11. Famine relief

1. Singer on famine relief

So far, when discussing death, we’ve been almost exclusively concerned with the ethics of killing other human beings. But we can control whether people live or die not only by means of killing them (or not) but also by means of letting them die (or not). We typically think that it is much worse to kill someone than to let someone die. Right now, for example, you are letting many people die. You could be doing something else (like writing a check to a reputable charity such as Oxfam), something that, if you did it, you would thereby save some person or perhaps several people from death. But you’re not doing that. You’re reading this instead! Why? And why isn’t what you’re doing just as bad as killing someone, since the end result (death) is the same in either case?


In his article, Singer challenges our common attitude toward letting die. He challenges our complacency. He claims that letting die is far worse than we typically take it to be. We tend to think that, although we are obligated not to kill another person, we are not similarly obligated to save someone from death. On the contrary, saving someone is something we usually think of as lying “above and beyond the call of duty”; it is, to use a fancy term, supererogatory (which just means: above and beyond the call of duty).

Notice two key features of an act that is supererogatory. First, it is not overall morally obligatory. Why? Because it goes “beyond” what we are required or obligated to do. Second, it is not overall morally wrong. Why? Because it lies “above,” rather than “below,” what we ought to do. Putting both these features together, we can say that such an act is morally optional; morally, it’s up to us whether to do it or not.

Singer does not deny that some acts may be supererogatory, but he certainly does deny that many acts (such as donating to famine relief) that we normally think of as supererogatory are in fact so. He believes that they are instead obligatory − they are wrong not to do. Thus, they are not morally optional.

In 1971, when Singer wrote his article, there were around 9 million refugees in East Bengal (Bangladesh), displaced by civil war, in grave danger of dying from starvation. In the article, Singer notes that the British government had up to that point donated a little less than £15 million (at the time worth roughly $36 million) toward helping these people. This might sound quite generous, but Singer also notes that by then the government had spent almost 20 times that much developing the Concorde (the supersonic jet that carried those few people who could afford it from New York to London or Paris in about 3-3½; hours, instead of the usual 7-8 hours). He finds this disparity in expenditure obscene. How can we justify spending so much on relatively frivolous projects while at the same time letting 9 million people starve to death?

Singer is not content just to make claims about our obligations to others. He argues for his position (see p. 538, left column). He does so by asking you to imagine that you have gone out for a stroll in the countryside (I am embellishing the story a little) and that, as you pass by a shallow pond, you see a small child face down in the water, in grave danger of drowning. It’s clear, he thinks (and thinks you’ll think), that you ought to wade into the pond and save the child. Doing so is not supererogatory; it’s something that morality requires of you. But then, what is the relevant difference between this situation and your current situation, one in which you know that there are people in distant lands in grave danger of starvation? None. Thus, saving them from starvation is not supererogatory after all; on the contrary, it’s something that morality requires of you.

Singer is challenging us to defend discriminating between the drowning child and the starving stranger(s), and he is arguing that this challenge cannot be satisfactorily met. He is in effect employing what is by now a familiar kind of argument, an argument of Type III (an argument from inference to the best explanation). He asks what principle underlies the judgment that you ought to save the drowning child. His answer may be put as follows:

M1: It is overall morally wrong under any circumstances to let something bad happen when it is in our power to prevent that thing from happening without thereby sacrificing anything of comparable moral significance.

And his argument may now be put as follows:

Arg. 1: (1) In the pond scenario, letting the child die is overall morally wrong.
(2) The best explanation of premise (1) is moral principle M1.
(3) If premise (2) is true, then M1 is true.
(4) M1 is true.
(5) If M1 is true, then our (affluent people’s) letting others starve to death is, under present circumstances, overall morally wrong.
(6) Our letting others starve to death is, under present circumstances, overall morally wrong.

Singer recognizes that some people may resist his argument, especially since taking the conclusion seriously would seem to require us to undertake a drastic and radical revision of our current, self-centered lifestyles. One objection (see p. 138, both columns) is that you have an obligation to help the drowning child in part because he happens to be nearby, whereas the same is of course not true of the distant stranger who is starving. Singer quickly dismisses this, embracing what might be called a “principle of geographical neutrality.” He acknowledges that proximity might make a difference, in that it might make it easier to judge who needs help and easier to provide the help needed, but he denies that distance in itself is a relevant feature. If (as is increasingly becoming the case in our “global village”) it is just as easy to provide effective aid to someone far away as to someone nearby, then proximity provides no good reason to choose to help the latter rather than the former.

Another concern that Singer addresses has to do with the fact that, whereas the pond scenario involves just one person who can give aid (you) and one person who needs it (the child), in the case of famine relief there are very many people on both “sides” of the issue. Some people appear to think that, if there are many people who can give aid but don’t, then the fact that you in particular can and don’t has no great moral significance − you’re just one of many. But he quickly scotches that idea. Would you somehow have done a less serious wrong, or perhaps no wrong at all, if you had not been the only one to pass the drowning child by − if there had been a dozen others, say, who could have helped but who, like you, didn’t? “One has only to ask this question,” Singer says, “to see the absurdity of the view that numbers lessen obligation” (p. 539, left column).

