Derek Goodwin

Peter Singer at Farm Sanctuary, a shelter in upstate New York for animals that have been rescued from stockyards, factory farms, and slaughterhouses, August 2006


We all want to know how to live. That includes not only knowing how to get what we want, but knowing what to want, and what we should and shouldn’t do. Peter Singer is prepared to tell us, and because his advice would require most of us to change our lives, and because it is offered with such force and clarity, he is an important figure in moral philosophy. He has also had a larger practical impact on the world than any other philosopher of our time. His 1975 book, Animal Liberation, led to effective movements to reduce the suffering of animals in factory farming, scientific experiments, and the testing of commercial products such as cosmetics, and it has persuaded many people to become vegetarians to one degree or another.

Singer’s claims about what well-off people in affluent societies should do to help those living in poverty elsewhere in the world have had less effect so far, but he hopes to remedy that with his latest book, The Life You Can Save. “The ultimate purpose of this book,” he says, “is to reduce extreme poverty, not to make you feel guilty.” But making the reader feel guilty is one of his specialties, and a key to his effectiveness as a writer. Whether he is describing cruel and pointless experiments on heroin addiction in monkeys or the devastating effects of easily reparable obstetric fistulas on impoverished third-world women who have given birth in their teens, he acts on our emotions, and the impact is heightened by his calm, cerebral style.

Singer’s exchanges with critics in Peter Singer Under Fire cover the full range of his views, but I will start with what he says about affluence and poverty. The Life You Can Save repeats and develops an argument he originally offered in 1972, in an article that has probably been read by more students of moral philosophy than any other text, ancient or modern.
1 He begins with an example: You are walking past a shallow pond, and you notice that a small child has fallen into the water and is about to drown. Should you wade in and rescue the child, even though it will ruin your shoes and get your clothes muddy?

Most people agree that anyone who didn’t rescue the child would be a moral monster. Even if the case is more demanding—the child has to be taken to a hospital, and that will make you miss a flight for which you have a nonrefundable ticket—it would still plainly be wrong not to save the child’s life. The next question is, what is the principle that explains why failing to rescue the child would be wrong? Singer argues that any plausible explanation of what is wrong in this case has vast implications. He offers the following simple principle: “If it is in your power to prevent something bad from happening, without sacrificing anything nearly as important, it is wrong not to do so.”

The Life You Can Save is mainly occupied with drawing the implications of this principle for people who have more money than they need to lead a decent life, in a world like ours where a billion people suffer from malnutrition and millions die from easily preventable diseases. He needs only two other premises: (1) Suffering and death from lack of food, shelter, and medical care are bad. (2) By donating to aid agencies, you can prevent suffering and death from lack of food, shelter, and medical care, without sacrificing anything nearly as important. It is typical of Singer to spell out the first of these two premises, just to make sure that the reader is not a complete moral idiot. But most of the book offers detailed empirical support for the second premise, describing and evaluating what aid agencies do, how effective it is, and how much it costs.

The facts are complicated, and there is controversy over the most useful forms of aid. Singer is fully aware of the problems posed by corruption in recipient countries, the importance of self-sustaining development rather than handouts, and the difficulty of ensuring that medical care and preventive measures are actually delivered to those who need them. He acknowledges that sometimes conditions may be so bad that there is nothing we can do.

But he makes a strong case that private aid has done a lot of good, and could do a lot more if it had more resources. He quotes estimates that programs to provide safe drinking water cost $250 per life saved; programs for mosquito nets against malaria cost $820 per life saved; surgery to cure cataract blindness, obstetric fistulas, and deformities like cleft palate costs from $500 to $1,500. And the per capita annual development aid that would enable people in extreme poverty to reach self-sustaining subsistence is estimated to be less than $200.


If you accept these estimates, what should you do? How much of your income and wealth could you contribute to the prevention of death, deformity, and chronic hunger “without sacrificing anything nearly as important?” According to The New York Review, the median household income of its subscribers is $123,600 a year. If that is your pretax income, my guess is that Singer’s principle applied literally would require you to give away more than half of it. You could survive perfectly well on what was left, though you’d have to move to cheaper housing, give up travel, restaurants, wine, opera, taxis, private schools, and cable television, and perhaps buy your clothes at Wal-Mart.

