A response to “The Double Effect Effect” by Charles Foster, Jonathan Herring, Karen Melham, and Tony Hope (CQ20(1))
Introduction
The doctrine of double effect and its associated intention/foresight distinction continues to be the subject of vigorous debate. In a recent article published earlier this year in the Cambridge Quarterly of Healthcare Ethics, Charles Foster, Jonathan Herring, Karen Melham, and Tony Hope have claimed that the doctrine has only limited application in the law, and that, given its limited application, academics have spent a disproportionate amount of effort debating its merits. The authors discuss the case law at length and show, convincingly in my view, that there is, in fact, little judicial support for the doctrine, at least as that doctrine has been traditionally understood in its Catholic version and in various modified forms of the Catholic version. The authors then go on to make a number of criticisms of the doctrine, on the basis of which they conclude that the doctrine is possibly unworkable.Footnote 1 However, their criticisms of the doctrine focus predominantly on what they concede to be “the core requirement of double effect,”Footnote 2 namely, the intention/foresight distinction, and they draw attention to some of the notorious difficulties in finding a satisfactory way of applying the distinction. It is principally, though not exclusively, on the basis of these difficulties that they suggest that the doctrine might “be kept safely penned in the philosopher’s lecture room”Footnote 3 rather than being “[a]llowed in the courtroom.”Footnote 4
In this article, I focus only on their criticisms of this aspect of the doctrine, namely, on the distinction between intention and foresight on which the doctrine relies, and argue that the authors are mistaken in concluding that this distinction may not be sufficiently workable so as to be allowed in the courtroom. They base this conclusion on the suggestion that, if there are two consequences of an action, the jury might not be able to decide which was intended and which was foreseen.Footnote 5 The difficulty with the claim is simply that the intention/foresight distinction is relevant to the definition of murder, for, at the present time, the mental element of the crime of murder remains intent. It is therefore impossible to recommend that the doctrine of double effect be kept out of the courtroom, if, as they contend, that recommendation is made on the basis that the intention/foresight distinction is difficult to apply. Yet the rejoinder, which they anticipate,Footnote 6 that the jury members have to make a decision about intention and foresight every day in the Crown Court is not itself an adequate answer to their contention. We need to show that there is a rational way by which the jury can arrive at a decision. There may, indeed, be different reasons for recommending that the other elements of the doctrine should be kept outside the courtroom, but the core requirement of the intention/foresight distinction is a fundamental part of the law. It is therefore essential to solve the problems the authors consider to beset the intention/foresight distinction, in order to ensure the integrity of the law of murder, for there will always be cases in which it will be necessary for a jury to determine if the requisite element of intent is satisfied. My argument will be that the distinction is not as unclear as the authors suggest.
Identifying the Appropriate Description of the Action by Reference to the Intended Consequence
The main criticism made by the authors is that intention and foresight are not always easy to distinguish. They quote from John Finnis, who has stated that whether a state of affairs is intended or merely foreseen can depend on how the agent's conduct is described.Footnote 7 The authors note that it follows from such a view that a headhunter may describe his cutting off an unfortunate person’s head as merely “adding to his collection” rather than “killing.”Footnote 8 If we respond that, at least “in our world,”Footnote 9 decapitation leads to deaths and so in our world it is killing, then why isn’t administering a lethal dose of painkillers also killing?Footnote 10
We begin with the claim that the way a person describes his or her activities can affect the legal assessment of that person’s intention,Footnote 11 and then, having explained why this is not so, we examine their claim that, if we say that headhunting is killing, we would have to describe a lethal dose of painkillers as killing as well on the very same grounds.
