1. Rachels on active and passive euthanasia

So far, we’ve been treating death as one of the most serious harms someone can suffer. But unfortunately, as you know, it sometimes happens that death would come as a benefit to someone. This is the case when that person’s quality of life is sufficiently low and there is no reasonable prospect of its improving. In such cases, euthanasia may be contemplated. Euthanasia may be defined as the intentional hastening of death for someone for whom death would be a benefit. (The word “euthanasia” derives from the ancient Greek for “dying well.”)

A distinction may be drawn between so-called active and passive euthanasia. This distinction is grounded in the distinction between killing and letting die. Active euthanasia involves hastening someone’s death by means of killing that person, whereas passive euthanasia involves hastening someone’s death by means of letting him die (rather than prolonging his life by means of initiating or continuing some form of medical intervention).

Rachels notes that the distinction between active and passive euthanasia is regarded by many people as being morally significant, so much so that active euthanasia is often taken to be overall morally wrong under all circumstances, whereas passive euthanasia is regarded as something that can be overall morally justified under certain circumstances. He takes this common view to be endorsed in a policy statement issued by the American Medical Association in 1973 (see p. 44, left column). It is a view that Rachels opposes.

Rachels begins by pointing out that in some cases in which death would come as a benefit to someone who is suffering, killing that person would often be a better way of putting an end to his suffering than would letting him die. Consider, for example, a patient who has incurable cancer of the throat and is in great pain, pain that cannot be satisfactorily alleviated. A lethal injection would be much more merciful than simply letting the patient’s life slip away slowly and agonizingly. (See p. 44, both columns.) So why in such a case would anyone think that we should abide by the common view that permits passive but forbids active euthanasia? The only possible reason that Rachels can imagine is that there is something about killing, about its very nature, that makes it so bad in and of itself that letting die is always to be preferred, regardless of whether killing might have better consequences than letting die. Indeed, this seems to be a very popular idea (as we have already seen in our discussions of abortion and famine relief). But it is an idea that Rachels believes has no merit.

To test whether there is such an intrinsic moral disparity between killing and letting die, Rachels proposes the philosophical analogue of a controlled experiment in science. He suggests that we consider a pair of cases that are exactly alike, except that one of them involves killing and the other involves letting die. Motives, circumstances, consequences are all to be held constant; the only admissible variable is the difference between killing and letting die. The point of considering such a pair of cases is this. If we find that we reach a different moral verdict in one case from the verdict we reach in the other, that can only be attributable to the sole difference between the cases, that is, to the difference between killing and letting die; in which case, we must conclude that this difference makes a difference, morally. However, if we find that we reach the same verdict in each case, then we must conclude that the bare difference between killing and letting die makes no difference, morally; in which case, the popular idea about the moral significance of the distinction between killing and letting die must be rejected, and hence so, too, must the common view about the distinction between active and passive euthanasia.

The cases that Rachels asks us to consider are these (see p. 45, right column). In one case, Smith stands to gain a large inheritance if his 6-year-old cousin dies. In order to speed things up, he sneaks into his cousin’s bathroom while the child is taking a bath, drowns him, and arranges things so that it looks like the child died accidentally. In the second case, Jones also stands to gain a large inheritance if his 6-year-old cousin dies. In order to speed things up, he sneaks into his cousin’s bathroom while the child is taking a bath but finds that the child is already face down in the water. He stands by, ready to push the child’s head back down should the need arise, but the need doesn’t arise and the child drowns “accidentally,” “all on his own.” (All these drowning children! First Singer, now Rachels.) Rachels asks: is Smith’s behavior any worse, from the moral point view, than Jones’s? His answer: it is not. His conclusion: killing is therefore no worse in itself than letting die, and so active euthanasia is also no worse in itself than passive euthanasia. This “bare-difference” argument, part of which is of Type II, may be put as follows:

It is important not to misconstrue Rachels’s conclusions. Two points in particular should be noted.

