Peter Singer & The Drowning Child

I’ve put together a new interactive activity that’s based on Peter Singer’s famous article, “Famine, Affluence & Morality”.

The activity is here.

It’s in an early beta-testing phase, invisible to the wider world, but at the stage where any feedback would be very useful.

I’m going for slightly less philosophical rigour than usual with this activity. My hope is that it will challenge people in terms of how they tend to think about their obligations towards other people in faraway places (or at least their behaviour in this regard). However, this “challenging people” aspect has meant that I’ve had to downplay (i.e., ignore!) some of the philosophical complexity. Nevertheless, I don’t want it to be a complete philosophical disaster, so if anybody notices any obvious omissions, lacunae, embarrassing slips, etc., then please say so here. It would be much appreciated. (And, of course, if you notice any programming errors, or it crashes, or something, it’d be very helpful if you could let me know.)

Also, it’s not quite finished in the sense that it needs a summing up page of analysis at the end, but it’s complete enough that it makes sense to play through it.

Thanks. Have fun. And you can donate to Oxfam here!

Leave a comment ?


  1. Good game and as you remark, it seemed less complicated to play than previous games.

    I wonder about the use of phrase “moral obligation”. I said that I had a moral obligation to save the child (except when he or she was in a distant country), but I don’t really believe that there is such a thing as a “moral obligation”.

    Isn’t there another word or phrase that would convey more or less the same idea, without bringing in the concept of moral obligation?

    I for one would save the nearby child because I could not live with myself if I did not, not because I am morally obliged to.

  2. Hi Amos

    Thanks for this.

    Yeah, I take your point about the phrase “moral obligation”, but actually if you had answered that you didn’t have a moral obligation, then the activity would have given you the option of choosing a more minimal conception of roughly the same idea.

    Also, so far, all but one person has stated they do think there’s a moral obligation to save the child (though obviously your point means that it isn’t entirely clear what people mean when they choose that option).

    The activity is actually quite complex, it’s just that if you choose the most common pathway through it, that isn’t obvious.

  3. “Moral Obligation” is difficult. I think it still applies here though. If the idea of moral obligation is that you would be acting immorally by not doing some thing, then it suggests that not saving the child is immoral. In this case in particular, i.e that there is no significant risk to yourself, it seems difficult to successfully argue that you would NOT be immoral by leaving the child to drown.

    I liked the direction that the bike question went but I expected the follow up to be something about a more expensive bike or one we’re more attached to. I can see why you didn’t go there as that wasn’t Singer’s point, but I personally find it an interesting concept. How much must we sacrifice before the “moral Obligation” disappears?

  4. I like these interactive games. I had fun thinking about this one, but I do have a reaction to the very last question – SPOILER ALERT


    In reference to the ‘do you have a moral obligation if you saved the child yesterday’ as an analogue to ‘do you have a moral obligation to donate today if you did tomorrow’:
    1. I do not think the activities of my donation today will be effected in one day. In fact, since most organizations have a plan for about a year for their programs the ‘saving of the child’ is not complete until the end of that plan. Also, taken to its logical extreme, we would have to donate constantly – making me question at what point the drain on my resources becomes a mitigating factor on my moral obligation. I didnt see any questions that tested that boundary. The late for work/soiled shirt/stolen bike would be a mitigating factor if it was cumulative – e.g. I would not be able to hold my job if I was constantly late, making me unable to feed myself, thus making me unfit to save the next child.

  5. I agree with Stophe, that the, “but I saved a drowning child last week,” argument is not really universalizable. There is always a person in trouble who can be helped by a donation. It would be like there being a big lake that always had a drowning child in it. At some point your moral obligation would have to change to putting up a fence or hiring a life guard rather than a never ending series of rescues. To me, at least, that’s analagous to contributing a decent, but not debilitating amount to charity per year, voting for the right politicians and maybe even doing some volunteer work. And if I did that last week, I’m under no obligation to do it again this week.

  6. give a man a fish and you feed him for a day.

    teach a man to fish and he will deplete the fish population, pollute the sea shore and his thousands descendants will starve. Or become illegal immigrants.

