the purposes of duress
between principals in the first degree on the one hand, and
principals in the second degree and aiders and abettors on the
other, being a decision of the House of Lords would be likely to
be treated as binding throughout England and Wales as well as
Northern Ireland. We did, however, say at, [1977] AC 755, 763:

“Whilst their Lordships feel bound to accept the decision of
the House of Lords in Lynch’s case they find themselves
constrained to say that had they considered (which they do
not) that that decision is an authority which requires the
extension of the doctrine to cover cases like the present
they would not have accepted it.”

Speaking only for myself, it was precisely because the three noble
and learned Lords in the majority in Lynch had expressly left open
the availability of duress as a defence to the actual participant in
a murder that I found it possible to accept the decision in Lynch
without criticism, and then only because the Abbott appeal was
solely concerned with the question so expressly left open. One
only needs to read the facts in Abbott to be aware of exactly
what the Board was being asked to do if it extended Lynch and
allowed the appeal.

The present case, in my opinion, affords an ideal and never
to be repeated opportunity to consider as we were invited
expressly to do by the respondent, the whole question afresh, if
necessary, by applying the Practice Statement (Judicial Precedent)
[1966] 1 W.L.R. 1234 to the decision in Lynch.

I therefore consider the matter first from the point of view
of authority. On this I can only say that at the time when Lynch
was decided the balance of weight in an unbroken tradition of
authority dating back to Hale and Blackstone seems to have been
accepted to have been that duress was not available to a
defendant accused of murder. I quote only from Hale and
Blackstone. Thus Hale’s Pleas of the Crown, vol. 1, p. 51:

“If a man be desperately assaulted, and in peril of death,
and cannot otherwise escape, unless to satisfy his assailant’s
fury he will kill an innocent person then present, the fear
and actual force will not acquit him of the crime and

– 6 –

punishment of murder, if he commit the fact; for he ought
rather to die himself, than kill an innocent.”

Blackstone’s Commentaries on the Laws of England. 1857 ed., vol.
4, p. 28 was to the same effect. He wrote that a man under
duress:

“. . . ought rather to die himself than escape by the murder
of an innocent.”

I forbear to quote the eloquent and agonised passage in the
dissenting speech of Lord Simon of Glaisdale in Lynch at p. 695 of
the report, or the more restrained exposition of Lord Kilbrandon at
p. 702 on the law as expressed in Reg. v. Dudley and Stephens
(1884) 14 QBD 273. These quotations are unnecessary since it
seems to have been accepted both by the majority in Lynch and
the minority in Abbott, that, to say the least, prior to Lynch
there was a heavy preponderance of authority against the
availability of the defence of duress in cases of murder.

I would only add that Article 8 of the Nuremberg statute
which was, at the time, universally accepted, save for its
reference to mitigation, as an accurate statement of the common
law both in England and the United States of America that:

“The fact that the defendant acted pursuant to the order of
his Government or of a superior shall not free him from
responsibility, but may be considered in mitigation of
punishment if the Tribunal determines that justice so
requires.”

“Superior orders” is not identical with “duress,” but, in the
circumstances of the Nazi regime, the difference must often have
been negligible. I should point out that under Article 6, the
expression “war crimes” expressly included that of murder; which,
of course, does not include the killing of combatants engaged in
combat.

What then is said on the other side? I accept, of course,
that duress for almost all other crimes had been held to be a
complete defence. I need not cite cases. They are carefully
reviewed in the case of Lynch and establish I believe that the
defence is of venerable antiquity and wide extent. I pause only to
say that although duress has, in my view, never been defined with
adequate precision, two views of its nature can no longer be
viewed as correct in the light of reported authority. The first is
that of Stephen in his History of the Criminal Law of England
(1883), who first promulgated the opinion that duress was not a
defence at all but, as in the Nuremberg statute, only a matter of
mitigation. The fact is that, where it is applicable at ail, in a
long line of cases duress has been treated as a matter of defence
entitling an accused to a complete acquittal. But in almost every
instance where duress is so treated a cautionary note has been
sounded excluding murder in terms sometimes more, and sometimes
less emphatic, from the number of crimes where it can be put
forward.

The second unacceptable view is that, possibly owing to a
misunderstanding which has been read into some judgments, duress

– 7 –

as a defence affects only the existence or absence of mens rea.
The true view is stated by Lord Kilbrandon (of the minority) Lynch
[1975] AC 653 and by Lord Edmund-Davies (of the majority) in
his analysis at p. 709.

Lord Kilbrandon said at p. 703:

“The decision of the threatened man whose constancy is
overbourne so that he yields to the threat, is a calculated
decision to do what he knows to be wrong, and is therefore
that of a man with, perhaps to some exceptionally limited
extent, a ‘guilty mind.’ But he is at the same time a man
whose mind is less guilty than is his who acts as he does
but under no such constraint.” [emphasis mine.]

In coming to the same conclusion Lord Edmund-Davies at pp. 709-
710 quoted from Professor Glanville Williams’ well known treatise
Criminal Law, 2nd ed. (1961) p. 751 para. 242:

“True duress is not inconsistent with act and will as a
matter of legal definition, the maxim being coactus voiui.
Fear of violence does not differ in kind from fear of
economic ills, fear of displeasing others, or any other
determinant of choice, it would be inconvenient to regard a
particular type of motive as negativing of will.”

After approving a paragraph from Lowry C.J., Lord Edmund-Davies
went on to say that two quotations from Lord Goddard C.J. in the
disgusting case of Reg. v. Bourne (1952) 36 Cr. App. R. 125 were
subject to criticism on this score. (See the report in Lynch at p.
710).

Before I leave the question of reported authority I must
refer to two other cases. The first is Reg. v. Kray (Ronald)
(1969) 53 Cr.App.R. 569 which was, to some extent, relied on by
the majority in Lynch, on the score of an obiter dictum of
Widgery L.J. at p. 578. I do not myself regard this passage as
authoritative. It depends on a concession by the Crown regarding
a party who was not before the Court of Appeal as his case had
been disposed of at first instance in order to found a submission
by the appellants. The dictum is also open to the criticism that
Widgery L.J. appeared to treat duress as making a person
otherwise than an “independent actor” which is contrary to the
analysis which I have accepted above.

The other reported authority is the famous and important
case of Reg. v. Dudley and Stephens (1884) 14 QBD 273. That
is generally and, in my view correctly, regarded as an authority on
the availability of the supposed defence of necessity rather than
duress. But I must say frankly that, if we were to allow this
appeal, we should, I think, also have to say that Dudley and
Stephens was bad law. There is, of course, an obvious distinction
between duress and necessity as potential defences; duress arises
from the wrongful threats or violence of another human being and
necessity arises from any other objective dangers threatening the
accused. This, however, is, in my view a distinction without a
relevant difference, since on this view duress is only that species
of the genus of necessity which is caused by wrongful threats. I
cannot see that there is any way in which a person of ordinary

– 8 –

fortitude can be excused from the one type of pressure on his will
rather than the other.

I shall revert to Dudley and Stephens when I come to
consider some of the issues of principle involved in our response to
the first certified question. But at this stage I feel that I should
say that in Abbott I would have been prepared to accept a
distinction between Abbott and Lynch on the basis of the argument
which appeared to attract Lord Morris of Borth-y-Gest at pp. 671-
672 of Lynch. I would not myself have immersed myself in the
somewhat arcane terminology of accessory, principal in the second
degree, and aiding and abetting. But it did seem to me then, and
it seems to me now, that there is a valid distinction to be drawn
in ordinary language between a man who actually participates in
the irrevocable act of murder to save his own skin or that of his
nearest and dearest and a man who simply participates before or
after the event in the necessary preparation for it or the escape
of the actual offender. It is as well to remember that, in the
case of Abbott the facts were that Abbott had dug a pit, thrown
the victim into it, subjected her in co-operation with others to
murderous blows and stab wounds and then buried her alive. It
seems to me that those academics who see no difference between
that case and the comparatively modest part alleged (falsely as is
now known) in Lynch to have been played by the defendant under
duress have parted company with a full sense of reality.
Nevertheless and in spite of this, and in the face of the somewhat
intemperate criticism to which this type of distinction has
sometimes been subjected since Abbott I am somewhat relieved to
know that the views of my noble and learned friends on the main
issue permit me to escape from such niceties and simply to say
that I do not think that the decision in Lynch can be justified on
authority and that, exercising to the extent necessary, the freedom
given to us by the Practice Statement (Judicial Precedent) [1966] 1
W.L.R. 1234 which counsel for the respondent urged us to apply, I
consider that the right course in the instant appeal is to restore
the law to the condition in which it was almost universally thought
to be prior to Lynch. It may well be that that law was to a
certain extent unclear and to some extent gave rise to anomaly.
But these anomalies I believe to be due to a number of factors
extraneous to the present appeal and to the intrinsic nature of
duress. The first is the mandatory nature of the sentence in
murder. The second resides in the fact that murder being a
“result” crime, only being complete if the victim dies within the
traditional period of a year and a day and that, in consequence, a
different crime may be charged according to whether or not the
victim actually succumbs during the prescribed period. The third
lies in the fact (fully discussed amongst many other authorities in
Reg. v. Hyam [1975] AC 55) that, as matters stand, the mens rea
in murder consists not simply in an intention to kill, but may
include an intent to commit grievous bodily harm. It has always
been possible for Parliament to clear up this branch of the law (or
indeed to define more closely the nature and extent of the
availability of duress as a defence). But Parliament has
conspicuously, and perhaps deliberately, declined to do so. In the
meantime, I must say that the attempt made in Lynch to clear up
this situation by judicial legislation has proved to be an excessive
and perhaps improvident use of the undoubted power of the courts
to create new law by creating precedents in individual cases.

– 9 –

This brings me back to the question of principle. I begin by
affirming that, while there can never be a direct correspondence
between law and morality, an attempt to divorce the two entirely
is and has always proved to be, doomed to failure, and, in the
present case, the overriding objects of the criminal law must be to
protect innocent lives and to set a standard of conduct which
ordinary men and women are expected to observe if they are to
avoid criminal responsibility.

