Therefore, J (or someone acting on her behalf) would be justified when acting in order to negate that
threat, even though a consequence of so doing was that the person who was the source of that threat
would die.
It is, of course, odd to think of this as a case of self-defence, since M can hardly be described as an
unlawful aggressor. But as Ward LJ rightly observes, there is no requirement in self-defence that the
attack be a criminal offence:
"The six year old boy indiscriminately shooting all and sundry in the school playground is
not acting unlawfully for he is too young for his acts to be so classified.... [H]owever, ...
in law killing that six-year old in self-defence or others would be fully justified and the
killing would not be unlawful. I see no difference in essence between that resort to
legitimate self-defence and the doctors coming to Jodie's defence and removing the threat
of fatal harm to her presented by Mary's draining her life blood". (At 1017.)
In something of a belt and braces approach, Ward LJ also noted that if one weighs up the respective
best interests of J and M, the scales were tipped heavily in J's favour because she was the only child
with any prospect of life extending beyond the following few months. However, this point matters, if at
all, only when the case is viewed as one of necessity. It is irrelevant to self-defence: if P attacks D
when D has only one hour to live and P is healthy, D is still entitled, if necessary, to kill P in
self-defence. His Lordship seems to think the point important because it resolves the dilemma created
by the conflict in the doctors' legal duties to act in the best interests of each of J and M:
"What are the doctors to do if the law imposes upon them a duty which they cannot
perform without being in breach of Mary's right to life if at the same time the respecting of
her right puts them in breach of the equally serious duty of respecting Jodie's right to
life?... In those circumstances it seems to me that the law must allow an escape through
choosing the lesser of two evils." (At 1016.)
Here Ward LJ helps himself to a second line of argument, that the surgeons may also have a defence of
necessity. This possibility is developed more fully by Brooke LJ, whose judgment is endorsed by Ward
LJ as a 'masterly analysis' (at 1013).
Bringing the case within necessity: Brooke LJ's judgment
In Brooke LJ's view, the case falls entirely within the defence of necessity, since it satisfies the
following three criteria, espoused originally by Stephen:
1.The act is needed to avoid inevitable and irreparable evil;
2.No more should be done than is reasonably necessary for the purpose to be achieved; and
3.The evil inflicted must not be disproportionate to the evil avoided. (At 1052)
These criteria reflect a characterisation of necessity as a lesser-evils defence: "The claim is that [D's]
conduct was not harmful because on a choice of two evils the choice of avoiding the greater harm was
justified." (At 1048.) It should immediately be remarked that this is not a comprehensive
characterisation. As was argued in Simester and Sullivan [pp. 629-33], there is no unitary rationale of
the necessity defence. It is capable of covering a range of justificatory reasons. Moreover, while the
lesser evils rationale does seem to apply here, the criteria set out above by Brooke LJ are insufficient by
themselves to distinguish Re A (which involved killing one to save one) from Dudley and Stephens
(which involved killing one to save three). More needs to be said.
To this need, Brooke LJ responds that "Mary is, sadly, self-designated for a very early death." (At 1051;
see at 1041-2.) Hence, unlike the case of Dudley and Stephens, there was no question of human
choice in selecting the candidate for death. Moreover, the balance of evils is tilted by the fact that "the
principles of modern family law point irresistibly to the conclusion that the interests of Jodie must be
preferred to the conflicting interests of Mary" (at 1052), since Jodie had a good prospect of living a
happy, fulfilled life and Mary had no prospect of life at all. His Lordship explicitly eschews reliance on
the proposition that M was an 'unjust aggressor': "None of the formulations of the doctrine of necessity
which I have noted in this judgment make any such requirement: in this respect [necessity] is different
from the doctrine of private defence." (At 1051.) Hence his Lordship's analysis of the case differs
sharply from that of Ward LJ. (Incidentally, it is an arduous task to reconcile this difference with Ward
and Brooke LJJ's assertions at 1011 and 1018 that they agree with each others' judgments.)
The account given by Brooke LJ is attractive but incomplete. As was stated in the last-but-one
paragraph, necessity cloaks a variety of different rationales, and the requirements of the ordinary
lesser-evils defence, as identified by his Lordship, need to be augmented before they can deal with the
sort of situation in Re A. Let us suppose, as was said to be true in Dudley and Stephens, that the cabin
boy was about to die in any event: that he was, in the language used by his Lordship, "self-designated
for death". Surely, nonetheless, the defendants would not be entitled to kill him? Likewise, doctors are
not free to accelerate the death of a terminally ill patient, V, merely in order to be able to transplant one
of V's organs into P, an otherwise healthy patient.
If one chooses to call Re A a case of necessity rather than self-defence, there are two important
features that must be relied upon to distinguish it from Dudley and Stephens. First, despite Brooke LJ's
refusal to rely on this fact, it matters that M was the source of the threat to J's life. This was not true of
either Dudley and Stephens or the transplant example given in the last paragraph; but it was true of the
unfortunate young man on the Herald of Free Enterprise. Secondly, even though M's death was
foreseen as a virtual certainty (and therefore intended, by virtue of the definition of intention in Woollin
[1999] 1 AC 82; see S&S p. 120-3), Re A was arguably not a core case of direct intention. Recall that
the criminal law acknowledges two varieties of intention [S&S p. 115].
D intends the actus reus if:
1.D directly intended the actus reus in the ordinary, paradigm sense of "intention" - i.e. if he acts
with the aim, object, or purpose of bringing the actus reus about; or
2.D recognised that the actus reus was a virtually certain consequence of his actions.