    Shoot him in the head and the problem is solved.

    Sorry. I don’t get morals.

    Right now I am concerned with feeding myself for the next 20-30 odd years.

  7. Hello Leo:

    Good to see you back.

    I’m not sure myself who or what obliges me to be moral or why anything at all obliges me to be moral.

  8. @Sophie & @Gene

    Yeah, I anticipated that sort of objection (the reason I specified “in the next few days” is because I wanted to weaken the “Ah, but I could always do it tomorrow” argument).

    A few points:

    1. Sure, there is a disconnect between the donation & the money working through the system, but nevertheless it presumably is the case that any particular donation could make the difference between the life and death of a person (albeit you’re never going to be aware of the counterfactual). I’m not sure the disconnect destroys the argument (though I think in a roundabout way there is a worry here).

    2. I don’t think the “Ah, but at some point we would have to do something other than performing endless rescues” objection works. Partly because there’s another question in there that tests that point (the water safety question). But also because it’s entirely possible to find an overseas agency that does more than simply providing one shot hits of food, etc. Singer actually deals with this argument in his article (aid agencies can promote sustainable growth, etc).

    3. Singer’s position in his original article just is that actually the logic of the argument requires that we give up large chunks of our material wealth (so yes, we have to keep on saving lives/giving donations).

  9. Your available responses don’t take into account that I have standing order set up to Oxfam, so I’m not clear if you’re talking about the next scheduled payment, or a new additional one.

    I presume from your post above, you’re talking about a new additional one.

    I have an intuition that it’s not distance that “cancels” the obligation, but the number of lives involved. This relates to the Chilean mining disaster where the whole world helped, *because* there were only a few of them.

  10. According to the argument, you are fulfilling your moral obligation already if you are donating to a place like Oxfam. The point is that a donation of that nature is equally morally obligatory as saving a child you see drowning near you.

  11. Another issue, although it is more of an intuition than well thought out, is that the difference between what I’ll call special circumstances moral obligations and institutional moral obligations is important. The drowning child examples all strike me as special circumstances. They are like contributing to a charity to help the victems of the Haitian earthquake or the Japanese tsunami. They are special, localized events that require special, localized effort.

    If the examples were retold as there being a bridge on my way to work where I know homeless people congregate, am I obligeted to buy extra oatmeal and deliver it on my way to work? If so, how often and how much? Those answers are much less clear than “Should I help a drowning child?”

    I understand that the point Singer is trying to make is that if you have the means to save a life, you should save that life. But I can’t shake the feeling that the circumstances surrounding the event matter and that seeing an event unfold in your presence is different, morally speaking, than being aware of suffering somewhere else in the world.

  12. I think you may want to stress a bit more that a no point will the life-saving donation become overly burdensome. I’m also not sure that you need to mention the aid agency since some people may have objections or intuitions about fullfilling duties through agencies/charities.

  13. Thanks guys, I’m reading all these.

    @Steve – Not sure about either of your points, because of wishing to remain true to Singer’s original argument (as it is, I’m focusing on the less radical version of his thesis). Singer clearly thinks – or he did back then – that we’re required to make fairly large sacrifices (presumably until the point at which the marginal utility of our sacrifice is trumped by the marginal cost).

    The problem with your second point is that Singer flags up aid agencies as the solution to the problem of acting at a distance (though maybe things have changed with the advent of the internet age, micro-credit charities, etc).

  14. @Jeremy if the intention is to follow Singer’s argument as you’ve written – right up until the burdens become very heavy – then I’m not sure the examples convey this strongly enough then. It certainly doesn’t seem all that budensome to save the drowning child at the loss of my crappy bike two weeks running. Nor does it seem overly burdensome to donate a little to save the distant child each month/week as my paycheque comes in. Perhaps I skipped through the quiz a little quickly and missed something.