First, Rachels is not saying that there is no difference between killing and letting die, or between active euthanasia and passive euthanasia. On the contrary, he acknowledges these differences. His claim is that these differences make no difference, morally.

Second, Rachels is not saying that it never matters, morally, whether we perform euthanasia by active means or passive. On the contrary, this can matter very much. Consider the patient with throat cancer mentioned above. In this case, on the assumption that no relevant details have been omitted, Rachels would presumably say that the choice is clear: choose active euthanasia rather than passive. Why? Not because active euthanasia is active, but rather because, in this instance, active euthanasia would be more merciful. Or consider another case in which a patient requests passive euthanasia but adamantly refuses active euthanasia. Then, again on the assumption that no relevant details have been omitted, Rachels would presumably say that the choice here is also clear: choose passive euthanasia rather than active. Why? Not because passive euthanasia is passive, but rather because, in this instance, passive euthanasia would be in keeping with the patient’s request. Rachels’s claim is that the bare difference between killing and letting die, and so too the bare difference between active and passive euthanasia, have no moral significance. This is perfectly compatible with saying that, because of some contingent circumstance that happens to obtain in some instance, one is to be preferred to the other on that occasion.

***Here is a good place to stop and consider Question 9 on the Practice Questions for Module 6.***

2. Critical evaluation of Rachels’s article

Let’s turn now to a consideration of some critical comments on Rachels’s article.

Nesbitt:

In his article, Nesbitt indicates that he accepts Rachels’s claim (in premise (2) of Arg. 1) that the Smith case is morally just like the Jones case but rejects the claim (in premise (1)) that these two cases are exactly alike except that one involves killing someone and the other involves merely letting someone die (see p. 49, both columns). His reason for saying this is that Rachels stipulates that Jones is prepared, should the need arise, to kill his cousin (actually, Nesbitt has changed the relationship: for some reason, he calls the child Jones’s nephew), and this, Nesbitt seems to think, means that Jones does not merely let the child die. He grants that Jones is just as reprehensible as Smith, but that’s precisely because Jones is prepared to do what Smith did. To test whether Jones would have been as reprehensible as Smith if he (Jones) had merely let the child die, he turns to a different case (see p. 51, left column) in which Jones is not prepared to kill the child should it happen that the child does not drown “all on his own.” And his claim is that, in this case, Jones, though certainly reprehensible, is not as reprehensible as Smith, and that therefore Rachels is mistaken in inferring that killing is no worse than letting die.

Perrett:

Nesbitt’s discussion is badly confused, for two reasons, both of which are pointed out by Perrett (see p. 52, right column – p. 53, left column).

First, Nesbitt fails to distinguish between two different kinds of moral evaluation: the evaluation of acts or behavior, on the one hand, and the evaluation of agents or persons on the other. (This is the same mistake as the one that I accused Slote of making in our discussion of famine relief.) Rachels is concerned with the question whether the kind of behavior that constitutes killing someone is in itself any worse than the kind of behavior that constitutes letting someone die. Yet Nesbitt focuses on the question whether Jones, a person, is as reprehensible as Smith, another person. (As I explained in my comments on Slote, these two kinds of evaluation can fall apart. A good person sometimes does bad things, and a bad person sometimes does good things. For example, a compassionate person might botch the rescue of a drowning child while a selfish person, spurred on by the promise of a large reward, succeeds in carrying out the rescue.) In this way, Nesbitt simply misses the point of Rachels’s discussion.

Secondly, and perhaps even more importantly, Nesbitt also misses the point of the kind of thought-experiment that Rachels conducts. As I said earlier, the idea is to set up an analogue of a controlled experiment in science, in which everything is held constant except for the variable that is being tested. In the cases that he presents, Rachels sticks to this formula by stipulating that Smith and Jones are mentally just alike. They are both motivated by greed and are both prepared to kill their respective cousins; it’s just their behavior that varies, in that Smith kills his cousin whereas Jones doesn’t kill his cousin but rather lets him die. When Nesbitt substitutes his own Jones-case for Rachels’s Jones-case, he deliberately departs from this formula by stipulating that Jones is not prepared to do what Smith did. But this serves only to contaminate the experiment, since now we cannot be certain whether any difference in the verdict that we might reach in each case is attributable to the difference in behavior or to the difference in mental state.