No one, who has read the case of Dudley and Stephens
whether in the Law Reports, or in the more popular and discursive
volume published by Professor Simpson Cannibalism in the Common
Law (1984), can fail to be moved by the poignant and anguished
situation to which the two shipwrecked mariners with Brooks (who
was not guilty) and the innocent boy of 17, who was the victim,
were exposed and which led the Home Secretary of the day to
commute a death sentence for murder to one of 18 months
imprisonment. Nevertheless, when one comes to examine the case
as one of legal principle it is, I believe, the case that the
conclusion reached by the judges of the Queen’s Bench Division
and voiced by Lord Coleridge C.J., not without manifest
compassion, has met with very wide acceptance. I quote from
(1884) 14 Q.B.D. 273, 286-288:

“Now it is admitted that the deliberate killing of this
unoffending and unresisting boy was clearly murder, unless
the killing can be justified by some well-recognised excuse
admitted by the law. It is further admitted that there was
in this case no such excuse, unless the killing was justified
by what has been called ‘necessity.’ But the temptation to
the act which existed here was not what the law has ever
called necessity. Nor is this to be regretted. Though law
and morality are not the same, and many things may be
immoral which are not necessarily illegal, yet the absolute
divorce of law from morality would be of fatal consequence;
and such divorce would follow if the temptation to murder
in this case were to be held by law an absolute defence of
it. It is not so. To preserve one’s life is generally
speaking a duty, but it may be the plainest and the highest
duty to sacrifice it. War is full of instances in which it is
a man’s duty not to live, but to die. The duty, in case of
shipwreck, of a captain to his crew, of the crew to the
passengers, of soldiers to women and children, as in the
noble case of the Birkenhead; these duties impose on men
the moral necessity, not of the preservation, but of the
sacrifice of their lives for others, from which in no country,
least of all, it is to be hoped, in England, will men ever
shrink, as indeed, they have not shrunk. . . It is not
needful to point out the awful danger of admitting the
principle which has been contended for. Who is to be the
judge of this sort of necessity? By what measure is the
comparative value of lives to be measured? Is it to be
strength, or intellect, or what? It is plain that the
principle leaves to him who is to profit by it to determine
the necessity which will justify him in deliberately taking
another’s life to save his own. In this case the weakest,
the youngest, the most unresisting, was chosen. Was it
more necessary to kill him than one of the grown men?
The answer must be no.”

– 10 –

It was pointed out in a footnote in this case at p.288
(attributed to Grove J.) that if the principle were once admitted
and the castaways not rescued, in the meantime it would have
been lawful for the strongest of the four men to eat his way
through the whole crew of the drifting boat, in order to be
rescued himself.

I must dissent profoundly from the statement of my
predecessor Bacon, a greater moralist perhaps in theory than in
practice, where, quoted by Lord Coleridge C.J. in Dudley and
Stephens, in his commentary on the maxim “necessitas inducit
privilegium quoad jura privata” he writes at p. 285:

“If divers be in danger of drowning by the casting away of
some boat or barge, and one of them get to some plank, or
on the boat’s side to keep himself above water, and another
to save his life thrust him from it, whereby he is drowned,
this is neither se defendendo nor by misadventure, but
justifiable.”

I also dissociate myself from the view of Rumpff J. in the South
African case S. v. Goliath 1972 (3) S.A.I (based, however, on
Roman Dutch law) and quoted in Abbott at [1977] AC 755, 771
as a justification for the opinion of the dissenting minority:

“It is generally accepted . . . that for the ordinary person
in general his life is more valuable than that of another.
Only they who possess the quality of heroism will
intentionally offer their lives for another. Should the
criminal taw then state that compulsion could never be a
defence to a charge of murder, it would demand that a
person who killed another under duress, whatever the
circumstances, would have to comply with a higher standard
than that demanded of the average person. I do not think
that such an exception to the general rule which applies in
criminal law, is justified.”

In general, I must say that I do not at all accept in relation
to the defence of murder it is either good morals, good policy or
good law to suggest, as did the majority in Lynch and the minority
in Abbott that the ordinary man of reasonable fortitude is not to
be supposed to be capable of heroism if he is asked to take an
innocent life rather than sacrifice his own. Doubtless in actual
practice many will succumb to temptation, as they did in Dudley
and Stephens. But many will not, and I do not believe that as a
“concession to human frailty” the former should be exempt from
liability to criminal sanctions if they do. I have known in my own
lifetime of too many acts of heroism by ordinary human beings of
no more than ordinary fortitude to regard a law as either “just or
humane” which withdraws the protection of the criminal law from
the innocent victim and casts the cloak of its protection upon the
coward and the poltroon in the name of a “concession to human
frailty.”

I must not, however, underestimate the force of the
arguments on the other side, advanced as they have been with
such force and such persuasiveness by some of the most eminent
legal minds, judicial and academic, in the country.

– 11 –

First, amongst these is, perhaps, the argument from logic
and consistency. A long line of cases, it is said, carefully
researched and closely analysed, establish duress as an available
defence in a wide range of crimes, some at least, like wounding
with intent to commit grievous bodily harm, carrying the heaviest
penalties commensurate with their gravity. To cap this, it is
pointed out that at least in theory, a defendant accused of this
crime under section 18 of the Offences against the Person Act
1861, but acquitted on the grounds of duress, will still be liable to
a charge of murder if the victim dies within the traditional period
of one year and a day. I am not, perhaps, persuaded of this last
point as much as I should. It is not simply an anomaly based on
the defence of duress. It is a product of the peculiar mens rea
allowed on a charge of murder which is not confined to an intent
to kill. More persuasive, perhaps, is the point based on the
availability of the defence of duress on a charge of attempted
murder, where the actual intent to kill is an essential prerequisite.
It may be that we must meet this casus omissus in your Lordships’
House when we come to it. It may require reconsideration of the
availability of the defence in that case too.

I would, however, prefer to meet the case of alleged
inconsistency head on. Consistency and logic, though inherently
desirable, are not always prime characteristics of a penal code
based like the Common Law on custom and precedent. Law so
based is not an exact science. All the same, I feel I am required
to give some answer to the question posed. If duress is available
as a defence to some crimes of the most grave why, it may
legitimately be asked, stop at murder, whether as accessory or
principal and whether in the second or the first degree? But
surely I am entitled, as in the view of the Common Sergeant in
the instant case of Clarkson and Burke, to believe that some
degree of proportionality between the threat and the offence must,
at least to some extent, be a prerequisite of the defence under
existing law. Few would resist threats to the life of a loved one
if the alternative were driving across the red lights or in excess
of 70 m.p.h. on the motorway. But, to use the Common
Sergeant’s analogy, it would take rather more than the threat of a
slap on the wrist or even moderate pain or injury to discharge the
evidential burden even in the case of a fairly serious assault. In
such a case the “concession to human frailty” is no more than to
say that in such circumstances a reasonable man of average
courage is entitled to embrace as a matter of choice the
alternative which a reasonable man could regard as the lesser of
two evils. Other considerations necessarily arise where the choice
is between the threat of death or a fortiori of serious injury and
deliberately taking an innocent life. In such a case a reasonable
man might reflect that one innocent human life is at least as
valuable as his own or that of his loved one. In such a case a
man cannot claim that he is choosing the lesser of two evils.
Instead he is embracing the cognate but morally disreputable
principle that the end justifies the means.

I am not so shocked as some of the judicial opinions have
been at the need, if this be the conclusion, to invoke the
availability of administrative as distinct from purely judicial
remedies for the hardships which might otherwise occur in the
most agonising cases. Even in Dudley and Stephens in 1884 when
the death penalty was mandatory and frequently inflicted, the

– 12 –

prerogative was used to reduce a sentence of death by hanging to
one of 18 months in prison. In murder cases the available
mechanisms are today both more flexible and more sophisticated.
The trial judge may make no minimum recommendation. He will
always report to the Home Secretary, as he did in the present
case of Clarkson and Burke. The Parole Board will always
consider a case of this kind with a High Court judge brought into
consultation. In the background is always the prerogative and, it
may not unreasonably be suggested, that is exactly what the
prerogative is for. If the law seems to bear harshly in its
operation in the case of a mandatory sentence on any particular
offender there has never been a period of time when there were
more effective means of mitigating its effect than at the present
day. It may well be thought that the loss of a clear right to a
defence justifying or excusing the deliberate taking of an innocent
life in order to emphasise to all the sanctity of a human life is
not an excessive price to pay in the light of these mechanisms.
Murder, as every practitioner of the law knows, though often
described as one of the utmost heinousness, is not in fact
necessarily so, but consists in a whole bundle of offences of vastly
differing degrees of culpability, ranging from brutal, cynical and
repeated offences like the so called Moors murders to the almost
venial, if objectively immoral, “mercy killing” of a beloved partner.

Far less convincing than the argument based on consistency
is the belief which appears in some of the judgments that the law
must “move with the times” in order to keep pace with the
immense political and social changes since what are alleged to
have been the bad old days of Blackstone and Hale. I have
already dealt with this argument in my respectful criticism of the
dissent in Hyam [1975] AC 55. The argument is based on the
false assumption that violence to innocent victims is now less
prevalent than in the days of Hale or Blackstone. But I doubt
whether this is so. We live in the age of the holocaust of the
Jews, of international terrorism on the scale of massacre, of the
explosion of aircraft in mid air, and murder sometimes at least as
obscene as anything experienced in Blackstone’s day. Indeed one
of the present appeals may provide an example. I have already
mentioned the so-called Moors murders. But within weeks of
hearing this appeal a man was convicted at the Central Criminal
Court of sending his pregnant mistress on board an international
aircraft at Heathrow, with her suitcase packed with a bomb and
with the deliberate intention of sending the 250 occupants, crew,
passengers, mistress and all to a horrible death in mid air. I
cannot forbear to say that if Abbott was wrongly decided, and had
the attempt succeeded, the miscreant who did this would have
been free to escape scot free had he been in a position to
discharge the evidential burden on duress and had the prosecution,
on the normal Woolmington principles (Woolmington v. Director of
Public Prosecutions [1935] AC 462, 482), been unable to exclude
beyond reasonable doubt the possibility of his uncorroborated word
being true. I must also point out in this context that known
terrorists are more and not less vulnerable to threats than the
ordinary man and that a plea of duress in such a case may be all
the more plausible on that account. To say this is not to cast
doubt on the reliability and steadfastness of juries. Counsel for
the appellants was able to say with perfect truth that, where
duress in fact has been put forward in cases where it was
available, juries have been commendably robust as they were in

– 13 –

the instant cases in rejecting it where appropriate. The question
is not one of the reliability of juries. It is one of principle.
Should the offence of duress be available in principle in such a
case as that of Hindawi where, of course, it was not put forward?
The point which I am at the moment concerned to make is that it
is not clear to me that the observations of Blackstone and Hale,
and almost every respectable authority, academic or judicial, prior
to Lynch are necessarily to be regarded in this present age as
obsolescent or inhumane or unjust owing to some supposed
improvement in the respect for innocent human life since their
time which unfortunately I am too blind to be able for myself to
perceive. Still less am I able to see that a law which denies such
a defence in such a case must be condemned as lacking in justice
or humanity rather than as respectable in its concern for the
sanctity of innocent lives. I must add that, at least in my view,
if Abbott were wrongly decided some hundreds who suffered the
death penalty at Nuremberg for murders were surely the victims
of judicial murder at the hands of their conquerors owing to the
operation of Article 8. Social change is not always for the better
and it ill becomes those of us who have participated in the cruel
events of the 20th century to condemn as out of date those who
wrote in defence of innocent lives in the 18th century.