The claim that how a person describes his or her activities can affect the assessment of that person’s intention is not new. It was perhaps most elegantly made by John Harris, in his book The Value of Life:
There is an immense, and I think insoluble, problem about deciding when a death (or any other consequence) is aimed at as an end or a means and when it is merely a side effect. . . . It is a feature of any action that its description is almost infinitely expandable or contractible. My crooking my finger, is my pulling the trigger, is my shooting at Samson, is my hitting Delilah, is my killing the president, is my orphaning her children, etc etc. The problem is to set a limit to the legitimacy of characterising the action one way rather than another.Footnote 12
In support of this claim, Harris cites the admittedly unfortunate example given by the Linacre Centre of the potholer stuck in a cave with people behind him and water rising to drown them. In one case, the man can be blown up to remove him from the exit that he is blocking. In another, a rock can be moved to open another escape route, but the man will be crushed as a consequence. The traditional analysis has been that, in the first case, the man’s death is intended, whereas, in the second, it is merely a foreseen side effect of the need to remove the rock and escape. Harris rightly regards such an analysis “as an example of the most comprehensive sophistry.”Footnote 13 However, this does not show that there is, as he claims, an immense and insoluble problem about deciding when a death or any other consequence is aimed at as an end or a means, and when it is merely a (foreseen) side effect (indeed, his very claim that the Linacre Centre’s analysis is an example of the most comprehensive sophistry surely rules out such a claim). For reasons we shall see, death is intended in both cases, and so the Linacre Centre’s purported illustration of the distinction is wrong. We will return to this, and the related abortion examples, later, but first some general remarks must be made.
Harris’s claim that there is an immense and insoluble problem about deciding when death is aimed at as an end or means, and when it is merely a side effect, is made on the basis that it is a feature of any action that its description is almost infinitely expandable or contractible. However, neither of these claims is justified. The question is whether the agent can credibly claim that his or her intent was merely to x rather than y. Thus, characterizing one’s conduct in one way rather than another so as to rig one’s account of what is intended as end or means, on the one hand, and what is merely a foreseen side effect of one’s conduct, on the other, will not necessarily prevent a court from finding that the putative side effect was in fact an intended consequence and so was not a side effect at all. Indeed, in many cases, whether or not conduct counts as one action rather than another depends on whether an outcome was intended, rather than vice versa. No more is this so than in the case of murder, for an element of murder remains intent in England and Wales. A defendant may deny that his or her conduct amounts to murder, but whether such a defense is sustainable—and therefore credible—will depend on ascertaining the relevant intention. If that intention is indeed found to be present, the recharacterization of one’s conduct so as to avoid that conclusion will fail. For this reason, Harris is mistaken to suggest that the infinitely expandable or contractible description of any action presents an insoluble problem about deciding when death is intended (as end or means) and when it is merely foreseen.
We can begin to see the hurdles that might face any such defendant by examining a famous example of the putative difficulty of distinguishing means, on the one hand, from side effects, on the other. Consider a bomber who bombs a passenger jet in order to recover on the insurance.Footnote 14 In this case, the bomber’s goal is to recover on the insurance. But are the deaths of the passengers the meansFootnote 15 to the end of recovering on the insurance, or are they a mere side effect of it? The bomber will of course describe his conduct in terms that make it appear plausible to say both that he merely intends to recover on the insurance (his end) and that the passenger deaths are not a means to his doing so but are merely a side effect. Yet as we shall see, such a description will not prevent the conclusion that the passenger deaths remain nonetheless intended at least as a means to recovery.
Lord Goff, writing extrajudicially,Footnote 16 was inclined to see this example as a severe case of reckless indifference to human life that should nonetheless count as satisfying the mental element of murder, a position that, as he himself conceded, would require a modification to the mental element of the crime of murder (that, at present, being an intent either to kill or to cause grievous bodily harm resulting in death). In other words, Lord Goff accepted that we are inclined to regard the conduct here as amounting to murder yet also considered that the deaths of the passengers could not be intended. It is true, of course, that the deaths of the passengers are not the ultimate end of the bomber, but the question is whether they are the means by which the bomber pursues his end, in the sense that he cannot achieve his end without killing some passengers if he chooses this way of making money.
Of course, the fact that he may not be able to achieve the end without some deaths does not of itself suffice to show that causing the passengers’ deaths is the means of recovering on the insurance. For there are many things that cannot be done without other consequences also eventuating, yet we regard those consequences as side effects. For example, I may not be able to become drunk without suffering a hangover, and a visit to the dentist will result in discomfort, yet neither the hangover nor the discomfort are means by which I become drunk or receive a filling. On the other hand, as Lord Goff concedes, we baulk at the suggestion that the passenger deaths in Williams’s example are a side effect of the action of recovering on the insurance. Why is it that we are inclined to think that death is intended so that the bomber can be charged with murder?