But you and I know that you are not going to do this, and so does Singer. In fact, Singer says he himself doesn’t give away as much as he thinks he should (a New Yorker profile ten years ago reported that he gave away 20 percent of his income). And that poses the question whether any of us believes Singer’s principle, plausible though it sounds. Does Singer himself really believe it? Or rather, in the sense in which he believes it, does it mean something different from what it appears at first to mean—something less radically demanding? The answer to these questions depends on the relation between morality and human motivation, a difficult topic in moral philosophy, for which Singer’s claims are crucially important. Singer’s principle seems to explain why we must rescue the drowning child. On the other hand, it seems to require us to do much more, in order to save hundreds of distant children. Most of us accept the first demand and balk at the second. Does that show that there is probably something wrong with the principle, even if we cannot easily say what it is?


There are three possible responses to this problem. The first is to say that the principle is correct, that it tells us we have a decisive reason to give away most of our income, and that our resistance, however natural, is due to motives of self-interest that do not provide reasons that outweigh the impartial demands of morality. The second is to say that the principle is incorrect, and to look for an alternative principle that explains why it is wrong not to save the drowning child but not wrong to keep most of our income. The third is to say that the principle is correct, but that it tells us only what morality requires us to do, not what we have decisive reason to do. What we have most reason to do, all things considered, depends on how much we care about not doing what is morally wrong; and since most of us care about many other things as well, we have significant reasons not to act in accordance with Singer’s principle.

You might think that Singer’s answer would be the first. That is what made his 1972 article so electrifying. But in later writings he explains, though with some ambivalence, that his answer is the third. This means that his position is more philosophically complex and less threatening than it at first appears. But before examining this aspect of his view, let me say something about the second type of response, which Singer clearly rejects. Responses of the second type are based on a different conception of morality from Singer’s, one according to which moral principles attempt to say how we should act in relation to other people, all things considered, by identifying and taking into account the full range of reasons bearing on our choices.

Someone who thinks of morality in this second way still faces a serious challenge from Singer’s argument, because Singer makes you feel that even for a person of modest means, the very strong reason to accept a sacrifice in the case of the drowning child does not disappear when we focus instead on an anonymous distant child. And efforts to explain away the latter reason often seem like lame excuses for a patent inconsistency. The question is whether there can be moral principles that include impartiality but are not dominated by it—principles that acknowledge both that everyone’s life is of equal value and that everyone has his own life to live.

One example of such resistance to Singer’s argument is the proposal of the moral and legal philosopher Liam Murphy that our responsibility to aid those in desperate need is a collective responsibility, and that each of us is required only to do our fair share. What makes Singer’s principle so demanding is that most affluent people contribute nothing to alleviate extreme poverty,
2 so that anyone who acknowledges a reason to do so is faced with the apparent demand to fill the gap and empty his pockets. If all well-off people gave a modest amount, the problem would be solved, and that, Murphy argues, is what each of us should contribute, whether or not others do their share. Morality should not make us hostages to the bad behavior of others.


This seems to distinguish the drowning child case from the case of generalized aid, but Singer points out that it doesn’t seem to work if one is faced with a dozen drowning children and a dozen bystanders. If all the others do nothing, it isn’t acceptable for you to rescue just one child and let the others drown. You have to help as many as you can. If the “fair share” limitation is to survive, it probably has to be coupled with a more stringent requirement of immediate rescue. But Singer would reply that it is morally irrelevant whether someone who needs help is right in front of you or halfway around the world.


Singer rejects all moral theories that, as a matter of principle, limit the subordination of our personal interests to the requirements of strict impartiality. He holds that you would be morally justified in keeping most of your income only if, improbably, doing so would promote the general welfare better than giving it away.
4 He is an adherent of utilitarianism, the moral theory developed by Jeremy Bentham, John Stuart Mill, and Henry Sidgwick, and favored by Singer’s teacher R.M. Hare. According to utilitarianism the rightness or wrongness of actions or omissions depends on their contribution to the overall balance of happiness minus unhappiness for all those sentient creatures affected, with the benefit or harm to any creature counting the same as comparable benefit or harm to any other. Your own happiness and that of the people you love do not count any differently than the happiness of strangers.