During the course of argument it was suggested that there
was available to the House some sort of half way house between
allowing these appeals and dismissing them. The argument ran
that we might treat duress in murder as analogous to provocation,
or perhaps diminished responsibility, and say that, in indictments

for murder, duress might reduce the crime to one of manslaughter.
I find myself quite unable to accept this. The cases show that
duress, if available and made out, entitles the accused to a clean
acquittal, without, it has been said, the “stigma” of a conviction.
Whatever other merits it may have, at least the suggestion makes
nonsense of any pretence of logic or consistency in the criminal
law. It is also contrary to principle. Unlike the doctrine of
provocation, which is based on emotional loss of control, the
defence of duress, as I have already shown, is put forward as a
“concession to human frailty” whereby a conscious decision, it may
be coolly undertaken, to sacrifice an innocent human life is made
as an evil lesser than a wrong which might otherwise be suffered
by the accused or his loved ones at the hands of a wrong doer.
The defence of diminished responsibility (which might well, had it
then been available to Dudley and Stephens, have prevailed there)
is statutory in England though customary in Scotland, the law of
its origin. But in England at least it has a conceptual basis
defined in the Homicide Act 1957 which is totally distinct from
that of duress if duress be properly analysed and understood.
Provocation (unique to murder and not extending even to “section
18” offences) is a concession to human frailty due to the extent
that even a reasonable man may, under sufficient provocation
temporarily lose his self control towards the person who has
provoked him enough. Duress, as I have already pointed out, is a
concession to human frailty in that it allows a reasonable man to
make a conscious choice between the reality of the immediate
threat and what he may reasonably regard as the lesser of two
evils. Diminished responsibility as defined in the Homicide Act
1957 depends on abnormality of mind impairing mental
responsibility. It may overlap duress or even necessity. But it is
not what we are discussing in the instant appeal.

– 14 –

I must add that, had I taken a different view, in the cases
of Bannister and Howe and, for rather different reasons, in the
case of Burke, I would have gone on to consider the questions
whether in any of these appeals the appellants had discharged the
evidential burden in duress, or whether, if they had, on the facts
described in the judgment of the Lord Chief Justice, the proviso
should not have been applied in every case. The case of Clarkson
is surely beyond dispute on the assumption that the second
certified question is not answered in his favour. But whatever
may be the characteristics of duress, even on the existing law the
ingredients of immediacy and absence of voluntary association (see
Reg. v. Fitzpatrick [1977] N.I. 20), must be essential components
of the evidential burden more or less on the lines of the draft bill
annexed to the Law Commission Report No. 83, to which I have
referred above. Even apart from this and on the assumption that
the matter should properly have been left to the jury, I am rather
more than doubtful whether any properly instructed jury could have
acquitted on the murder charges in either of the instant cases or
on the facts of Abbott. It is not necessary to express a concluded
opinion on this since, for the reasons I have adumbrated above, I
consider that these appeals should be dismissed and the certified
questions answered respectively (1) no, (2) yes, (3) yes. If so, the
questions relating to the proviso and evidential burden do not
arise. So far as I have indicated, the decision of this House in
Lynch [1975] AC 653 should be regarded as unsatisfactory and the
law left as it was before Lynch came up for decision. The
decision in Abbott [1977] AC 755 should be followed, and, unless it
can be distinguished on the facts, that in Richards [1974] Q.B. 776
should be overruled.

LORD BRIDGE OF HARWICH

My Lords,

The defence of duress, as a general defence available at
common law which is sufficient to negative the criminal liability
of a defendant against whom every ingredient of an offence has
otherwise been proved, is difficult to rationalise or explain by
reference to any coherent principle of jurisprudence. The theory
that the party acting under duress is so far deprived of volition as
to lack the necessary criminal intent has been clearly shown to be
fallacious: Director of Public Prosecutions for Northern Ireland v.
Lynch [1975] AC 653, 7Q9H-711A per Lord Edmund-Davies. No
alternative theory seems to provide a wholly satisfactory
foundation on which the defence can rest. The law, therefore,
might have developed more logically had it adopted the view of
Stephen, expressed in his History of the Criminal Law of England
(1883), that duress should be a matter, not of defence, but of
mitigation. If this course had been followed, it might sensibly
have led to the further development that, in the case of murder,
duress, like provocation, would have sufficed to reduce the offence
from murder to manslaughter. But that is not the law and, though
it is open to Parliament to decide that it ought to be, that course
is not open to us. We have to accept the law as we find it and,
given the lack of any clear underlying principle to which we can
refer, we must not, I think, be wholly surprised if the solution to

– 15 –

the problem posed by the first certified question arising in these
appeals fails to remove all the anomalies which some may discern
in this field of the law.

If we take the majority decisions of this House in Lynch
and of the Privy Council in Abbott v. The Queen [1977] AC 755,
as establishing the present law, duress is a complete defence to a
murderer otherwise guilty as a principal in the second degree, it is
no defence to a murderer guilty as a principal in the first degree.
Technically, of course, the two decisions were made in two
distinct jurisdictions, though three Lords of Appeal (Lord
Wilberforce, Lord Kilbrandon and Lord Edmund-Davies) were party
to both. In this situation it is an odd quirk of the system
operated by two ultimate appellate tribunals, each deciding by a
majority, that their two decisions should have the combined effect
of affirming a distinction which four out of the seven participants
in the decisions (Lord Simon of Glaisdale and Lord Kilbrandon in
Lynch, Lord Wilberforce and Lord Edmund-Davies in Abbott)
expressly rejected as untenable. The only speech which gives any
positive, even if somewhat lukewarm, support to the distinction is
that of Lord Morris of Borth-y-Gest in Lynch. He said, at p. 671:

“The issue in the present case is therefore whether there is
any reason why the defence of duress, which in respect of a
variety of offences has been recognised as a possible
defence, may not also be a possible defence on a charge of
being a principal in the second degree to murder. I would
confine my decision to that issue. It may be that the law
must deny such a defence to an actual killer, and that the
law will not be irrational if it does so.”

Later, referring to the “actual killer,” he said: “There, I think,
before allowing duress as a defence it may be that the law will
have to call a halt.” Lord Morris supported the distinction by
illustrations of theoretical cases where principals in the second
degree (as in the case of Lynch itself) might be seen as playing a
relatively minor role in a murderous enterprise. These passages
are naturally referred to in the majority judgment in Abbott, but
this hardly strengthens support for the distinction when one
remembers, first, that Lord Kilbrandon was one of the majority,
secondly, that the essential attitude of the majority in Abbott to
the decision in Lynch is expressed by saying that: “… their
Lordships, whilst loyally accepting the decision in Lynch’s case, are
certainly not prepared to extend it.”

As is pointed out in Smith and Hogan, Criminal Law, 5th
ed. (1983), p. 211 there is no necessary correspondence between
degrees of culpability of parties to a murder and the technical
distinction between principals in the first and second degrees,
which would make the latter a rational cut-off point at which the
defence of duress ceases to be available. My noble and learned
friend, Lord Griffiths, gives cogent examples to illustrate this. I
can find nothing whatever to be said for leaving the law as it
presently stands. Ineluctably, as it seems to me, we must either
move forward and affirm the view of the minority in Abbott that
duress is available as a defence to murder generally, or depart
from Lynch and restore the law as it was generally accepted
before Lynch, whereby duress was not a defence available to any
party otherwise guilty of murder.

– 16 –

All other considerations apart, I should myself have found a
sufficient reason for deciding in favour of the latter course in the
consideration that it was never open to the House in its judicial
capacity to make such a fundamental reform of the law as the
introduction of duress as a defence to murder involved. The
passages in Lynch in the speeches to this effect of Lord Simon of
Glaisdale (695D-696C) and of Lord Kilbrandon (699H-701B) seem to
me to carry conviction. But that conviction is now immensely
strengthened by the knowledge that Parliament, even against the
background of the plainly unsatisfactory present state of the law,
has in ten years taken no action on the Report of the Law
Commission, No. 83. If duress is now to be made available
generally as a defence to murder, it seems to me incontrovertible
that the proper means to effect such a reform is by legislation
such as that proposed by the Law Commission. Not only is it for
Parliament to decide whether the proposed reform of the law is
socially appropriate, but it is also by legislation alone, as opposed
to judicial development, that the scope of the defence of duress
can be defined with the degree of precision which, if it is to be
available in murder at all, must surely be of critical importance.

My Lords, I have had the advantage of reading in advance
the speeches of my noble and learned friends, Lord Griffiths and
Lord Mackay of Clashfern. I entirely agree with them and
gratefully adopt their fuller reasoning, in addition to my own short
observations, as leading to the conclusion that the appeals should
be dismissed and the first certified question answered in the
negative. I would wish to emphasise in particular my concurrence
with Lord Griffiths in the weight he attaches to the opinion of
Lord Lane C.J. as expressed in the judgment of the Court of
Appeal (Criminal Division) appealed against.

For the reasons given in the speeches of my noble and
learned friends, the Lord Chancellor and Lord Mackay of
Clashfern, I would answer the second and third certified questions
in the affirmative.

LORD BRANDON OF OAKBROOK

My Lords,

I have had the advantage of reading in draft the speech
prepared by my noble and learned friend, Lord Mackay of
Clashfern. I agree with it, and for the reasons which he gives I
would dismiss the appeal.

I cannot pretend, however, that I regard the outcome as
satisfactory. It is not logical, and I do not think it can be just,
that duress should afford a complete defence to charges of all
crimes less grave than murder, but not even a partial defence to a
charge of that crime. I say nothing as to treason, for that is not
here in issue. I am persuaded, nevertheless, to agree with my
noble and learned friend by three considerations. First, it seems
to me that, so far as the defence of duress is concerned, no valid
distinction can be drawn between the commission of murder by one
who is a principal in the first degree and one who is a principal in

– 17 –

the second degree. Secondly, I am satisfied that the common law
of England has developed over several centuries in such a way as
to produce the illogical, and as I think unjust, situation to which I
have referred. Thirdly, I am convinced that, if there is to be any
alteration in the law on such an important and controversial
subject, that alteration should be made by legislation and not by
judicial decision.