In Dudley and Stephens, the cabin boy's death was intended in the core or direct sense: the defendants
aimed to kill him, in order then to eat him. In Re A, M's death was no part of the doctors' aim or
purpose, but was at least a virtually certain consequence of what they set out to achieve. (We return to
this issue below.) It is only by supplementing Brooke LJ's analysis with these distinctions that the rule in
Dudley and Stephens can safely be evaded.
Lack of a criminal intent? Robert Walker LJ's judgment
A distinction between these two varieties of intention is at the heart of the Thomist double effect
doctrine, which however ordinarily requires that the beneficial purpose of the treatment (e.g. to alleviate
pain) be directed at the same person who suffers the unwanted side-effect (e.g. an acceleration of
death). Although, formally, the double effect doctrine is inapplicable to Re A, in which two patients are
involved rather than one, the underlying distinction upon which that doctrine rests is essential, it is
submitted, to support Brooke LJ's analysis of the case as one of necessity. Moreover, the same
distinction is explicitly relied upon by Robert Walker LJ (see at 1063). His Lordship combines this
distinction with a finding that surgery would be in the best interests of M as well as J (cf. 1063e); hence
the operation 'would not be unlawful. It would involve the positive act of invasive surgery and M's death
would be foreseen as an inevitable consequence of an operation which was intended, and was
necessary, to save J's life. But M's death would not be the purpose or intention of the surgery, and she
would die because tragically her body, on its own, was not and never had been viable.' (At 1070.) Thus
Robert Walker LJ's analysis relies on a variant of the double effect doctrine.
Both Ward and Brooke LJJ reject the conclusion that the operation was in M's best interests, and that
conclusion will not be discussed here. What is of greater concern is the assertion by his Lordship that
"the doctrine of double effect prevents the doctor's foresight of accelerated death from counting as a
guilty intention." (At 1063.) This sort of reasoning appeared in Gillick v. West Norfolk and Wisbech
Area Health Authority[1986] AC 112, 190, where Lord Scarman once said that "the bona fide exercise
by a doctor of his clinical judgement must be a complete negation of the guilty mind" [See S&S pp.
199-200, 334.] It has been criticised trenchantly and rightly by Ashworth ('Criminal Liability in a Medical
Context: The Treatment of Good Intentions' in Simester and Smith, Harm and Culpability (1996) 173).
Robert Walker LJ's (and Lord Scarman's) reasoning depends on a proposition that there is a difference
between the law's definition of 'intention' and its definition of "guilty intention". No such difference
exists. Either D fulfilled the definition of intention set out in Woollin or he did not. Re A, Gillick, and
any other criminal case, should be approached by asking whether D's conduct constitutes the actus
reus and mens rea of an offence; then by asking separately whether any defence is available. If D acts
in a situation of necessity, that conclusion is a matter of defences and does not mean that D in some
way lacked a "guilty" intent.
What sort of intent to kill?
One final difficulty. It is clear that, as a matter of mens rea, the doctors intended to kill M. The case
falls well within the definition of intention in murder, as laid down in Woollin [1999] 1 AC 82. At the
same time, it is arguable that the judges in Re A wrongly took it for granted that this is a case of
foresight of virtual certainty. M's death seems too close, too intimately bound-up, with the intended
operation to separate her from J, for that death plausibly to be characterised merely as a side-effect,
even a virtually certain side-effect. [Compare the case of the spelaeologists, discussed in S&S at p
120-1.]
Since the law acknowledges only two categories of intention, this suggests that M's death is better
treated as something directly intended by the surgeons who operate. However, that conclusion does not
seem right either. M's death was neither sought for its own sake, nor sought as a means to an end. Her
death supplied no part of the reasons why the doctors were operating; it was not an aim, object, or
purpose of the operation; the doctors would not have regarded themselves as having "failed" in any
sense if, by a miracle, M had survived. [See, on these characteristics of direct intent, S&S p 115-6.]
This suggests that, apart from core cases of direct intention (i.e. means and ends), and cases of virtual
certainty (i.e. virtually certain side-effects), there is a further subcategory of intention which applies to
inseparable accompanying effects. M's death was an inseparable accompanying effect of the
intended operation to separate her from J. [For further discussion of this category, see Simester, Moral
Certainty and the Boundaries of Intention (1996) 16 OJLS 445.]
Why might all this matter? It does not affect the issue of mens rea, because inseparable accompanying
effects lie well inside the boundaries set down in Woollin. But it may matter to the availability of a
defence like necessity. If death is merely a foreseen certain side-effect, there is room for distinguishing
the case from Dudley and Stephens. Similarly, the double effect doctrine in Thomist theology depends
for its application on the difference between a doctor who directly seeks to kill, as a means to end pain,
and a doctor who seeks to end pain knowing that death will be accelerated as a side-effect. But if death
is an inseparable accompanying effect, and not merely a side-effect, it is unclear whether this
distinction, upon which the necessity defence in Re A rests, can be drawn so easily. Unfortunately, their
Lordships failed even to notice this difficulty.
Conclusions
Future criminal cases will find little material with which to generalise in Re A. Robert Walker LJ's
judgment can largely be disregarded, and the analyses of Ward and Brooke LJJ tread different paths.
Indeed, their Lordships' mutual declarations of agreement are undermined by the reasoning in their
judgments. No ratio decidendae emerges with clarity from the decision. Nonetheless, authoritative
dicta may be drawn upon to support arguments about the scope of self-defence (in Ward LJ's judgment)
and especially necessity (in Brooke LJ's judgment). And one may be confident in future that a defence
to murder will be available to D in situations where a blameless victim is, by her conduct, posing an
unjustified threat to the lives of others, at least provided the victim's death is not directly sought and is
only a virtually certain side-effect of the life-preserving actions taken by D.