  15. Interesting. As I have strong moral objections to the way in which I have seen charities operate in places such as Bosnia and Sierra Leone, I choose not to support them. I would however, have no hesitation in assisting a “drowning child” in the given circumstances. So, does this then show a lack of moral courage on my part by refusing to contribute on separate moral grounds?

  16. @Steve – Thanks. I think you’re right that there is this tension. Trouble is, there are two versions of Singer’s argument. In his 1971 article, he gives both versions, but emphasizes the fact that his argument suggests we take on large burdens (or give up significant goods, etc). But in the article he wrote in the New Internationalist, he talks specifically about giving up something roughly the cost of a CD, so the stronger version isn’t there.

    I think the logic of his argument suggests that the stronger version is in play. But I’m aware that most people simply won’t buy into that, so I’m try to set it up with the weaker argument while preserving the integrity of the argument.

    Maybe I haven’t succeed, though!

  17. Extending the arguments to the distant child begs the question that the only difference is the distance. However, to me that is not the case. The distant child moves the argument from one specific child that is in your direct experience to, effectively, all children in danger everywhere. The argument then becomes “do we have a moral obligation to save all children everywhere?”. Or “if we have a moral obligation to save a single child, do we have a moral obligation to save all children?”. I can see that some people would answer yes to that and act in accordance with their beliefs. But anyone who buys a coffee at Starbucks or gives their own child a teddy bear for Christmas, is drawing the line somewhere. Does Singer accept his own argument and live a life of poverty, dressed in rags, dumpster diving for food? If he doesn’t, he draws that line too.

    Even if you accept the obligation to save all children, what moral obligation do you have to make sure you couldn’t deploy your funds differently and have a greater effect?

  18. GrafVonBek – I guess it just depends how the moral calculus turns out. Singer makes the point that there is quite a lot of choice in terms of how, and to whom, we give our money (given that we are obliged to do so), and that except perhaps in the case of systematic governmental abuse, there isn’t good evidence that suggests that charities are so inefficient, etc., that our money makes no difference (or makes things worse).

    I think once on starts thinking in these sorts of broader ways, then everything rapidly becomes very complicated, because one could imagine a utilitarian argument that holds something to the effect that a horrible famine every now and then is necessary, because its effect is to focus the world’s attention on the problem of starvation, with the effect that more people are saved in total than would be otherwise (I’m not saying that argument works, but it doesn’t seem to be ruled out of court).

  19. @Jeremy – Singer took part in BBC Radio 4’s Iconoclasts programme (with the RSA) a few years ago; in it he admitted that he doesn’t give anywhere near as much as he thinks he should (10% of his income): I think he might fail this test! It’s still an enormously powerful argument though.

  20. @jeremy – I think I just “rapidly complicated” my own thinking by considering whether donations to one charity could possibly be considered as detrimental to the funds to another and thus morally dubious! Back to the moral drawing board for me…

  21. @Steve – Ah, thanks, I didn’t know that. I might mention it. It is a powerful argument. Kinda cool that something so simple can (potentially) have such profound implications.

    @GrafVonBek – Yeah, there are many layers of moral complication once one begins to think about this stuff.

  22. On the test itself:

    “Yup, that’s got to be the morally sound response…”

    “Excellent, we reckon that response shows you’ve got your head screwed on right.” – Well, who are you to judge?

    Surely your post-test confirmations are providing a bias.

    I tried the test by starting off with a negative answer, and pretty soon it’s assumed I’m a moral nihilist (see next comment).

    I find the test is telling me how to respond rather than asking me.

  23. Answering in the positive does not allow that I might be an expressivist: I might be a moral nihilist, in terms of my view on the objective existance of morals, but who accepts that human beings have evolved with certain behaviours that socially and historically we have developed into moral codes that complement our biological feelings to some extent, coupled with our rational assessment of how to satisfy those feelings most effectively as a group.

    This test, and the simplistic arguments do not examine anything useful as far as I can tell.

  24. Saving a local drowing child opportunities are rare. Even if I did the same last week, it’s still rare and I would be struck by the coincidence, but that coincidence would not prevent my action.