LORD GRIFFITHS

My Lords,

As a general rule I support the view that in criminal
appeals to this House it is desirable wherever possible to have one
speech, so that the judges and practitioners may turn to one
source for authoritative guidance. Clarity, certainty and, wherever
possible, simplicity are invaluable attributes of the criminal law
which must be understood by laymen and especially by jurymen as
well as lawyers. This will usually be better achieved by the
distillation of the consensus view of the House in one speech
rather than leaving judges of first instance to pick their way
through five speeches in an attempt to apply the principle of the
decision to the trial currently taking place before them. There
are, however, exceptions to every rule and as I believe that we
should now depart from the decision of this House in Director of
Public Prosecutions for Northern Ireland v. Lynch [1975] AC 653, I
feel that I should shortly state the reasons for my opinion.

For centuries it was accepted that English criminal law did
not allow duress as a defence to murder. It was so stated in
Hale’s Pleas of the Crown, repeated by Blackstone in his
Commentaries, and so taught by all the authoritative writers on
criminal law. It was accepted by those responsible for drafting
the criminal codes for many parts of the British Empire and they
provided, in those codes, that duress should not be a defence to
murder. In Reg. v. Tyler and Price (1838) 8 C. & P. 616,
Denman C.J. told the jury in emphatic language that they should
not accept a plea of duress that was put up in defence to a
charge of murder against those who were not the actual killers.
Fifty years later, in Reg. v. Dudley and Stephens (1884) 14 Q.B.D.
273, the defence of necessity was denied to the men who had
killed the cabin boy and eaten him in order that they might
survive albeit only Stephens was the actual killer. The reasoning
that underlies that decision is the same as that which denies
duress as a defence to murder. It is based upon the special
sanctity that the law attaches to human life and which denies to a
man the right to take an innocent life even at the price of his
own or another’s life.

There are surprisingly few reported decisions on duress but
it can not be gainsaid that the defence has been extended,
particularly since the second war, to a number of crimes. I think
myself it would have been better had this development not taken
place and that duress had been regarded as a factor to be taken
into account in mitigation as Stephen suggested in his History of
the Criminal Law in England (1883). However, as Lord Morris of

– 18 –

Borth-y-Gest said in Director of Public Prosecutions for Northern
Ireland v. Lynch [1975] AC 653, 670, it is too late to adopt that
view. And the question now is whether that development should
be carried a step further and applied to a murderer who is the
actual killer, and if the answer to this question is no, whether
there is any basis upon which it can be right to draw a distinction
between a murderer who did the actual killing and a murderer who
played a different part in the design to bring about the death of
the victim.

The first suggestion that the defence of duress might be
available to a person guilty of murder appears to have emerged in
Reg. v. Kray (Ronald) (1969) 53 Cr.App.R. 569, 576-578 in the
judgment of the Court of Appeal delivered by Widgery L.J.

It is instructive to see the circumstances in which it arose:

“We now consider the effect on these two applicants
(the Kray brothers) of the defence put forward by Anthony
Barry. The case against Anthony Barry was that he was an
accessory before the fact to the McVitie murder, and the
Crown relied primarily on his having carried a gun from the
Regency Club to Evering Road, knowing Reginald Kray
intended to use it in the murder of McVitie. Barry
admitted from the outset that he had done this, but pleaded
that he had acted under duress being in fear for the safety
of himself and his family if he failed to carry out the order
of the Krays.

“In support of this defence Barry gave evidence of an
immediate threat made in relation to his unwillingness to
carry the gun and also to prior conduct of the Kray twins
which, he said, had placed him in terror of them. The
immediate threat was said to have been a message
transmitted from Reginald Kray by Hart to the effect that,
if Barry would not take the gun to Evering Road, the Krays
would come back to the Regency Club; a message which
Barry interpreted as meaning that he would get hurt as well
as McVitie. The general background of terror was supported
by evidence of previous threats and acts of violence
committed by the Krays or their henchmen at the Regency
Club and similar acts which had been reported to Barry, but
which he had not seen.

“Barry’s counsel indicated the general nature of his
defence at an early stage in the trial, and counsel for the
other accused were naturally apprehensive lest this should
let in a great deal of otherwise inadmissible evidence
detrimental to the Krays. An attempt was made at the
outset to get a ruling as to the admissibility of this
evidence, but the judge had no information on which to rule
and wisely declined. Barry’s counsel attempted to cross-
examine along these lines when the Crown witnesses to the
Cornell murder were giving evidence, but he made little
progress in the face of objections. Later, however, a
substantial body of evidence of the vicious and violent
reputation of the Kray twins was let in either during cross-
examination of Crown witnesses or in Barry’s own evidence
and that of his witnesses.

– 19 –

“Mr. Platts-Mills contended before us that the whole
of this evidence was inadmissible since Barry had never laid
an adequate foundation for the defence of duress. He
concedes that although duress is not available in murder to
a person charged as a principal it is available to an
accessory, but he submits the accessory must show he had
no alternative and in this case Barry had the alternative of
taking the gun to the police instead of Evering Road.
Accordingly, says Mr. Platts-Mills, Barry never had a viable
defence of duress and the judge should have discharged the
jury as soon as this became apparent.

“Mr. Wrightson takes a broader view and accepts that
Barry had a valid defence if his will was so overcome by
threats and fear that he had no independent choice and
ceased to be an independent actor. He accepts that it was
open to Barry to support this defence by evidence of threats
and violent conduct within his own knowledge, and also by
recounting incidents of which he had heard and the news of
which had affected his mind. Mr. Wrightson’s complaint is
that the evidence admitted went beyond these limits and
related to events early in 1966 which were not proved to
have come to Barry’s notice. He further complains that on
occasions a witness called to prove that Barry had been told
of a particular incident went on to testify to the truth of
that information, which Mr. Wrightson submits was irrelevant
and inadmissible since it did not go to the state of Barry’s
mind.

“It is evident to us that both the learned judge and
counsel for the Crown did their best to confine this
evidence to the limits contended for by Mr. Wrightson, but
this was not always possible. Having examined all such
incidents to which our attention has been drawn, we are
satisfied that they give rise to no miscarriage of justice and
do not render the verdict unsafe or unsatisfactory. We are
further satisfied that Barry had a viable defence on the
basis left to the jury by the learned judge, namely, that by
reason of threats he was so terrified that he ceased to be
an independent actor, and that the evidence of violent
conduct by the Krays which Barry put before the Court was
accordingly relevant and admissible.”

Like the Lord Chancellor, I regard this decision as of little
authority. The decision proceeded on a concession by counsel that
the defence of duress was available to an accessory before the
fact to murder and I have myself doubts about the correctness of
allowing the defence at all if, as it appears, Anthony Barry had
got himself into this predicament by reason of his association with
known violent criminals (see Reg. v. Fitzpatrick [1977] N.I. 20). It
is, however, understandable that the judge would not wish to limit
the full extent of the gangleader’s villainy being laid before the
jury.

Widgery L.J. who had given the judgment in Reg. v. Kray
(Ronald) (1969) 53 Cr. App.R. 569, repeated the same qualification
in Reg. v. Hudson [1971] 2 QB 202, when allowing the defence in
a case of perjury he said: “… it is clearly established that
duress provides a defence in all offences including perjury (except

– 20 –

possibly treason or murder as a principal),” but apart from saying
that the court had been referred to much authority he gave no
reason for limiting the exception to murder as a principal.

In Director of Public Prosecution for Northern Ireland v.
Lynch [1975] A.C. 633, of the majority who held that duress should
be available to an aider and abettor to a murder, in that case the
driver of the getaway car, Lord Morris of Borth-y-Gest said of the
actual killer at p. 671:

“… the person is told that to save his life he himself
must personally there and then take an innocent life. It is
for him to pull the trigger or otherwise personally to do the
act of killing. There, I think, before allowing duress as a
defence it may be that the law will have to call a halt.”

Lord Wilberforce said at p. 685: “I would leave cases of direct
killing by a principal in the first degree to be dealt with as they
arise.” Lord Edmund-Davies, at p. 715, foreshadowed his opinion
in Abbott v. The Queen [1977] AC 755, by citing a passage from
Smith and Hogan Criminal Law, 3rd ed., p. 166:

“The difficulty about adopting a distinction between the
principal and secondary parties as a rule of law is that the
contribution of the secondary party to the death may be no
less significant than that of the principal.”

To illustrate this one only has to point to the case of a “contract”
killing.

Thus it seems to me, my Lords, that even after Lynch the
whole weight of authority denied the defence of duress to the
actual killer. This view had the unanimous support of the Supreme
Court of South Australia in Reg. v. Brown and Merely [1968]
S.A.S.R. 467, when the majority denied the defence of duress to
an aider and abettor to murder and Bray C.J., who dissented on
this issue, nevertheless said at p. 499: “I repeat also that as at
present advised I do not think duress could constitute a defence to
one who actually kills or attempts to kill the victim.”

In Abbott v. The Queen [1977] AC 755, the majority in the
Privy Council applied the law of duress in accordance with English
authority and denied it as a defence to a murderer who took part
in the actual killing. The minority would have extended the
defence even to the actual killer, pointing out the illogicality of
allowing it to the principal in the second degree or the aider and
abettor and denying it to the principal in the first degree.

Since that time the whole question of duress has been
studied by the Law Commission (see Law Commission Report No.
83 on “defences of general application”, dated 27 July 1977). The
report sets out the arguments for and against the defence and
deals in particular with whether it should apply to murder. They
balanced the argument based upon the sanctity of human life that
denies the defence to a murderer against the argument urged by
the majority in Director of Public Prosecutions for Northern
Ireland v. Lynch [1975] AC 653, that the law should not demand
more than human frailty can sustain. They preferred the latter
argument and accordingly recommended that a defence of duress

– 21 –

should be available to all crimes including murder. But in the
draft Bill they annexed to their report they prescribed the defence
in far narrower terms than it had hitherto been defined by the
judges and they introduced conditions which clearly go beyond the
bounds of judicial creativity and would require legislation. It is
worth reminding oneself of the first two clauses:

“1. – (1) The following provisions of this section provide a
defence (referred to below in this Act as ‘the defence of
duress’) in place of the defence of duress at common law
(which is consequently abolished except in relation to
offences committed before the passing of this Act).

1. 1. Subject to section 2 and subsection (5) below, a person
    shall not be guilty of an offence by virtue of any action
    taken by him under duress.
  2. A person shall be regarded for the purposes of this
    section as having taken any action under duress if he was
    induced to take it by any threat of harm to himself or
    another and at the time when he took it he believed
    (whether or not on reasonable grounds) –
2. 1. that the harm threatened was death or serious
    personal injury (physical or mental);
  2. that the threat would be carried out immediately
    if he did not take the action in question or, if not
    immediately, before he could have any real
    opportunity of seeking official protection; and

provided, however, that in all the circumstances of the case
(including what he believed with respect to the matters
mentioned in paragraphs (a) to (c) above and any of his
personal circumstances which are relevant) he could not
reasonably have been expected to resist the threat.