What of the idea that we only have to do the “part” that we would have to do if everyone else did his “part”? Suppose, for example, that the Bengal emergency would be solved if everyone gave just £5 to the relief fund. Does it then follow that all I must do is give £5 to this fund? Unfortunately not, says Singer. That I would be obligated only to give £5 under the circumstances in which everyone gives £5 does not imply that I am obligated only to give £5 under present circumstances, circumstances in which relatively few people donate money to the fund.

What about the fact that the pond scenario is a “one-off,” in the sense that it happens only once, whereas the current situation is one in which, as soon as one person has been saved from starvation, there is another to take his place (or, perhaps, the same person one day later)? Singer doesn’t address this issue directly (although on pp. 542-43 he does make a related point about population control), but we can imagine what he might say. Just consider a variation on the pond scenario, one in which the emergency that confronts you is repeated a number of times. On Monday, while out walking, you come across a child in a pond, whom you duly rescue. Then the same thing happens on Tuesday, and Wednesday, and Thursday… What are you going to say on Saturday? “Too bad, kid, but I take weekends off”? Surely not. Surely it would be just as wrong to pass the child by on Saturday as on the first day.

***Here is a good place to stop and consider Questions 1 and 2 on the Practice Questions for Module 6.***

2. Critical evaluation of Singer’s article

Singer’s article has been criticized by both Arthur and Slote.

Arthur’s critique:

Arthur argues that the principle on which Singer relies (principle M1) has absurdly demanding consequences (see p. 546, left column). Suppose that you have two good kidneys and I have none. My life is being sustained by artificial means, and I am in grave danger of dying soon. As it happens, you are well positioned to help me out, since your kidney would suit me perfectly. Are you obligated to donate a kidney to me, a perfect stranger? Surely not. Surely such a magnanimous action on your part would be supererogatory. But, Arthur points out, principle M1 implies that (very likely) you are obligated to help me, since (very likely) you can prevent something bad happening (namely, my dying from kidney failure) without thereby sacrificing anything of comparable moral significance (since the significance to you of the loss of one of your kidneys doesn’t come close to the significance to me of the loss of my life). His argument (an argument of Type V) may be put as follows:

Arg. 2: (1) If M1 were true, then it would be overall morally obligatory for a normal person under normal circumstances to donate a kidney to someone who needed it to survive.
(2) It isn’t the case that it is overall morally obligatory for a normal person under normal circumstances to donate a kidney to someone who needs it to survive.
(3) M1 is not true.

To reinforce his point, Arthur notes that we typically take people in general to have “negative” rights (rights, as he puts it, of “noninterference”) not to be harmed, but not also “positive” rights (rights of “recipience”) to be helped. For example, you have a right not to be harmed by me, but you don’t have a right to my kidney. This is not to say that no one ever has any positive rights. On the contrary, such rights can arise out of special relationships, such as the relationship between parent and child or doctor and patient. Suppose, for example, that Mr. Smith has a small son, Sam, and Mr. Jones also has a small son, Jim. Sam of course has a right not to be harmed by Mr. Smith, but he also has a right not to be harmed by Mr. Jones. Likewise, Jim has a right not to be harmed by Mr. Jones and he has such a right against Mr. Smith. The right not to be harmed is one that, in general, everyone has against everyone (unless it has been waived or forfeited − consider the harm of punishment). Now, Sam also has a positive right to be fed, clothed, and sheltered by Mr. Smith, and Jim has such a right against Mr. Jones. These positive rights, it seems reasonable to think, arise out of the parent-child relationship. Notice that, because this is so, we would not normally say that Sam also has a positive right to be fed, clothed, and sheltered by Mr. Jones, or that Jim has such a right against Mr. Smith. And, Arthur says, it is when rights are violated that especially serious wrong is done. If you were to take one of my kidneys (without permission), you would do something seriously wrong, since you would thereby violate my right not to be harmed; but you wouldn’t do something seriously wrong if you merely refused to donate a kidney to me, since you wouldn’t thereby violate any right of mine.

empty_bowlArthur’s criticism of M1 seems plausible. This principle really does seem too demanding. (But we shouldn’t be too quick to reach this verdict. Perhaps it is a mere rationalization on our part, a symptom of unwarranted complacency.) But even if so, Arthur’s criticism of Singer would seem easy to deflect.

Notice, first, that Arthur’s discussion of negative and positive rights, while perhaps accurate, is a red herring, since Singer doesn’t put his point, or argue for his position, in terms of rights. He never mentions rights at all! Arthur would say that neither the drowning child nor the starving stranger has a right to your help. Singer could, and I dare say would, agree. He doesn’t say that these people have a right to your help; he only says that you ought to help them, that it would be wrong of you not to do so.