The answer to this question brings us to the heart of the issue and is the key to solving the riddle posed by the so-called hard cases. The problem is solved by taking proper account of the nature of the foreseen consequence concerned. It is the gravity of the foreseen consequence that is all-important in distinguishing these cases. Because we regard the foreseen consequence of some conduct as so serious, and because we therefore expect anyone engaging in conduct leading to that consequence to refrain from the conduct, we do not tolerate the description of such consequences in terms that trivialize them, as would be the case if we referred to the deaths of the passengers as a side effect. Lives are not the kind of thing that one can meaningfully trade for lesser evils such as recovering on the insurance fraudulently, without intentionally doing so.Footnote 17 If it were otherwise, it would be meaningful for me to cure someone of their migraine by killing them. But because the gravity of killing them far outweighs the gravity of the migraine, we do not regard it as credible for someone to say that he or she merely intended to cure the migraine, given that death is a much worse fate than that of suffering a migraine. Anyone who genuinely insisted that this was his or her intention does not commit a mistake about intention and foresight. Similarly, with the authors’ example of decapitation, it is not credible for the perpetrator to say that he merely intended to add to his collection.
The same point applies to the different case of the bomber who plants a bomb in a building, certain that it will explode when the bomb disposal squad attempts to diffuse it. Antony Duff believes that, in this case, even though the bomber is certain the squad will be working on the bomb at the time it explodes, the bomber may nonetheless only intend to endanger life or cause alarm.Footnote 18 This is mistaken. Neither possibility—that he intends to endanger life or to cause alarm—is credible. In the case of causing alarm, the gravity of the consequence of causing the deaths of the bomb disposal squad completely outstrips the gravity of the end of causing alarm. If the bomber knows that it will explode when they are working to diffuse it and does nothing, his intention becomes an intention to kill the disposal squad as a means (even if not the means) of causing alarm. And in the case of an intention to endanger life, it is simply not logical to say that one intends to endanger life by killing someone—it is not logically coherent to describing killing as a form of merely endangering life.
In short, the gravity of the consequence means that there is a transition from quantity (the amount of severity, say) to quality (the kind of action he can be described as engaging in) in the way we characterize that consequence—it is such a severe or catastrophic consequence that engaging in an act that leads to such a consequence, when there is a choice about whether or not to engage in it, means that the consequence feeds back into the character of the act such that we call such conduct “killing passengers” or “killing the bomb disposal unit” and do not leave any room for the distinction between an unintended side effect and some other aimed-for consequence.Footnote 19 An important point often overlooked in debates about the bomber examples is that our concepts and the distinctions we draw have not fallen from the sky but are given their meanings by us, relative to our purposes of classification.Footnote 20 We refuse to call some consequences a side effect because of their gravity and of our correlative expectations that people would not engage in such conduct when they know with certainty that it will lead to such grave consequences.Footnote 21
But does my claim here not amount merely to the claim that the gravity of the consequence makes us deem the consequence to be intended? No. If, for example, instead of having the effect of making a patient’s hair fall out, administering chemotherapy were lethal, we would expect doctors to refrain from engaging in the conduct, just as we would expect someone to refrain from using a gun. The gravity of the consequence outweighs any gain, and this factor impinges on what it is the people in question can be meaningfully said to be aiming for. Similarly, the bomber chooses to recover on the insurance or to cause alarm at the expense of people’s lives, not at the expense of a scratch on the paintwork of the plane or the building. The gravity of the consequence he is prepared to accept for the sake of his goal, together with the fact that he has a choice about whether to engage in the conduct, having foreseen that grave consequence, is what makes this case different from a case in which the only result is a scratch in the paintwork, and this is what rules out a distinction between an intended versus unintended outcome in these cases.
The authors might now objectFootnote 22 as follows: this analysis could be brought to apply to palliative pain relief, thereby making it indistinguishable from euthanasia. For in palliative care, ex hypothesi, the patient’s life is abbreviated as a result of providing the relief. The patient thereby dies. Does this consequence therefore not outweigh the benefit of giving the relief on the criterion of the gravity of the consequence?Footnote 23 But it seems to me that pain relief of terminally ill patients desperate for the relief is not a case in which the gravity of the consequence means that there is a transition from quantity to quality in the same way as in the bomber example. The patient is not only terminally ill but very close to the end. The finality of his or her situation is accepted by all. It is accepted that good medical practice is to administer relief in such circumstances, and, when a patient may be desperate for the relief, there is an expectation on the doctor to provide that relief. It seems to me quite reasonable to view death in this case differently from death of innocent passengers in an aircraft.