(4) The fact that any official protection which might have
been available in the circumstances would or might not have
been effective to prevent the harm threatened is immaterial
for the purposes of subsection (3)(b) above.

1. 1. The defence of duress does not apply in any case where
    on the occasion in question the defendant was voluntarily
    and without reasonable cause in a situation in which he
    knew he would or might be called upon to commit the
    offence with which he is charged or any offence of the
    same or a similar character under threat of death or serious
    personal injury (whether to himself or to anyone else) if in
    the event he should refuse to do so.
  2. In this section ‘official protection’ means the protection
    of the police, of the authorities governing any prison or
    other custodial institution, or of any other similar authority
    concerned in the maintenance of law and order.

– 22 –

(7) The fact that one party to any action is exempt by
virtue of this section from criminal liability for that action
shall not affect the question whether anyone else is guilty
of an offence by virtue of being a party to that action.

2. – (1) On a trial on indictment the defendant shall not,
without leave of the court, be entitled to rely on the
defence of duress unless he has served on the prosecutor at
least seven clear days before the hearing a notice in writing

(a) indicating his intention to rely on the defence;

(b) giving particulars of the words or conduct
constituting the threat which induced him to take the
action in question; and

(c) giving any information then in his possession to
identify or assist in identifying any persons making
the threat and any persons other than himself on
whom the harm threatened would have been inflicted
if the threat had been carried out.

(2) In any proceedings for an offence it shall be for the
prosecution to prove that the defence of duress does not
apply, but only if there is sufficient evidence to raise an
issue with respect to whether or not it does.”

I can not refrain from commenting that if duress is
introduced as a merciful concession to human frailty it seems hard
to deny it to a man who knows full well that any official
protection he may seek will not be effective to save him from the
threat of death under which he has acted, but such is the effect
of clauses l(3)(b) and 1(4) when read together.

But what, I think, is significant is the fact that although
the report clearly recognised that English law did not extend the
defence of duress to the actual killer and recommended that the
law should be changed, Parliament never acted upon this advice.
The report was laid before Parliament by the Lord Chancellor in
July 1977 but no steps have been taken to introduce a Bill upon
the lines they recommended. This must at least be some
indication that the community at large are not pressing for a
change in the law to remedy a perceived injustice.

Against this background are there any present circumstances
that should impel your Lordships to alter the law that has stood
for so long and to extend the defence of duress to the actual
killer? My Lords, I can think of none. It appears to me that all
present indications point in the opposite direction. We face a
rising tide of violence and terrorism against which the law must
stand firm recognising that its highest duty is to protect the
freedom and lives of those that live under it. The sanctity of
human life lies at the root of this ideal and I would do nothing to
undermine it, be it ever so slight.

On this question your Lordships should, I believe, accord
great weight to the opinion of the Lord Chief Justice who by
virtue of his office and duties is in far closer touch with the

– 23 –

practical application of the criminal law and better able to
evaluate the consequence of a change in the law than those of us
who sit in this House. This is what he had to say in his judgment
in this case [1986] Q.B. 626, 641:

“It is true that to allow the defence to the aider and
abettor but not to the killer may lead to illogicality, as was
pointed out by this Court in Reg. v. Graham (Paul) [1982] 1
W.L.R. 294, where the question in issue in the instant case
was not argued, but that is not to say that any illogicality
should be cured by making duress available to the actual
killer rather by removing it from the aider and abettor.

“Assuming that a change in the law is desirable or
necessary, we may perhaps be permitted to express a view.
The whole matter was dealt with in extenso by Lord Salmon
in his speech in Abbott v. The Queen [1977] AC 755 to
which reference has already been made. He dealt there
with the authorities. It is unnecessary for us in the
circumstances to repeat the citations which he there makes.
It would, moreover, be impertinent for us to try to restate
in different terms the contents of that speech with which
we respectfully agree. Either the law should be left as it
is or the defence of duress should be denied to anyone
charged with murder, whether as a principal in the first
degree or otherwise. It seems to us that it would be a
highly dangerous relaxation in the law to allow a person who
has deliberately killed, maybe a number of innocent people,
to escape conviction and punishment altogether because of a
fear that his own life or those of his family might be in
danger if he did not; particularly so when the defence of
duress is so easy to raise and may be so difficult for the
prosecution to disprove beyond reasonable doubt, the facts
of necessity being as a rule known only to the defendant
himself. That is not to say that duress may not be taken
into account in other ways, for example by the parole
board.

“Even if, contrary to our views, it were otherwise
desirable to extend the defence of duress to the actual
killer, this is surely not the moment to make any such
change, when acts of terrorism are commonplace and
opportunities for mass murder have never been more readily
to hand.”

My Lords, in my view we should accept the advice of the
Lord Chief Justice and the judges who sat with him, and decline
to extend the defence to the actual killer. If the defence is not
available to the killer what justification can there be for extending
it to others who have played their part in the murder. I can, of
course, see that as a matter of commonsense one participant in a
murder may be considered less morally at fault than another. The
youth who hero-worships the gangleader and acts as lookout man
whilst the gang enter a jeweller’s shop and kill the owner in order
to steal is an obvious example. In the eyes of the law they are
all guilty of murder, but justice will be served by requiring those
who did the killing to serve a longer period in prison before being
released on licence than the youth who acted as lookout.
However, it is not difficult to give examples where more moral

– 24 –

fault may be thought to attach to a participant in murder who
was not the actual killer; I have already mentioned the example of
a contract killing, when the murder would never have taken place
if a contract had not been placed to take the life of the victim.
Another example would be an intelligent man goading a
weakminded individual into a killing he would not otherwise
commit.

It is therefore neither rational nor fair to make the defence
dependent upon whether the accused is the actual killer or took
some other part in the murder. I have toyed with the idea that it
might be possible to leave it to the discretion of the trial judge
to decide whether the defence should be available to one who was
not the killer, but I have rejected this as introducing too great a
degree of uncertainty into the availability of the defence. I am
not troubled by some of the extreme examples cited in favour of
allowing the defence to those who are not the killer such as a
woman motorist being highjacked and forced to act as getaway
driver, or a pedestrian being forced to give misleading information
to the police to protect robbery and murder in a shop. The short,
practical answer is that it is inconceivable that such persons would
be prosecuted, they would be called as the principal witnesses for
the prosecution, and if by any chance they were prosecuted.

As I can find no fair and certain basis upon which to
differentiate between participants to a murder and as I am firmly
convinced that the law should not be extended to the killer, I
would depart from the decision of this House in Director of Public
Prosecutions for Northern Ireland v. Lynch [1975] AC 653 and
declare the law to be that duress is not available as a defence to
a charge of murder, or to attempted murder. I add attempted
murder because it is to be remembered that the prosecution have
to prove an even more evil intent to convict of attempted murder
than in actual murder. Attempted murder requires proof of an
intent to kill, whereas in murder it is sufficient to prove an intent

to cause really serious injury.

It can not be right to allow the defence to one who may be
more intent upon taking a life than the murderer. This leaves, of
course, the anomaly that duress is available for the offence of
wounding with intent but not to murder if the victim dies
subsequently. But this flows from the special regard that the law
has for human life, it may not be logical but it is real and has to
be accepted.

I do not think that your Lordships should adopt the
compromise solution of declaring that duress reduces murder to
manslaughter. Where the defence of duress is available it is a
complete excuse. This solution would put the law back to lines
upon which Stephens suggested it should develop by regarding
duress as a form of mitigation. English law has rejected this
solution and it would be yet another anomaly to introduce it for
the crime of murder alone. I would have been more tempted to
go down this road if the death penalty had remained for murder.
But the sentence for murder although mandatory and expressed as
imprisonment for life, is in fact an indefinite sentence, which is
kept constantly under review by the parole board and the Home
Secretary with the assistance of the Lord Chief Justice and the
trial judge. I have confidence that through this machinery the

– 25 –

respective culpability of those involved in a murder case can be
fairly weighed and reflected in the time they are required to serve
in custody.

I have had the advantage of reading the speeches of the
Lord Chancellor and Lord Mackay of Clashfern and I agree with
the opinions they have expressed on the second and third questions
raised before your Lordships.

LORD MACKAY OF CLASHFERN

My Lords,

At the request of all parties to the appeals by Burke, Howe
and Bannister, they were conjoined and have been heard together.
They arise out of two separate cases and the issues in each are
similar. The appeal of Clarkson arises out of one of these cases
and by agreement of the parties to it and the other appeals it has
also been heard with the others.

The material facts in Burke’s case are, that on 9 July 1983
Burke shot dead one Henry Botton at point blank range with a
sawn off shotgun. The victim was due to give evidence at Inner
London Crown Court on 18 July 1983 in a trial in which Burke’s
co-defendant, William Clarkson, was accused with others of
conspiracy to handle stolen goods. The victim was to have given
evidence in support of the defence of duress by Clarkson, which
was raised by one of Clarkson’s co-defendants. Burke’s defence at
his trial was that he had agreed to shoot the victim only because
of his fear that Clarkson would kill him if he did not, but when it
came to the event, the gun actually went off accidentally and the
killing was therefore unintentional and amounted to no more than
manslaughter. At the trial it was submitted on behalf of Burke
that he was entitled to be acquitted completely of murder and
manslaughter by reason of duress and that if this submission failed
he was entitled to be acquitted of murder and found guilty of
manslaughter by reason of accident in an unlawful act. The judge
who presided at the trial, the then Common Sergeant, Judge Tudor
Price, directed the jury that Burke, as the actual killer, was not
entitled to rely upon the defence of duress to the charge of
murder but since he considered there was evidence in support of
the defence of duress he left it to the jury in respect of
manslaughter. He further directed the jury that they could not
convict Clarkson of murder unless they convicted Burke of murder
and that if Burke was guilty of manslaughter then Clarkson could
be convicted, at most, only of manslaughter and that if the
appellant was acquitted on the grounds of duress then Clarkson,
the author of the duress, must be convicted of manslaughter.

Howe and Bannister were indicted together with two other
men, Murray and Bailey, at Manchester Crown Court on two
counts of murder and one of conspiracy to murder contrary to
section 1 of the Criminal Law Act 1977. The particulars of the
first count referred to the murder of Mitchell Elgar on 10 October
1983, the particulars of the second count referred to the murder
of Martin Pollitt on 11 October 1983 and the conspiracy to murder

– 26 –

related to a conspiracy on 12 October 1983 to murder John
Redfern.