In Bentham’s version of utilitarianism the sole measures of happiness and unhappiness are pleasure and pain. In Singer’s version what counts is the satisfaction of preferences, including but not limited to the preference for pleasure and against pain. For example, human beings typically have a strong preference to go on living, and are prepared to accept considerable pain and other sacrifices, if necessary, to avoid their own death. Utilitarianism simply counts the preferences of all persons, and other creatures, impartially in determining what any individual should do.

However, as Singer emphasizes, this does not mean that we should continually calculate the effects of each of our actions as we go through life. First, as a practical matter, the ends of utilitarianism are better achieved if we follow rough general rules of conduct, avoiding harm and being kind, considerate, and reliable in our dealings with others. Second, some of the greatest sources of happiness in human life are attachments and commitments that involve giving priority to the interests of particular people, for example one’s children. The general happiness is best served if people’s close personal relations are not governed by concern for the general happiness. Still, Singer says, “this doesn’t mean that parents are justified in providing luxuries for their children ahead of the basic needs of others.”

Third, there is special value from a utilitarian standpoint in accepting and inculcating certain strict rules such as the rule against murder, which blocks the use of utilitarian calculation as a way of deciding whether to kill an innocent person deliberately. The reason is that the general bad effects of not having such a rule, with a widespread revulsion against murder, would be much worse than the loss of the occasional happiness-enhancing murder that the rule would prevent. Something similar is true of the rule that torture is always wrong. Here is what Singer says:

Given the well-documented tendency of police and guards to abuse prisoners, and the low probability that torture will yield useful information, that rule seems likely to have the best consequences. Yet, I would argue, if I find myself in the highly improbable scenario where only torturing a terrorist will enable me to stop a nuclear bomb from going off in the middle of New York City, I ought to torture the terrorist. What the individual ought to do, and what the best moral rule directs one to do, are not necessarily identical.

This is the utilitarian strategy for explaining the apparent force of all strict rules of ordinary morality, such as the wrongness of lying, promise-breaking, and theft. None of those rules is morally basic, but their automatic acceptance is useful. This is another feature of utilitarianism that is widely contested, by those who believe that an adequate moral theory cannot be just a matter of adding up costs and benefits but needs to explain how torture, murder, and betrayal are wrong in themselves.


In a different way, Singer applies the idea of a disparity between what individuals really ought to do and what is the best rule for their guidance when formulating his recommendations about aid from the affluent to the destitute. The principle he believes to be morally correct would require the affluent to lower their standard of living drastically, but he recognizes that to insist on this might be counterproductive:

Daunted by what it takes to do the right thing, they may ask themselves why they are bothering to try. To avoid that danger, we should advocate a level of giving that will lead to a positive response.

So he proposes a target that, if everyone met it, would generate more than enough to solve the problem of extreme poverty: “5 percent of annual income for those who are financially comfortable, and rather more for the very rich.” He provides helpful tables of progressive rates to allow you to calculate your recommended contribution, and the median New York Review household would contribute $6,180. If your income is higher, say $200,000, you would contribute $12,600; if it is $500,000, you would contribute $48,450. If the top 10 percent of American earners alone contributed at these rates, it would raise $471 billion a year to help the world’s poorest billion people. But even if many will not contribute, Singer believes that promulgating a norm of this kind is likely to be more helpful than battering people with the more stringent demand of his original principle, which would require most readers of The New York Review to contribute more than half of their gross income.

He supplements the argument with comments on Larry Ellison’s $200 million yacht and the $45 million the Metropolitan Museum paid recently for a Duccio Madonna and Child (using philanthropic contributions):

In buying this painting, the museum has added to the abundance of masterpieces that those fortunate enough to be able to visit it can see. But if it only costs $50 to perform a cataract operation in a developing country, that means there are 900,000 people who can’t see anything at all, let alone a painting, whose sight could have been restored by the amount of money that painting cost. At $450 to repair a fistula, $45 million could have given 100,000 women another chance at a decent life. At $1,000 a life, it could have saved 45,000 lives…. How can a painting, no matter how beautiful and historically significant, compare with that? If the museum were on fire, would anyone think it right to save the Duccio from the flames, rather than a child? And that’s just one child.