This response might, however, be thought to raise a fresh objection: couldn’t the same be said of cases of euthanasia, cases in which death is intended? Isn’t the patient in such cases sufficiently close to the end that the benefit of relief outweighs death? But it should be noted that the reason death is regarded as intended in euthanasia has to do with not the gravity of the consequence but simply the fact that no objective can meaningfully be aimed for other than the death of the patient. In this case, death is the end or purpose of the conduct and not a meansFootnote 24 or a side effect, because there is no meaningful sense in which the patient is relieved from his or her pain and suffering—nobody exists to experience the relief, and “relief” is par excellence an experience noun. Killing someone is not a means of ending their pain in the way that giving pain-relieving drugs is. It would be misleading to say that we are ending their pain when, in reality, we are killing the patient. That would be like saying we cured someone’s migraine by killing them. We do not just end their pain, but we kill them, that is, it is their life that we end. It is true of course that they no longer experience pain. But that is because they no longer experience anything—nobody exists at that point to experience anything. We can no more single out an inability to experience pain than anything else. For example, we end their ability to feel an itch, and more generally to move, speak, and sleep, as well as to experience pain. In short, we simply end their life. By contrast, in palliative care, the pain relief is literally experienced, and the drugs are used for the express purpose of relieving the patient’s pain, which is their main function. The pain-relieving properties in the drug enable the doctor to achieve the aim of relieving pain independently of the patient’s death, whereas, in euthanasia, the relevant properties of the poison are its powers to bring about death.
Applying the Analysis to the Hard Cases and My Test of Credibility
Let us now apply the analysis to the hard cases. Previously we mentioned the Linacre Centre’s example of a potholer who becomes stuck in a hole in a cave and who must either be blown apart in order for those trapped behind him to escape from the cave or be indirectly crushed when the party trapped removes a rock from an alternative escape route. As noted, traditionally, in the former case his death is regarded as intended, but in the latter it is regarded as merely foreseen. It is this analysis that Harris rightly labeled as “an example of the most comprehensive sophistry.” We are now in a position to see why this is so. The gravity of the consequence in both cases—the potholer’s death—is the same. It therefore matters little whether that consequence is brought about by a stick of dynamite or by removing a rock whose removal will result in the potholer’s foreseeably certain death. True, the consequence being weighed against his death—the deaths of the others in the cave—is equally serious, but this does not mean that the potholer’s death is any less grave. Compare the consequence of the rock being moved—the death of the potholer—with a different consequence—the death of some lice on the rock. In the case of the lice, it is unlikely that the cave explorers will even turn their minds to the possible deaths of the lice. And this is because the consequence is regarded as too trivial to warrant even an expectation that they should turn their minds to such a consequence and to weigh it up against their need to escape. The situation is, of course, different in the case of the potholer. The true position is that the potholer’s death is intended in each case,Footnote 25 but that the potholer’s death may be justified on the basis of an extraordinary emergency.Footnote 26
The same analysis applies to the famous hysterectomy case in which a mother needs a hysterectomy while she is pregnant. Again, traditionally the analysis has been that, by contrast with the craniotomy case, the fetus’s death is a mere side effect of the aim to remove her womb.Footnote 27 The fetus is moved (resulting in its death) merely as a consequence of having to remove the womb. Yet the fact that, in our world, the current state of technology does not allow us to entertain, as a genuine possibility, the idea that the baby might survive outside the womb has an impact on how we choose to describe the doctor’s conduct. Given that the timing of the need to remove the womb corresponds to the timing of the pregnancy, it is both the womb and the fetus that need to be removed. To focus on the womb at the expense of the fetus is to leave out an important circumstantial consideration that changes the stakes involved. The true position is that the fetus’s life will be terminated in order to save the mother. Assuming, for the sake of argument, that the fetus has personhood rights,Footnote 28 the consequence for the fetus is going to be grave, and so the fetus’s death must be justified, once again on the alternative basis of an extraordinary emergency.