In October 1983 Howe and Bailey were aged 19, Bannister
was aged 20 and Murray was 35. Howe had one conviction for a
motoring offence. Bannister had convictions for dishonesty
offences but none for violence. Murray had 25 previous court
appearances including appearances for offences of violence and in
1974 was sentenced to a term of eight years’ imprisonment for
offences of assault with intent to rob and robbery. Bannister met
Murray whilst at Risley Remand Centre. Howe was introduced to
Murray by Bailey and all became acquainted with each other for a
period of four days prior to the first murder and were in each
other’s company almost all the time from the period 6 October to
13 October. Murray was the dominant figure. He was dishonest,
powerful, violent and sadistic. Through acts of actual violence or
threats of violence, Murray gained control of each of the
appellants who became fearful of him. Mitchell Elgar, referred to
in count 1, was killed at a remote spot in the Goyt valley,
Derbyshire. His death was preceded by a savage kicking and
beating, and acts of torture and sexual perversion perpetrated by
Bannister, Howe and Bailey. The coup de grace was executed by
Bailey who strangled Mitchell Elgar with a headlock. Before the
attack, Murray had told Howe and Bannister that Elgar was a
‘grass’ and was to be killed and Bannister was threatened with
violence by Murray if he did not give Mitchell Elgar “a bit of a
battering.” Martin Pollitt referred to in count 2 was picked up by
the four men, to whom I have referred, on 11 October 1983 and
they took him to the same place as that in which Mitchell Elgar
was killed on the previous day. Murray told Howe and Bannister
to kill Pollitt, which they did by strangling him with Bannister’s
shoelace, each holding one end. In relation to count 3, the same
procedure had been followed. However Redfern, the intended
victim, suspected something was afoot and managed to escape
from what otherwise would, have inevitably been his death. Howe
and Bannister each admitted to being parties to the killings and
the conspiracy to kill in the circumstances I have described but
they alleged that they acted in fear of their own lives because of
the conduct of Murray. They alleged that they feared that Murray
would treat them in the same way as Mitchell Elgar had been
treated if they did not comply with his directions.

At the trial, counsel for the prosecution were content to
assent to the proposition that in respect of the murder of Mitchell
Elgar death had been caused by Bailey strangling the victim
although the kicks and punches would have resulted in death
moments later even in the absence of strangulation. On that
basis, the prosecution were content that the judge who presided at
the trial should leave the defence of duress to the jury in respect
of count 1 and also in respect of count 3. On count 2, the judge
rejected the submissions made on behalf of both appellants that
the defence of duress should be allowed.

In directing the jury to consider the plea of duress in
respect of counts 1 and 3, the judge directed the jury that the
test to be applied was whether “the threat was of such gravity
that it might well have caused a reasonable man placed in the
same situation as the defendants to act as the defendants did” and
to pose the question “would a sober person of reasonable firmness

– 27 –

sharing the defendants’ characteristics have responded to the
threats by taking part in the killing.”

Burke was convicted of murder and Howe and Bannister
were convicted on both counts of murder with which they were
charged and also on the count of conspiracy to murder. All
appealed to the Court of Appeal (Criminal Division) [1986] Q.B.
626 where the appeals were heard together and dismissed. In
dismissing the appeals the court certified three points of law of
general public importance were involved in the decisions to dismiss

the appeals, namely:

1. 1. Is duress available as a defence to a person charged
    with murder as a principal in the first degree (the
    actual killer)?
  2. Can the one who incites or procures by duress
    another to kill or to be a party to a killing be
    convicted of murder if that other is acquitted by
    reason of duress?
  3. Does the defence of duress fail if the prosecution
    prove that a person of reasonable firmness sharing the
    characteristics of the defendant would not have given
    way to the threats as did the defendant?”

The Court of Appeal granted leave to appeal against their
decision to this House. Clarkson’s appeal was also heard by the
Court of Appeal (Criminal Division) at the same time as the
others was also dismissed and leave to appeal to this House was
granted. Although he is concerned in the first question already
mentioned as a basis for his concern in the second, which arises
only if Burke is successful on the first, Clarkson’s appeal is
concerned with the second question in respect of which he
contends that if Burke was acquitted by reason of duress he could
not be convicted of murder as one who had incited or procured by
duress Burke to kill or to be a party to a killing.

It will be convenient to deal with the matters arising in
these appeals by reference to the three questions of law which
have been certified and the order in which the questions are posed
by the Court of Appeal.

Question 1

The question whether duress is available as a defence in law
to a person charged with murder as a principal in the first
degree (actual killer) has not been the subject of a previous
decision of this House. The matter received consideration
in this House in Director of Public Prosecutions for
Northern Ireland v. Lynch [1975] AC 653.

Lynch had driven a motor car containing a group of the
I.R.A. in Northern Ireland on an expedition in which they
shot and killed a police officer. He was tried along with
two other men on a count that he murdered the police
constable and was convicted and sentenced to life
imprisonment. This House, by a majority of three to two,
allowed Lynch’s appeal and ordered a new trial pursuant to

– 28 –

section 13 of the Criminal Appeal (Northern Ireland) Act
1968. At the new trial Lynch was allowed to plead the
defence of duress but this defence was rejected by the jury
and Lynch was again convicted.

It was accepted by the majority of the House in Lynch that
at that time the balance of such judicial authority as
existed was against the admission of the defence of duress
in cases of first degree murder. The writers were generally
agreed in saying that the defence was not available in
murder although later writers appear to have said so
following Hale. The references are Hale’s Pleas of the
Crown (1736), vol. 1, pp. 51, 434; East’s Pleas of the Crown
(1803), vol. 1, p. 294; Blackstone, Commentaries on the
Laws of England, 1809 ed., vol. 4, p. 30; Glanville Williams
Criminal Law, 2nd ed. (1961), p. 759, para. 247; Russell on
Crime, 12th ed. (1964), vol. 1, pp. 90-91; Smith and Hogan
Criminal Law, 3rd ed. (1973) pp. 166-167. Since the
fundamental passage is that from Hale at p. 51, I think it is
appropriate to quote it in full:

“If a man be menaced with death, unless he will
commit an act of treason, murder, or robbery, the
fear of death does not excuse him, if he commit the
fact; for the law hath provided a sufficient remedy
against such fears by applying himself to the courts
and officers of justice for a writ or precept de
securitate pacis.

Again, if a man be desperately assaulted, and in peril
of death, and cannot otherwise escape, unless to
satisfy his assailant’s fury he will kill an innocent
person then present, the fear and actual force will
not acquit him of the crime and punishment of
murder, if he commit the fact; for he ought rather to
die himself, than kill an innocent: but if he cannot
otherwise save his own life, the law permits him in
his own defence to kill the assailant; for by the
violence of the assault, and the offence committed
upon him by the assailant himself, the law of nature
and necessity, hath made him his own protector cum
debito moderamine inculpatae tutelae, as shall be
farther shewed, when we come to the chapter of
homicide se defendendo.”

Counsel for the appellants, Burke, Bannister and Howe, in
his very detailed and careful submission accepted this
position as reflecting the law up to the time of Lynch.
Since that time, on this question there has been the decision
of the Privy Council in Abbott v. The Queen [1977] A.C.
755, a majority decision in which the minority consisted of
Lord Wilberforce and Lord Edmund-Davies who, along with
Lord Morris of Borth-y-Gest, had constituted the majority
in the Lynch’s case. Counsel for these appellants submitted
that your Lordships should hold that the reasoning of the
majority in the Lynch case should be applied and extended
to cover the present cases. He recognised that this would
involve a change in the law on this matter but argued that
the change was one which your Lordships should properly

– 29 –

decide to make as the consequence of the decision of the
House in Lynch.

In approaching this matter, I look for guidance to Lord
Reid’s approach to the question of this House making a
change in the prevailing view of the law in Myers v.
Director of Public Prosecutions [1965] A.C. 1001, 1021-1022,
where he said:

“I have never taken a narrow view of the functions of
this House as an appellate tribunal. The common law
must be developed to meet changing economic
conditions and habits of thought, and I would not be
deterred by expressions of opinion in this House in old
cases. But there are limits to what we can or should
do. If we are to extend the law it must be by the
development and application of fundamental principles.
We cannot introduce arbitrary conditions or
limitations: that must be left to legislation. And if
we do in effect change the law, we ought in my
opinion only to do that in cases where our decision
will produce some finality or certainty. If we
disregard technicalities in this case and seek to apply
principle and common sense, there are a number of
other parts of the existing law of hearsay susceptible
of similar treatment, and we shall probably have a
series of appeals in cases where the existing technical
limitations produce an unjust result. If we are to
give a wide interpretation to our judicial functions
questions of policy cannot be wholly excluded, and it
seems to me to be against public policy to produce
uncertainty. The only satisfactory solution is by
legislation following on a wide survey of the whole
field, and I think that such a survey is overdue. A
policy of make do and mend is no longer adequate.
The most powerful argument of those who support the
strict doctrine of precedent is that if it is relaxed
judges will be tempted to encroach on the proper
field of the legislature, and this case to my mind
offers a strong temptation to do that which ought to
be resisted.”

In the present appeal, as I have said, the reason advanced
on behalf of the appellants to allow the defence of duress
to persons in the appellants’ position as the actual killers is
based upon the assertion that this House in Lynch allowed it
to a person who was charged with murder as a principal in
the second degree otherwise described as an aider and
abettor and that there was no relevant distinction between
that case and the case of the actual killer. He submitted
that the reasoning of the majority in the Lynch case when
logically applied to the circumstances of the present case
led to the result that the defence of duress should have
been admitted here and that the appeal should accordingly
be allowed.

Counsel for the Crown submitted that the appeal should be
refused, that the existing law did not allow the defence of
duress to an actual killer or principal in the first degree

– 30 –

and that ii no proper distinction could be made between this
and the Lynch case the House should decline to follow
Lynch because in his submission the reasoning in Lynch was
flawed.

The first question accordingly that arises in this appeal is
whether any distinction can be made between this case and
the Lynch case. It is clear from the speech of Lord Morris
of Borth-y-Gest that he did not regard it as a necessary
consequence of his view that the defence of duress should
be available to a principal in the first degree. At [1975]
A.C. 653, 671, he says:

“The issue in the present case is therefore whether
there is any reason why the defence of duress, which
in respect of a variety of offences has been
recognised as a possible defence, may not also be a
possible defence on a charge of being a principal in
the second degree to murder. I would confine my
decision to that issue. It may be that the law must
deny such a defence to an actual killer, and that the
law will not be irrational if it does so.