    But all this is a poor comparison to the giving to Oxfam. Let’s strike up a better comparison.

    There are many children around the world who are starving or otherwise in trouble, so one or two local, or distant, children in need that I can help directly does not compare. The failure isn’t in the response (save a local child, therefore donate), but in the comparison itself.

    What if it was the habit of all local children to fall into deep water and if most other local adults did little to help because they could not swim. If I attempted all rescues I would exhaust myself, fail to eat and die, so preventing further rescues. Surely, I would pace myself to do what I could. Do Western aid workers helping in foreign aid centres give up all their food to the starving? Do they become part of the problem? Or do they have they pace themselves, have their own supplies? If an aid working gave a bit of their personal supply last week, shouldn’t they give it all, all the time, by extension?

    So, by comparison, in an extreme situation I am not only not obliged to rescue every drowing child, and not only is it impossible for me to do so, but I would feel I’m obliged to pace myself, feed, sleep, work, and yet contribute all of my spare time saving as many children as I could.

    Or, perhaps I could, instead of actually saving children, be better fighting for better education of children, the teaching of swimming to the very young and to adults, the fencing off of dangerous waters, the education of parents in the protection of their children, and so on. History is full of activists that have made far more effective change by changing the system than by saving individuals – though they may do that as well. There’s even the strange situation where rich philanthopists donate large sums, from their fortunes that have been acquired by unsavory means.

    So, what about these starving children that Oxfam serve? Maybe I can help some of them immediately by donating to Oxfam. But the argument, if I donated last week I should donate this week, can be extended indefinitely, until I am poor, in debt, out of work, starving and looking for help from Oxfam myself. Clealry there’s a balance to be had. And that balance might be better served by political activism that changes our government, which can (or could), through taxes, make a far better donation to foreign aid.

    And what about the governments that are responsible in these foreign lands for the state of their people? Perhaps a better result would be to make a military move against them, enforcing our own control, our own engineering solutions, including birth control, to prevent suffering.

    Iraq may eventually, over many decades, or maybe centuries, actually be better off as a result of the war on Iraq, than it would have been left to its own devices under the Hussain family. The problem is that we can’t be certain of that; we can’t be certain that all the deaths that occurred as a direct result of the war on Iraq weigh less than those saved in the very long term. But the Singer argument that uncertainty of outcome should not be a bar to action surely applies here?

    The extended analysis, from simplistic local cases to complex global political ones does not work. It trivialises the complexity of the foreign aid problem by comparing it to simplistic local ones.

    As with many moral arguments the individual objections can be overcome quite easily. All the push-back aruments are fine in supporting the donation, as far as they go, when comparing simplistic logical statements. But as a compound argument they fail. And when other realistic alternatives are assessed the logic of individual arguments fail, since those very arguments can justify war.

  25. @Ron – The test follows Peter Singer who notes that “people can hold all sorts of eccentric positions, and perhaps from some of them it would not follow that death by starvation is in itself bad…Those who disagree need read no further.”

    And you’re right there is certainly some confirmation bias going on here (though, of course, counterfactually we can’t know how much difference it will make). My intention here isn’t to create a rigorous examination of people’s moral intuitions, but rather push people towards the tension that Singer sets up between the drowning child scenario and our behavior vis-a-vis people suffering in faraway places, etc.

  26. The difficulty to me is that in all the given scenarios it is probable that your action will save a life, whereas the final question demands that you act in a way that arguable is unlikely on its own to make any difference (at all). I don’t believe that if I give oxfam £20 in the next couple of days they will change their behaviour at all. In the same way that if I vote in an election it almost certainly will not in itself change how the country’s run. The micromanagement of such a large organisation to invest £20 more or less depending on my donation would be wasteful, and it’s incredibly unlikely that my donation would tip any balance in the spending strategies of Oxfam.

    Assuming I’m right on this, are you really making fair comparisons? My willingness to sacrifice a relatively small cost (my shirt) to save a child’s life who probably would die would not seem to be comparable with making a sacrifice of a similar worth (assuming i paid £20 for my shirt) to donate to an organisation which I think it’s probable will not make a difference.