Similarly, in the conjoined twin case of Jodie and Mary,Footnote 29 the suggestion that Mary’s death was a mere side effect of the intention to save Jodie was rightly rejected by the majority. Such a suggestion would rightly be considered by Harris as another example of the most comprehensive sophistry. Even if it could be accepted that killing Mary was “indirect”—a description that could easily be contested—such a conclusion would be irrelevant, because it is possible of course to intend outcomes indirectly. The true position is that Mary was being sacrificed for the sake of Jodie, because, as the judges themselves pointed out, the consequences for Mary would be grave—her death.
The gravity of the consequence, then, is the crucial factor, and whether such a consequence could ever be merely foreseen is controlled by what I would call a “test of credibility.” Neither the headhunter case, the potholer case, or the hysterectomy case satisfies this test, as we have seen. To see its relevance, consider the well-known example of the strategic bomber during a war who wants to bomb a munitions factory, knowing that innocent civilians will be killed, in order to shorten the war.Footnote 30 This case is contrasted with the terror bomber who targets civilians on purpose to lower enemy morale.Footnote 31 The notion of “target” here is doing a lot of the work. The picture we might adopt in the case of the strategic bomber is of the civilian houses surrounding the infrastructure targeted, and it seems to make sense on such a picture to consider the bomber as targeting the factory, knowing that civilians may be killed in the vicinity, but not as actually targeting the civilians, hence the concept of collateral damage. But if we change the example such that the civilians become human shields placed inside the munitions factory, intuitions change. The notion of collateral damage here, when it is not possible to target the factory without also targeting the civilians, seems no longer to be credible, even though we accept that the ultimate end may only be to disable the infrastructure. True, unlike the terror bomber, there’s still a desire to limit casualties, but for that very reason there may be an expectation here that the conduct not be undertaken. Compare an archer in the style of 1970s TV programs showing off his skills by firing arrows at the target, close to which a woman is standing. Because it is possible to hit the target without hitting the woman, it is credible to believe that the bull’s-eye only is targeted. Place the woman in front of the target, and it is no longer credible that he can aim to hit the target at all. Assume the arrow is strong enough and the bow is powerful enough for the arrow to travel through her to the target. Still, in such a case he can no longer hit the target without killing the woman. But if we think this is because killing the woman is now the means, it is difficult to see how the case differs in substance from the human shields in the munitions factory, and to overcome the problem it suffices to imagine placing the woman behind the arrow board instead, where it is “foreseen” that the arrow will travel through the target board and hit the woman behind. Here, there is no credibility in regarding the archer in such a case as intending only to hit the target board, with hitting the woman as a mere side or concomitant effect of that aim, because of the gravity of the consequence—injury to her that could be fatal.
Conclusion
The aim of this article has been to take up some of the challenges laid down by the authors in their article. The conceptual issues they have raised about the utility and workability of the intention/foresight distinction for medical law are important and required a detailed response. I have attempted to provide that response here in order to demonstrate the continued importance of the distinction, particularly in the context of the law of murder in England and Wales, where the mental element remains intent. In particular, I have tried to show that the main reason for the difficulty in deciding whether death is intended or merely foreseen for the purposes of the law of murder has been a failure to appreciate that the gravity of the consequence can itself feed back into the character of the act being performed such that any act producing that consequence in which the consequence is actually foreseen will be regarded as an act that intends to produce that consequence. It is this observation that explains why someone who claims not to have intended that consequence may not be considered credible, and why I have developed the test of credibility as a means of deciding this question in the hard cases. Far fewer cases than those envisaged under the double effect doctrine are cases in which intention is absent. Of the cases discussed in this article, only the case of pain relief (in which it is foreseen that death will be hastened) is a case in which death will not be intended. This means that some of the reasons advanced by the authors for their claim that the intention/foresight distinction “is not as clear-cut as may be thought”Footnote 32 do not in fact support that contention and that they have arguably exaggerated the difficulties in applying it. By providing arguments for the continued importance of the intention/foresight distinction here, I hope to have advanced the debate.