“Though it is not possible for the law always to be
worked out on coldly logical lines there may be
manifest factual differences and contrasts between
the situation of an aider and abettor to a killing and
that of the actual killer.”

He goes on to distinguish the case of a person in the
position of an aider and abettor who saves his own life at a
time when the loss of another life is not a certainty with
the position of a person who is told that to save his life he
must himself personally there and then take an innocent
life. There, says Lord Morris at pp. 671-672:

“I think, before allowing duress as a defence it may
be that the law will have to call a halt. May there
still be force in what long ago was said by Hale?
‘Again, if a man be desperately assaulted, and in peril
of death, and cannot otherwise escape, unless to
satisfy his assailant’s fury he will kill an innocent
person then present, the fear and actual force will
not acquit him of the crime and punishment of
murder, if he commit the fact; for he ought rather to
die himself, than kill an innocent.” (see Hale’s Pleas
of the Crown, vol. 1, p. 51).

“Those words have over long periods of time
influenced both thought and writing but I think that
their application may have been unduly extended when
it is assumed that they were intended to cover ail
cases of accessories and aiders and abettors.”

Lord Wilberforce at p. 685 said:

“I would decide that the defence is in law admissible
in a case of aiding and abetting murder, and so in
the present case. I would leave cases of direct

– 31 –

killing by a principal in the first degree to be dealt
with as they arise.”

Lord Edmund-Davies at p. 715, referring to the opinion of
Bray C.J. in Reg. v. Brown and Morley [1968] S.A.S.R. 467,
said:

“His conclusion, at p. 499, was that ‘. . . the trend
of the later cases, general reasoning, and the express
authority of the Privy Council in Sephakela’s case
prevent the acceptance of the simple proposition that
no type of duress can ever afford a defence to any
type of complicity in murder. I repeat also that as
at present advised I do not think duress could
constitute a defence to one who actually kills or
attempts to kill the victim.’

“It appears to me, with respect, that the reliance
placed by Bray C.J. on Sephakela’s case is misplaced,
though I concur when he says, at p. 496, that:

‘there is nothing, in my view, in Sephakela’s case to
prevent us from holding that there can be
circumstances in which duress can be a defence to a
person charged with murder as a principal in the
second degree.’

“Such was the role of Lynch, and this House is
accordingly not now called upon to deal with the
reservation of the Chief Justice in relation to a
person who under duress ‘actually kills or attempts to
kill the victim.’ As to the actual killer, while I
naturally seek to refrain from prejudging future cases,
I think it right to say that I agree with the
observation of Smith and Hogan, Criminal Law, 3rd
ed. p. 166 that; ‘the difficulty about adopting a
distinction between the principal and secondary parties
as a rule of law is that the contribution of the
secondary party to the death may be no less
significant than that of the principal.”

In my opinion, it is plain from these quotations that the
majority of this House in Lynch, and particularly Lord
Morris, were reaching a decision without committing
themselves to the view that the reasoning which they had
used would apply to an actual killer. To take one example,
it would have been impossible to cite Bray C.J. in support
of the proposition that the defence of duress should be
allowed in a charge of murder unless this distinction had
been taken.

While therefore Lynch was decided by reasoning which does
not extend to the present case, the question remains
whether there is a potential distinction between this case
and that of Lynch by which to determine whether or not
the defence of duress should be available. I consider that
Smith and Hogan were perfectly right in the passage cited
from that work by Lord Edmund-Davies to which I have
already referred. I have not been able to find any writer~

– 32 –

of authority that is able to give rational support for the
view that the distinction between principals in the first
degree and those in the second degree is relevant to
determine whether or not duress should be available in a
particular case of murder. Whatever may have divided Lord
Wilberforce and Lord Edmund-Davies on the one hand, from
Lord Simon of Glaisdale and Lord Kilbrandon on the other,
it is apparent that all agree that this is not a distinction
which should receive practical effect in the law.

I believe that the discussions of this matter have shown that
at one extreme, namely that of the person who actually
kills by a deliberate assault on a person who is then
present, there is a fair body of support for the view that
the defence of duress should either not be allowed or that
the practical result will be even if it is allowed that it will
never be established while there is also strong support for
the view that at the other extreme minor participation
which the law regards as sufficient to impute criminal guilt
should be capable of being excused by the defence of
duress. A similar consideration was no doubt present to the
mind of Hume, the eminent writer on the Scottish criminal
law, where in his work Commentaries on the Law of
Scotland respecting Crimes (1829) 53 in relation to the
defence in Scotland known as coercion, after a reference to
the case of James Graham who claimed that he had been
forced by Rob Roy and his gang to take part in an armed
robbery, he says:

“But generally, and with relation to the ordinary
condition of a well-regulated society, where everyman
is under the shield of the law, and has the means of
resorting to that protection, this is at least somewhat
a difficult plea, and can hardly be serviceable in the
case of a trial for any atrocious crime, unless it has
the support of these qualifications: an immediate
danger of death or great bodily harm; an inability to
resist the violence; a backward and inferior part in
the perpetration; and a disclosure of the fact, as well
as restitution of the spoil, on the first safe and
convenient occasion.” (underlining added).

So far, I have not found any satisfactory formulation of a
distinction which would be sufficiently precise to be given
practical effect in law and at the same time differentiate
between levels of culpability so as to produce a satisfactory
demarcation between those accused of murder, who should
be entitled to resort to the defence of duress and those who
were not.

The House is therefore, in my opinion, faced with the
unenviable decision of either departing altogether from the
doctrine that duress is not available in murder or of
departing from the decision of this House in Lynch. While
a variety of minor attacks on the reasoning of the majority
were mounted by counsel for the Crown in the present case,
I do not find any of these sufficiently important to merit
departing from Lynch on these grounds. I do, however,
consider that having regard to the balance of authority on

– 33 –

the question of duress as a defence to murder prior to
Lynch, for this House now to allow the defence of duress
generally in response to a charge of murder would be to
effect an important and substantial change in the law. In
my opinion too, it would involve a departure from the
decision in the famous case of Reg. v. Dudley and Stephens
(1884) 14 QBD 273. The justification for allowing a
defence of duress to a charge of murder is that a defendant
should be excused who killed as the only way of avoiding
death himself or preventing the death of some close relation
such as his own well-loved child. This essentially was the
dilemma which Dudley and Stephens faced and in denying
their defence the court refused to allow this consideration
to be used in a defence to murder. If that refusal was
right in the case of Dudley and Stephens it cannot be wrong
in the present appeals. Although the result of recognising
the defence advanced in that case would be that no crime
was committed and in the case with which we are
concerned that a murder was committed and a particular
individual was not guilty of it (subject to the consideration
of the second certified question) that does not distinguish
the two cases from the point of view now being considered.

To change the law in the manner suggested by counsel for
the appellants in the present case would, in my opinion,
introduce uncertainty over a field of considerable
importance.

So far I have referred to the defence of duress as if it
were a precisely defined concept but it is apparent from the
decisions that it is not so and I cannot do better in this
connection than refer to what Lord Simon of Glaisdale said
on this point in Lynch at p. 686:

“Before turning to examine these considerations, it is
convenient to have a working definition of duress –
even though it is actually an extremely vague and
elusive juristic concept. I take it for present
purposes to denote such [well grounded] fear, produced
by threats, of death or grievous bodily harm [or
unjustified imprisonment] if a certain act is not done,
as overbears the actor’s wish not to perform the act,
and is effective, at the time of the act, in
constraining him to perform it. I am quite uncertain
whether the words which I have put in square
brackets should be included in any such definition. It
is arguable that the test should be purely subjective,
and that it is contrary to principle to require the
fear to be a reasonable one. Moreover, I have
assumed, on the basis of Reg. v. Hudson [1971] 2 Q.B.
202 that threat of future injury may suffice, although
Stephen’s Digest of the Criminal Law art. 10 is to
the contrary. Then the law leaves it also quite
uncertain whether the fear induced by threats must
be of death or grievous bodily harm, or whether
threatened loss of liberty suffices: cases of duress in
the law of contract suggest that duress may extend
to fear of unjustified imprisonment; but the criminal
law returns no clear answer. It also leaves entirely

– 34 –

unanswered whether, to constitute such a general
criminal defence, the threat must be of harm to the
person required to perform the act, or extends to the
immediate family of the actor (and how immediate?),
or to any person. Such questions are not academic,
in these days when hostages are so frequently seized.”

To say that a defence in respect of which so many
questions remain unsettled should be introduced in respect of
the whole field of murder is not to promote certainty in the
law. In this connection it is worth observing that when in
the Law Commission Report No. 83 on “defences of general
application,” the Law Commission recommended that the
defence of duress should be available in murder they
suggested a definition of duress which is, I believe,
considerably narrower than that generally thought to be
available in the present law in respect of other offences.
In particular they required that the defendant must believe
that “the threat will be carried out immediately, or if not
immediately, before he can have any real opportunity of
seeking official protection” and they suggested that the fact
that any official protection which might have been available
in the circumstances would or might not have been effective
to prevent the harm threatened should be immaterial in this
context. It is of interest and importance to notice that this
point figured long before in Hale’s statement which I have
quoted. It is to be noted that it was of this very part of
Hale’s statement that Lord Wilberforce said in Lynch at p.
682:

“Even if this argument was ever realistic, he would
surely have recognised that reconsideration of it must
be required in troubled times.”

I notice that in the Law Commission Report No. 143, dated
28 march 1985, which contains a report to the Law
Commission in respect of the codification of the criminal
law by a team from the Society of Public Teachers of Law,
doubt is expressed on the soundness of this recommendation
in Report No. 83. This particular matter does not arise in
the circumstances of the present case, but the great
difficulty that has been found in obtaining a consensus of
informed opinion upon it is just one illustration of the
uncertain nature of what would be introduced into this most
important area of the criminal law if the defence of duress
were to be available.

Since the decision in Lynch the Law Commission have
published in their Report No. 83, to which I have referred,
the result of an extensive survey of the law relating to
duress and have made recommendations upon it which have
been laid before Parliament. In my opinion, the problems
which have been evident in relation to the law of murder
and the availability of particular defences is not susceptible
of what Lord Reid described as a solution by a policy of
make do and mend. While I appreciate fully the gradual
development that has taken place in the law relating to the
defence of duress I question whether the law has reached a
sufficiently precise definition of that defence to make it

– 35 –

right for us sitting in our judicial capacity to introduce it
as a defence for an actual killer for the first time in the
law of England. Parliament, in its legislative capacity,
although recommended to do so by the report of the Law
Commission, has not taken any steps to make the defence
of duress available generally to a charge of murder even
where it has the power to define with precision the
circumstances in which such a defence would be available.