***Here is a good place to stop and consider Questions 10 and 11 on the Practice Questions for Module 6.***

So says Perrett in defense of Rachels against Nesbitt’s criticism, and he (Perrett) is absolutely right. This doesn’t mean, though, that everything Rachels says is beyond reproach. On the contrary, there are a number of difficulties that we need to address.

First difficulty:

First, perhaps the most obvious point to raise concerns the relevance of Rachels’s discussion to the question whether there is any moral significance to the distinction between active and passive euthanasia. As I noted at the outset, euthanasia consists in the intentional hastening of death for someone for whom death would be a benefit. Yet the pair of cases involving Smith and Jones by means of which Rachels proposes to test the significance of the distinction between active and passive euthanasia are cases in which death is a harm to the child who dies. This is decidedly odd.

Suppose that for some reason Rachels’s argument fails and that, contrary to what he says, we should declare Smith’s behavior worse than Jones’s after all. This would of course be in keeping with the popular view, which Rachels attacks, that killing is worse in itself than letting die. Surely, though, this popular view is attractive (if it is attractive at all) only in cases in which death is a harm. The underlying idea seems to be this: causing a harm is worse than simply allowing a harm to occur. Suppose there is indeed such an asymmetry between causing and allowing. What, then, should we expect in cases in which it is a benefit rather than a harm that is at issue? Presumably this: causing a benefit is better than simply allowing a benefit to occur. Indeed, this idea has some appeal. Consider yet another drowning case. Black and White are on a pier and see someone struggling in the water. Black throws out a life-preserver with a rope attached. The drowning person grabs hold of the life-preserver and Black pulls him to the pier. All the while, White (who has a sharp knife with him) could prevent the rescue by cutting the rope, but he doesn’t. Now, Black caused the rescue, whereas White simply allowed it to occur. Whose behavior was better? Arguably, Black’s. In any case, we certainly have no reason whatsoever to declare Black’s behavior worse than White’s.

My point is simply this: if active euthanasia involves causing a benefit, while passive euthanasia involves allowing a benefit to occur, then, if the causing-allowing distinction has any moral significance in itself, we should expect active euthanasia to be better in itself than passive. We certainly have no reason to declare the former worse than the latter. The common view that active euthanasia is overall morally wrong under all circumstances, whereas passive euthanasia can be overall morally justified under certain circumstances, would seem therefore not to have any plausibility to begin with! The best way to attack it, I believe, is simply by pointing out this fact, rather than by way of a consideration of cases in which death is a harm to the one who dies.

Second difficulty:

It might seem that, in arguing for the point I’ve just made, I have overlooked the fact (one upon which I have relied in our previous discussions of abortion and famine relief) that a right to life is, more particularly, a right not to be killed and is not a right not to be allowed to die. Thus active euthanasia of someone involves violating that person’s right to life, whereas passive euthanasia does not. That is why the former is morally beyond the pale, whereas the latter is not.

But there’s an obvious problem with this retort. To see just what it is, we need to make another distinction regarding euthanasia, one that cuts across the passive-active distinction on which we’ve been focusing. This new distinction is the distinction between voluntary and involuntary euthanasia. (To say that the distinctions “cut across” one another is to say that they “subdivide” in the following way: active euthanasia can be either voluntary or involuntary, and so too can passive.) Voluntary euthanasia is euthanasia that is performed at the autonomous request of the person who is euthanized; involuntary euthanasia is euthanasia that is not performed at the autonomous request of the person who is euthanized. Here it is important to note the following points.