Peter Singer Under Fire is a valuable though uneven volume in which Singer, responding to fifteen critics, explains and defends his views. It begins with a seventy-four-page intellectual autobiography that is of great interest. He was born in Australia after World War II to Jewish refugees from Austria. Three of his grandparents died in the Holocaust, and no doubt this background has left its mark on him, in the form of a countervailing hope for the transformative power of morality.

Singer formed his basic views early. As an undergraduate he did not react as expected to the classic argument that utilitarianism must be wrong because it would justify framing one innocent person for a crime if that were the only way to prevent the lynching of five innocent people by a mob intent on revenge. Instead of taking this as a refutation of utilitarianism, he thought it might well be the right thing to do. This exemplifies the moral equivalence between harms you produce and harms you fail to prevent that is such an important element of utilitarianism, and that determines his controversial later views on the irrelevance of the distinction, in medical ethics, between killing and letting die.

His beliefs about our duties toward animals and toward people in desperate need were formed when he was a graduate student at Oxford. He later drew another consequence from his rejection of “speciesism”: the denial of the sanctity of human life, according to which all human beings, whatever their particular characteristics, have a value that no other creatures have, and should therefore be kept alive and above all not killed. Singer holds that human beings without a conception of their own future existence, such as newborns, have no interest in not being killed. He defends not only abortion but the killing of irreversibly comatose patients, and even infanticide for severely disabled infants, instead of permitting them to die slowly of afflictions that could be cured but are not, as is now sometimes done.

This last position has provoked outrage, but he insists that severe disabilities impair the quality of life, and that this should have consequences. In response to an essay by the late Harriet McBryde Johnson, an advocate for the disabled who was seriously disabled herself, Singer says:

We could even rescind the ban on thalidomide—why should pregnant women not take advantage of this drug, so helpful in reducing morning sickness, if the fact that your child is likely to be born without arms or legs has no tendency to reduce his or her quality of life? If this sounds grotesque, that is because the view that implies it is so difficult to take seriously.

On the other hand, he recognizes that as a practical matter, there probably has to be a legal bright line against killing after a certain point, and that although it is morally arbitrary, “there is something to be said for taking birth as the cut-off line, at least in terms of the criminal law.”

The most interesting critical essays in Peter Singer Under Fire, which prompt the most interesting responses from Singer, are those by Bernard Williams on humanism (his term for the solidarity with our own species that Singer condemns as “speciesism”); Don Marquis on abortion and infanticide; Judith Lichtenberg on moral motivation; Richard Arneson on the stringency of morality’s demands; and Michael Huemer on metaethics, that is, Singer’s conception of the meaning of ethical claims and the grounds of their correctness. The last of these topics brings together many of the concerns of the other essays, and poses the most philosophically difficult questions about Singer’s position.

Unlike Williams, a famous opponent of utilitarianism, Singer believes that the content of ethics depends on our taking up what Sidgwick called “the point of view of the universe,” and thereby transcending not only our individual point of view but the point of view of our society and of our species. Of course Singer does not believe that the universe has a point of view, but he thinks this is an apt metaphor for the human capacity to take up a standpoint of impartial and equal concern for the welfare of all sentient beings.

Singer’s conviction that this is the standpoint from which moral judgments must be made depends on a theory of the meaning of moral language that he takes from R.M. Hare. Hare holds that moral judgments express what he calls “universal prescriptions,” which means, roughly, that when I say that someone morally ought to do something, I am not saying merely that I want him to do it, but that I would want everyone in similar circumstances to do the same thing. This includes wanting it to be done in all the hypothetical cases with the same structure, in which I myself occupy each of the roles of those affected by the act.

To decide whether you in your present circumstances are morally required to give away half your income to Oxfam, for example, you have to decide whether you would want this to happen in the full set of structurally similar circumstances in which you were each of the potential donors and each of the potential beneficiaries. Hare claims that the only way to decide whether to endorse this universal prescription is to ask yourself what you would actually want if you were each of the persons affected, and then to weigh up the strengths of their preferences pro and con. This amounts to a criterion of preference utilitarianism. Moral requirements derive from an equal and impartial consideration for the interests of all.