It has also been suggested for consideration whether, if the
defence of duress is to be allowed in relation to murder by
the actual killer, the defence should have the effect, if
sustained, of reducing the crime to that of manslaughter by
analogy with the defence of provocation. Provocation itself
was introduced into the law by judicial decision in
recognition of human frailty, although it is now the subject
of a statutory provision and it was suggested that the same
approach might be taken now with regard to duress. In this
connection it is worthy of note that in the judgment in S.
v. Goliath S.A.L.R. [1972] (3) 1, 465 in which Rumpff J.A.
examined the question of whether compulsion could
constitute the defence on a charge of murder from which
Lord Wilberforce quoted as a statement of principle a
substantial passage. Rumpff J.A. went on in a later passage
to say at pp. 480-481:

“Whether an acquittal will follow on a charge of
murder because of compulsion, will depend on the
particular circumstances of each case and the whole
factual complex will have to be carefully investigated
and judged with the greatest circumspection. In the
simple case where A kills B just to save his own life,
the strength of the compulsion would be a decisive
factor and the compulsion would have to be so strong
that although not vis absoluta, it would still be
comparable with it, in the sense that the reasonable
man in the particular circumstances would not be able
to withstand it.

For the purposes of replying to the reserved questions
it is unnecessary to determine in what light the
defence of compulsion must be seen i.e. whether it
operates because of the lawfulness of the compelled
act or whether it excludes the full fault.”

And a little later he said, at p. 482:

“I am of the opinion that the second question which
was stated:

[‘that is ‘whether the special defence of
compulsion can ever in law constitute a
complete defence to a charge of murder so as
to entitle an accused to an acquittal?’]

must be answered by a qualified ‘yes,’ in the sense
that a complete defence will depend on the
circumstances of each case.”

– 36 –

From this I take it that Rumpff J.A. was of opinion that
the defence might, depending on the circumstances, either
lead to an acquittal or to a reduction of the charge from
murder to a lower category of unlawful killing.

In my opinion, we would not be justified in the present
state of the law in introducing for the first time into our
law the concept of duress acting to reduce the charge to
one of manslaughter even if there were grounds on which it
might be right to do so. On that aspect of the matter the
Law Commission took the view that where the defence of
duress had been made out it would be unjust to stigmatise
the person accused with a conviction and there is clearly
much force in that view.

The argument for the appellants essentially is that Lynch
having been decided as it was and there being no practical
distinction available between Lynch and the present case
this case should be decided in the same way. The opposite
point of view is that since Lynch was concerned not with
the actual killer but with a person who was made guilty of
his act by the doctrine of accession the correct starting
point for this matter is the case of the actual killer. In
my opinion, this latter is the correct approach. The law
has extended the liability to trial and punishment faced by
the actual killer to those who are participants with him in
the crime and it seems to me, therefore, that where a
question as important as this is in issue the correct starting
point is the case of the actual killer. It seems to me plain
that the reason that it was for so long stated by writers of
authority that the defence of duress was not available in a
charge of murder was because of the supreme importance
that the law afforded to the protection of human life and
that it seemed repugnant that the law should recognise in
any individual in any circumstances, however extreme, the
right to choose that one innocent person should be killed
rather than another. In my opinion, that is the question
which we still must face. Is it right that the law should
confer this right in any circumstances, however extreme?
While I recognise fully the force of the reasoning which
persuaded the majority of this House in Lynch to reach the
decision to which they came in relation to a person not the
actual killer. It does not address directly this question in
relation to the actual killer. I am not persuaded that there
is good reason to alter the answer which Hale gave to this
question. No development of the law or progress in legal
thinking which have taken place since his day have, to my
mind, demonstrated a reason to change this fundamental
answer. In the circumstances which I have narrated of a
report to Parliament from the Law Commission concerned
inter alia with this very question it would seem particularly
inappropriate to make such a change now. For these
reasons, in my opinion, the first certified question should be
answered in the negative.

It follows that, in my opinion, the House should decline to
follow the decision in Lynch. In my opinion, the reasoning
which persuaded this House in Reg. v. Shivpuri [1986] 2
W.L.R. 988 that it was appropriate to reconsider its earlier

– 37 –

decision in Anderton v. Ryan [1985] AC 560 applies equally
to the present case, although the decision in Lynch is of
longer standing than was that of Anderton v. Ryan. Up to
the present time, the courts have been declining to allow an
actual killer to plead the defence of duress while allowing it
to a person charged with murder who was not the actual
killer as is illustrated in the circumstances of these appeals.
The Lord Chief Justice in Reg. v. Graham (Paul) [1982] 1
W.L.R. 294, 297 illustrated how technical and puzzling in
practice the distinction could be. In my opinion, it would
not be right to allow this state of affairs to continue. I
recognise that this decision leaves certain apparent
anomalies in the law but I regard these as consequences of
the fact that murder is a result related crime with a
mandatory penalty. Consequently no distinction is made in
penalty between the various levels of culpability.
Differentiation in treatment once sentence has been
pronounced depends upon action by the Crown advised by
the executive Government although that may be affected by
a recommendation which the court is empowered to make.
Where a person has taken a minor part in a wounding with
intent and is dealt with on that basis he may receive a very
short sentence. If sufficiently soon after that conviction
the victim dies on the same facts with the addition of the
victim’s death caused by the wounding, he may be sentenced
to life imprisonment. This is simply one illustration of the
fact that very different results may follow from a set of
facts together with the death of a victim from what would
follow the same facts if the victim lived.

I turn now to the second certified question. In the view
that I take on question one the second does not properly
arise. However, I am of opinion that the Court of Appeal
reached the correct conclusion upon it as a matter of
principle.

Giving the judgment of the Court of Appeal Lord Lane C.J.,
said [1986] 1 Q.B. 626, 641-642:

“The judge based himself on a decision of this Court
in Reg. v. Richards [1974] Q.B. 776. The facts in
that case were that Mrs. Richards paid two men to
inflict injuries on her husband which she intended
should “put him in hospital for a month.” The two
men wounded the husband but not seriously. They
were acquitted of wounding with intent but convicted
of unlawful wounding. Mrs. Richards herself was
convicted of wounding with intent, the jury plainly,
and not surprisingly, believing that she had the
necessary intent, though the two men had not.

“She appealed against her conviction on the ground
that she could not properly be convicted as accessory
before the fact to a crime more serious than that
committed by the principals in the first degree. The
appeal was allowed and the conviction for unlawful
wounding was substituted. The Court followed a
passage from Hawkins’ Pleas of the Crown, vol. 2. c.
29, para. 15:

– 38 –

‘I take it to be an uncontroverted rule that
[the offence of the accessory can never rise
higher than that of the principal]; it seeming
incongruous and absurd that he who is punished
only as a partaker of the guilt of another,
should be adjudged guilty of a higher crime
than the other.’

“James L.J. delivering the judgment in Reg. v. Richards,
said, at p. 780:

‘If there is only one offence committed, and that is
the offence of unlawful wounding, then the person
who has requested that offence to be committed, or
advised that that offence be committed, cannot be
guilty of a graver offence than that in fact which
was committed.’

“The decision in Reg. v. Richards has been the subject of
some criticism – see for example Smith and Hogan, Criminal
Law, 5th ed. (1983), p. 140).

“Counsel before us posed the situation where A hands a gun
to D informing him that it is loaded with blank ammunition
only and telling him to go and scare X by discharging it.
The ammunition is in fact live, as A knows, and X is killed.
D is convicted only of manslaughter, as he might be on
those facts. It would seem absurd that A should thereby
escape conviction for murder.

“We take the view that Reg. v. Richards [1974] Q.B. 776
was incorrectly decided, but it seems to us that it cannot
properly be distinguished from the instant case.”

I consider that the reasoning of the Lord Chief Justice is
entirely correct and I would affirm his view that where a
person has been killed and that result is the result intended
by another participant, the mere fact that the actual killer
may be convicted only of the reduced charge of
manslaughter for some reason special to himself does not, in
my opinion in any way, result in a compulsory reduction for
the other participant.

I turn now to question three. On this question the learned
Lord Chief Justice said [1986] Q.B. 626, 642-643:

“Finally we turn to the second ground of appeal in
the case of Howe and Bannister, namely, that the
judge was wrong in directing the jury that there is an
‘objective’ element in the defence of duress. The
judge directed the jury on this point as follows: The
test is whether the threat was of such gravity that it
might well have caused a reasonable man placed in
the same situation as the defendant to act as the
defendant did.’ And a little later on the judge put it
in this way: ‘Would a sober person of reasonable
firmness sharing the defendant’s characteristics have
responded to the threats by taking part in the killing

– 39 –

“It seems to us that this direction was in accordance
with the judgment of this court in Reg. v. Graham
(Paul) [1982] 1 WLR 294. Consequently this ground

of appeal likewise fails.”

In Graham, Lord Lane C.J. giving the judgment of the Court
of Appeal (Criminal Division) said at p. 300:

“As a matter of public policy, it seems to us
essential to limit the defence of duress by means of
an objective criterion formulated in the terms of
reasonableness. Consistency of approach in defences
to criminal liability is obviously desirable.
Provocation and duress are analogous. In provocation
the words or actions of one person break the self-
control of another. In duress the words or actions of
one person break the will of another. The law
requires a defendant to have the self-control
reasonably expected of the ordinary citizen in his
situation. It should likewise require him to have the
steadfastness reasonably to be expected of the
ordinary citizen in his situation. So too with self-
defence, in which the law permits the use of no more
force than is reasonable in the circumstances. And,
in general, if a mistake is to excuse what would
otherwise be criminal, the mistake must be a
reasonable one.

It follows that we accept Mr. Sherrard’s submission
that the direction in this case was too favourable to
the appellant. The Crown having conceded that the
issue of duress was open to the appellant and was
raised on the evidence, the correct approach on the
facts of this case would have been as follows. (1)
Was the defendant, or may he have been, impelled to
act as he did because, as a result of what he
reasonably believed King had said or done, he had
good cause to fear that if he did not so act King
would kill him or (if this is to be added) cause him
serious physical injury? (2) If so, have the
prosecution made the jury sure that a sober person of
reasonable firmness, sharing the characteristics of the
defendant, would not have responded to whatever he
reasonably believed King said or did by taking part in
the killing? The fact that a defendant’s will to resist
has been eroded by the voluntary consumption of
drink or drugs or both is not relevant to the test.”

In my opinion, what the Lord Chief Justice said in the
present case and in Graham was entirely correct. In my
opinion, this question also falls to be answered “yes.”

I, therefore, consider that these appeals should be dismissed,
the first certified question answered in the negative and the
second and third in the affirmative.

– 40 –

Source: https://www.bailii.org/