Secondly, and more importantly, even if Arthur’s rejection of M1 is sound, this does not badly damage Singer’s case. Now, it might seem to damage it, since M1 is of course an integral part of Arg. 1. But what is really crucial to Singer’s case is the judgment, which he thinks you will share, that it would be wrong to pass the drowning child by. If M1 is unacceptable, as Arthur claims, then of course it cannot provide the best explanation for this judgment. What, then, might do a better job? Well, Singer himself provides an alternative principle; in fact, he does so in the very same paragraph as the one in which he introduces M1 (see again p. 538, left column). This other principle, which I will call M2, involves a double weakening of M1 and can be put as follows:

M2: It is overall morally wrong under any circumstances to let something very bad happen when it is in our power to prevent that thing from happening without thereby sacrificing anything of any moral significance.

(The double weakening of M1 is indicated by the two italicized terms.) And now Singer’s argument can be re-cast simply by substituting M2 for M1 as follows:

Arg. 3: (1) In the pond scenario, letting the child die is overall morally wrong.
(2) The best explanation of premise (1) is moral principle M2.
(3) If premise (2) is true, then M2 is true.
(4) M2 is true.
(5) If M2 is true, then our (affluent people’s) letting others starve to death is, under present circumstances, overall morally wrong.
(6) Our letting others starve to death is, under present circumstances, overall morally wrong.

dry_earthAs far as I can tell, nothing that Arthur says provides any reason to doubt any of the premises of this argument. In particular, his criticism of M1 seems not to apply to the weaker M2. If this is the case, then Singer’s conclusion survives unscathed, even if we accept that there is in general something worse about killing (perhaps because it involves the violation of a right) than about letting die (which perhaps does not involve the violation of a right).

***Here is a good place to stop and consider Question 3 on the Practice Questions for Module 6.***

Slote’s critique:

Slote in effect rejects both M1 and M2. He does so by proposing what he calls “an ethics of caring.” His claim is that whether some act is overall morally right or overall morally wrong can depend on whether and to what degree it exhibits empathy. He says that whether one person shows empathy toward another person often does and, moreover, ought to vary according to the level of intimacy that exists between the two (see p. 549, left column). In saying this, Slote is indirectly challenging Singer’s principle of geographical neutrality and claiming that, contrary to what Singer claims, the pond scenario and the famine-relief scenario are not morally equivalent, precisely because the former involves helping (or not helping) someone who is “up close and personal,” whereas the latter does not. The failure to help the nearby child would betray a marked deficiency in “normally or fully developed human empathy,” whereas the failure to help a far-off stranger would not. In place of M1 and M2, Slote is thus endorsing something like the following:

M3: It is overall morally wrong under any circumstances to let something (very) bad happen to someone who is near and/or dear to us when it is in our power to prevent that thing from happening without thereby sacrificing anything of (comparable) moral significance.

Notice that this principle would justify discriminating between the nearby child and the far-off stranger, precisely because the former is near to us and the latter is not. Slote claims that this difference between the cases not only explains why we (tend to) judge turning away in the pond scenario more harshly than in the famine-relief scenario but justifies this judgment (see p. 552, both columns).

It’s important to recognize that Slote is not rejecting Singer’s claim that we are obligated to do more than we usually do with regard to famine relief. He leaves open the question whether and to what degree we are obligated to help out, as opposed to its being supererogatory for us to do so. He simply denies that we are obligated to help out to the extent that Singer claims we are. (See p. 555, both columns.)

What should we make of Slote’s remarks? Certainly what he says is reassuring, since it absolves us of the great wrong that Singer claims we routinely commit when we fail to help those who are in need to the extent that we could, but is it true? I’m afraid that there is reason to doubt what Slote says. The primary problem, as I see it, is that Slote conflates two distinct kinds of moral evaluation. First, there is the evaluation of acts. It is with this kind of evaluation that we have been almost exclusively concerned so far in this course. Such evaluation is what is at issue when we declare an act morally right, wrong, obligatory, or supererogatory. Then there is the evaluation of agents, that is, people who perform acts. We’ve neglected this kind of evaluation, but that isn’t to say it isn’t important. On the contrary, it is. Such evaluation is what is at issue when we declare a person virtuous (because he is, say, compassionate or caring or courageous or conscientious…) or vicious (because he is cold-hearted or callous or cowardly or oblivious of moral demands…). Slote appears to think that these two kinds of evaluation must somehow complement one another, but it is hard to see why we should think this. Doesn’t it seem that they can fall apart? In particular, doesn’t it seem that someone can do what he ought to do and yet be cold-hearted, or fail to do what he ought to do and yet be compassionate? For example, you might rescue the nearby child, not out of compassion, but because you anticipate a reward. Haven’t you done the right thing, even if you’ve done it for the wrong reason? Your lack of compassion doesn’t mean that, when you rescued the child, you didn’t do what you ought to have done, even if it does mean that you didn’t do it in the right spirit. Notice that Singer doesn’t say anything about the spirit in which you should or should not rescue the child or save a stranger from starvation. He simply says that, just as you ought to do the former, so ought you to do the latter. Is there anything in what Slote says that shows that Singer is mistaken in making this claim? I confess that I cannot find it, which leaves me in the decidedly uncomfortable position of not knowing how to defend my self-centered lifestyle against Singer’s attack. How about you?

***Here is a good place to stop and consider Question 4 on the Practice Questions for Module 6.***

PHI 121