Hare thus attempts to ground utilitarianism in the analysis of moral language and the universal prescriptions he believes such language entails. However, this turns out to be a very weak foundation, since Hare does not believe that there is any rational requirement to engage in moral judgment, to govern one’s conduct by moral standards, or to use moral language at all. If someone just didn’t care about what he would want everyone to do in parallel circumstances, then morality would have no grip on him. He could say, “Of course if I were starving I would want rich people to contribute to famine relief, but so what? Whether my conduct is ‘wrong’ in that sense doesn’t interest me.”


The more radical and demanding a moral claim is, the more pressing is the question what reason we have to live in accordance with it. This is an acute problem for Singer, both because he espouses a radical morality and because, like Hare, he is drawn to the view, associated with David Hume, that all our reasons for action depend on our desires, and that our desires themselves are neither rational nor irrational:

So if some people prefer to follow my arguments and give almost all their money to aiding those in great need, there is nothing at all irrational about this. Equally, giving away money to those in great need is not rationally required.

When he is thinking in this vein, Singer searches outside morality for reasons to be moral, and he has argued, for example in his book Practical Ethics,
5 that living by principles of radical impartiality will make us individually happy by giving meaning to our lives—something we presumably care about already. But another approach also tempts him, one that he suggests in a later book, The Expanding Circle —namely that the impartiality that underlies utilitarian morality is itself a requirement of reason:

That one’s own interests are one among many sets of interests, no more important than the similar interests of others, is a conclusion that, in principle, any rational being can come to see.

This idea is found in Sidgwick, and it is opposed to the Humean outlook because it asserts that not only our factual and logical beliefs but also our motives are subject to rational requirements. In other words, it is contrary to reason to prefer eating a meal in a good restaurant to saving a child from death by malnutrition at the same cost. In the end, however, Singer doubts that the appeal to reason can serve as a basis for ethics, since if someone doesn’t care about the interests of others, telling him he has an objective reason to care won’t get him to act any differently.

I think Singer has taken the wrong path here. Huemer argues forcefully and convincingly that Singer’s overall moral position is more consonant with Sidgwick’s reason-based outlook than with Hume’s desire-based one. Sidgwick’s argument requires that reason should be capable of getting us to change our desires at the most basic level, and also to refuse to act on some desires for reasons that do not derive from other, already existing desires. Yet that is precisely what seems to happen when, in response to Singer’s arguments, someone becomes a vegetarian or a major donor to Oxfam for reasons of consistency.

The heart of Singer’s case is that it is inconsistent to be very concerned about human pain but hardly at all about animal pain, and inconsistent to care about the drowning child in front of you but not about the dying child in a distant country. When people change their conduct to avoid such inconsistency, their motives have been altered by reasoning. Singer also works on our feelings, but in large part he is calling on us to transcend by rational thought the motivational habits of nature and culture toward a truer understanding of how we have reason to live—he is not just describing a moral life that is rationally optional.

Such a reason-based defense of utilitarianism would be opposed not only by antirationalists but also by other rationalists, many of whom think that the impartiality demanded by utilitarianism oversimplifies the complex territory of true reasons for action. Sidgwick, though he defends impartiality as a requirement of reason, despairs of reconciling it with the competing principle of rational self-interest, which seems also intuitively self-evident. That is one of the problems that discourages Singer from accepting reason as the basis of morality. But a reason-based morality may have to take on both these types of reason, and perhaps others, in constructing the principles by which we should live—principles that aspire to the reflective endorsement of everyone, even when their personal interests conflict.

That is the likely outcome if we take as the aim of ethics, as Singer is reluctant to do, the discovery of standards that everyone has reason to meet. The result might be a pluralistic morality that combines impartiality with respect for each individual’s inviolability and personal autonomy.
7 It would probably include some version of the relatively modest duties to others in need that Singer recommends as a “realistic” fall-back from what morality truly requires.

Singer’s own message is very different. He believes that the moral truth is very demanding, and in spite of his expressed ambivalence, one gets the sense from his writings that deep down he is convinced that everyone is not just morally required to give equal consideration to the interests of all, but has decisive reasons to do so—and that the recognition of those reasons has the power over time to transform the world. Those of us who are not persuaded that he is correct about the content of morality can nevertheless be grateful to him for posing these questions so vividly.

This Issue

March 25, 2010