First, in stipulating that the request in question must be autonomous, I am allowing for the possibility that sometimes a person might make a non-autonomous request for euthanasia. Recall from our discussion of prostitution that to say that someone makes an autonomous decision is to say, roughly, that he (a) makes the decision on the basis of an adequate understanding of its probable consequences, (b) intends that these consequences occur, and (c) is free from undue constraints (such as being pressured into making the decision against his will). Clearly, it would be problematic to honor a request for euthanasia that fails to satisfy one or more of conditions (a)-(c). But, just as clearly, an autonomous request for active euthanasia has great moral significance. In particular, it constitutes a waiver of one’s right to life. In such a case, the moral barrier against killing that is constituted by a right to life has been eliminated. That is why there is an obvious problem with the retort that we are now considering. (Some might balk at accepting this position on religious grounds, holding that a right to life cannot be waived by the one who holds it, since it is not up to the right-holder but up to God when one’s life on Earth comes to an end. I won’t pursue this point here.)

Secondly, there are cases in which a person may not currently be able to make an autonomous request for active euthanasia (because he no longer has the capacity to do so) but did make such a request in the past regarding the sort of circumstances he now finds himself in. Such a prior request introduces difficulties (Did he really have an adequate understanding then of the sort of circumstances he is in now? Does his prior request pertain to precisely these circumstances? And so on), difficulties that may make it hard to tell whether the request should be honored, but at least in principle it would seem that such a request should have the same moral weight as a current request.

Third, we must guard against thinking of involuntary euthanasia as euthanasia that is performed against the wishes of the person euthanized. That would of course be grotesque, although the term “involuntary,” which is commonly used in this context, may seem to suggest it (perhaps a better term would be “non-voluntary”). No, involuntary euthanasia is euthanasia that is performed in cases in which the person in question cannot make an autonomous decision either for or against euthanasia (and has not made such a decision in the past). In such a case, of course, we cannot say that the person has waived his right to life, and so here, it might seem, we would have good reason to opt for passive rather than active euthanasia, if a decision in favor of euthanasia is warranted at all. But is this really so? Let’s turn our attention for a moment from humans to nonhuman animals. Perhaps you have had a pet whose life became intolerable. What did you do? Perhaps you took your pet to the vet and asked him to put an end to your pet’s misery. If your vet agreed to do so, he surely did not say, “Yes, sad to say, Rover’s time is up. So, stop feeding him, and don’t give him anything to drink. He’ll die within a few days, and that will be that.” Who would even think of taking such a course? (Think again of Rachels’s example of the patient with throat cancer.) No, in all likelihood, your vet gave your pet a lethal injection, thereby ensuring a quick, humane death. Now, of course, your pet did not, because he could not, make an autonomous request for such treatment. Still, that was obviously the right way to treat him. But if that was the right way to treat your pet, why would it not also be the right way to treat at least some human beings who cannot make an autonomous request for active euthanasia but whose lives are equally as miserable as your pet’s and for whom a quick, humane death is the best way to put an end to their misery?

Third difficulty:

I have been pretending that the distinction between killing and letting die (and so, too, the distinction between active and passive euthanasia) is clear: the former involves causing a death, whereas the latter involves allowing a death. But the fact is that the matter is not at all straightforward.

Consider two patients, Sam and Sheila, each of whom has two doctors. Dr. A gives Sam a lethal injection; Dr. B could immediately counteract this by giving Sam an antidote, but he does not do this. As a result, Sam of course dies. It’s clear, I think, that Dr. A killed Sam, whereas Dr. B let Sam die. Turn, now, to Sheila, who is on life support. Dr. C “pulls the plug” (a terrible phrase − almost as bad as “collateral damage”); Dr. D could immediately counteract this by putting the plug back in, but he does not do this. As a result, Sheila also dies, of course. In summary:

Dr. A:     Gives Sam a lethal injection.
Dr. B:     Refrains ​from giving Sam an antidote.

Dr. C:     Pulls the plug on Sheila.
Dr. D:     Refrains from replacing the plug.

It’s clear that, like Dr. B, Dr. D let his patient die. How, though, should we describe Dr. C’s behavior? Doesn’t it seem as if he behaved in a way that is very similar to the way in which Dr. A behaved? Now, Dr. A killed Sam. Shouldn’t we therefore say that Dr. C likewise killed Sheila? And yet pulling the plug is commonly taken to be a paradigm of passive euthanasia, of letting die! What are we to make of this?

I’m not sure. Perhaps it can be argued that Dr. C didn’t kill Sheila but only let her die because what really killed her was the underlying condition, whatever it was, from which she was suffering. This condition threatened her life, which is why she was put on life support in the first place. She was therefore already in the process of her dying. This process was interrupted when she was put on life support. By pulling the plug, Dr. C merely let the process resume. Well, perhaps so. But think of another case. Sally has a heart condition that threatens her life. She decides to undergo surgery. During surgery, she is hooked up to all sorts of machines that keep her alive. Dr. E pulls the plug on one of these machines and, as a result, Sally dies. Did Dr. E merely let Sally die from her underlying condition, or did he kill her? I bet your answer is: he killed her. But if Dr. E killed Sally, why then didn’t Dr. C kill Sheila after all?

However in the end we choose to describe the behavior of Drs. A-E, I think we must agree that there’s a clear difference between withholding treatment, as Drs. B and D did, and withdrawing treatment, as Dr. C did. Whereas the former is clearly passive (“keeping hands off”), the latter just as clearly involves considerable activity (“putting hands on”). Even if we decide in the end that, although Dr. A killed Sam, Dr. C didn’t kill Sheila but rather let her die, still the sort of active letting die in which Dr. C engaged is quite distinct from the passive letting die in which Drs. B and D engaged. We thus have a three-fold distinction: killing, active letting die, and passive letting die. That being the case, it becomes even less clear why anyone would think that the distinction between euthanasia by means, say, of a lethal injection and euthanasia by means of pulling the plug should in itself be morally significant.

Fourth difficulty:

As I reported above, Rachels believes that the AMA policy statement that he cites endorses the common view that active euthanasia is overall morally wrong under all circumstances, whereas passive euthanasia can be overall morally justified under certain circumstances. This is perhaps not an unreasonable interpretation of that statement, but I think that it is nonetheless erroneous.

In commenting on the AMA’s statement, Rachels complains that, after identifying the crucial issue as that of “the intentional termination of the life of one human being by another,” the statement then mistakenly denies that “the cessation of the employment of extraordinary means to prolong the life of the body when there is irrefutable evidence that biological death is imminent” constitutes the intentional termination of life (see p. 46, both columns). But Rachels’s diagnosis of a mistake here is too hasty. As I see it, the AMA intends to condemn all euthanasia, whether active or passive. Its condoning the cessation of the employment of extraordinary means to keep the patient alive is intended to be consistent with such a blanket condemnation of euthanasia. Perhaps I can best explain this point by reference to a famous case, the case of Karen Ann Quinlan.

Karen Ann Quinlan

In 1975, when she was 21 years old, Karen took some prescription drugs, went to a party, had a few drinks, and collapsed. She stopped breathing twice. Paramedics took her to a hospital, where she lapsed into a persistent vegetative state and was put on a ventilator. (“Vegetative”! Yet another gruesome term.) Her parents requested that she be taken off life support, but the hospital refused to comply with the request. A celebrated legal battle ensued. In 1976, Joseph Quinlan, Karen’s father, was appointed as her legal guardian. He repeated his request that Karen be removed from the ventilator. Given his legal authority to make such a request, the hospital now complied. Karen defied all expectations and continued to live in a vegetative state for several more years, eventually dying from pneumonia in 1985.

It might seem that what Joseph Quinlan was requesting all along was that Karen receive passive euthanasia, but I suspect that that is not so. In order to explain his request that Karen be taken off life support, Joseph reportedly said that he didn’t want Karen to die; what he wanted was that she should be in a “natural” state, not hooked up to a ventilator. To see just what’s at issue, consider the following table:

In this table, there are three possible states for the patient, who is currently hooked up to a life support machine, to be in: alive on the machine, alive off the machine, and dead. There are six basic rankings, A-F, of these possibilities, where “1” represents what is most preferred, “2” what is next preferred, and “3” what is least preferred. Thus in case A, it is deemed best that the patient be alive on the machine and worst that she be dead; second best would be her being alive off the machine. And so on for the remaining cases B-F. Suppose, now, that there are three options regarding how to treat the patient: (a) do nothing (i.e., leave her as she is on the machine), in which case she will in all probability continue to live in her present condition; (b) give her a lethal injection, in which case she will certainly die; or (c) pull the plug (i.e., take her off the machine), in which case she will very likely die, although there is some possibility that she will continue to live. Which option should be chosen? The answer depends on which case we’re dealing with.

Consider case A. Here it is clear what option to choose (all else being equal − that is, on the assumption that no relevant details have been overlooked): (a). Why? Because she is already in the most preferred state (being alive on the machine), and choosing either of the other options would remove her from this state. So, too, in case B: the obvious option is (a).

In cases D and F, however, the obvious option is (b). (Again, I am assuming that “all else” is being held “equal.”) Why? Because this certainly would lead to the patient’s being in the most preferred state (death), whereas choosing (a) would almost certainly not achieve this, while choosing (c), though likely to achieve it, would not be guaranteed to do so.

Cases C and E are different, though. Here the most preferred state is that the patient be alive off the machine. What option to choose in case C is not at all clear. The decision would have to depend on how much worse death would be than continued life on the machine and on how likely death would come about as a result of pulling the plug; all that is clear is that (b) should not be chosen. In case E, however, it is clear what option to choose: (c), i.e., pull the plug. Why? Because taking this option, though likely to lead to the patient’s death (which in this case would be preferable to her remaining in her current state), is not guaranteed to do so; on the contrary, it leaves at least some chance of her attaining an even better state (being alive off the machine), and no other option can manage this.

Two key points are to be noted here. First, I believe that Joseph Quinlan took Karen’s situation to be that of case E. Secondly, and most importantly, pulling the plug in such a case does not amount to an attempt to perform euthanasia, passive or active. Why? Because euthanasia involves the intentional termination of life, whereas in case E such termination, though certainly expected, is not intended. (We addressed the distinction between intention and foresight in our discussion of killing in time of war. You would be well advised to take another look at it, including especially the discussion of the Principle of Double Effect. That principle is endorsed by the Catholic Church ─ and, as it happens, Joseph Quinlan was a practicing Catholic.)

***Here is a good place to stop and consider Questions 12, 13, and 14 on the Practice Questions for Module 6.***

3. The law

Finally, and very briefly, it’s worth noting that, even if we should accept that euthanasia is in principle morally justifiable, and even if we should also accept, as Rachels claims, that the distinction between active and passive euthanasia in itself has no moral significance, it does not follow that the law should permit active euthanasia or, indeed, any form of euthanasia. Whether it should do so depends on a number of factors, including whether permitting some form of euthanasia would have unintended consequences that are so bad that it would be better not to permit it at all, even if prohibiting it would itself have undesirable consequences (in particular, the consequence of forcing people to remain alive contrary to their most fervent wishes). What consequences of legalizing euthanasia could be so bad as to warrant its continued prohibition? Some people fear that permitting euthanasia would usher in a “culture of death” in which respect for life would be eroded. Others fear that scarce medical resources would be rationed in such a way that they would be denied to people who are deemed to be “good candidates” for euthanasia, even if these people have no wish to die. Still others have raised yet other concerns. Such worries are surely legitimate. Whether there is a satisfactory way of putting them to rest is a complicated matter that deserves careful consideration.

What do you think?

***Here is a good place to stop and consider Question 15 on the Practice Questions for Module 6.***