    If a man came up to me and said, “I am the messiah, if you give me £5 I will end all natural disasters and war.” I would not give him the £5, because I presume the chances that giving him £5 would have his claimed impact are too low. But, if I’ve correctly applied your logic, even if the chances are so so ricidulously low, I would be obliged to give him the £5?

  27. @Anthony – I think that’s a genuine problem – it’s the problem of small effects, basically. James Garvey (TPM’s editor) and I have discussed it in the context of climate change: counterfactually speaking, no particular switching off of a light is going to make any difference to anything, so how can there be a moral obligation to switch off a light, etc?

    But I’m not certain it’s obviously the case that a donation of $50 will not (literally) save a life. So, for example, there are charities that claim that $20 will provide clean water for 20 days. Many thousands of people die because of contaminated water. Clearly the charities that are addressing this problem cannot do so if people don’t contribute. It seems entirely possible that your $20 could actually make the difference between somebody dying and not dying (although you’ll never know it – hence the question about uncertainty in the test).

    But yeah, it is a genuinely tricky issue, I think.

  28. I suppose my stance is best described as sympathetic towards Moral Particularism. This in its strongest expression claims that there are no defensible moral principles. Such principles are at best crutches that a morally sensitive person would not require and indeed the use of such crutches might even lead to moral error. cf
    My own viewpoint for what it is worth is that faced any set of circumstances which demand a judgement for action, one does what is likely to have the best all round effect. Accordingly one does not feel good or bad about what one has decided. Notwithstanding your suggestion that Moral nihilists do not complete the test I had little difficulty in so doing. I always saved the child but sending money, I was not so sure.
    I am not sure like is being compared with like here. Surely when saving a child one is in a “hands on” situation whether the child lives or dies is absolutely and utterly up to you. I cannot see how giving money to a charity compares sufficiently with this, such as to make the demand for charitable donations a logical or necessary outcome.
    I do give money to charity but am always worried that is often misused and or little progress seems to have been made over many years in solving problems which may well be insoluble. Are we just allowing more people to be born that is to say condemning the unborn just to suffer a terrible life.
    It is some time since I read Singer’s paper on saving the child. So far as I remember little or no details were given about the child, sex, age, character, and so on. But these are not necessary for the point Singer wished to make. How ever in the real world as opposed to a Philosophical conjectures and thought experiments, such details could be known to the person observing the plight of the child which could result in a different decision. For instance were it known to the observer that the child in difficulties was the local hooligan who viscously tormented and harmed as many people on his/her estate as possible especially vulnerable people. In fact a widow and her handicapped daughter driven to distraction, had committed suicide as a result thereof. The hooligan’s parents simple but honest people, were near breaking point concerning their child and police and social services seemed ineffective so far as bringing about a change. I suggest in these circumstances the observer may well turn a blind eye to the situation in the pond and walk on. I could not say for certain that I would not also do the same.
    My point here is that we never encounter situations which are identical in every respect. Sometimes it may be better to lie, steal, kill, or break a promise and it is for reasons as I have outlined, I feel that one should act for what seems to bring about if possible the best conclusion for all concerned. This looks like a utilitarian decision, but sometimes events can occur in which the greater happiness is sacrificed for some higher ideal.

  29. Jeff Alexander

    About the bicycle question:
    If I knew I was in an area where bicycles are frequently stolen I would then question whether or not the child was really in distress or if the child was collaborating with thieves.

  30. I have two problems with the scenario.

    1. The analogy is unfair.

    Remember the Feed the World Campaign in the 1980s? It raised enough cash to fix the problem, and sustainably so (that is a regular supply of food was established to feed all of the people who needed it at the time)… but after a few years, the problem grew (literally – population increased!) and the same number of people who were starving before are starving now, despite a regular supply of extra food becoming available.

    Capacity to assist is only superficially covered in Singer’s drowning child scenario. “You are strong and physically capable, etc… but you will have to go home to change clothes.”

    So you do the rescue – but what if the same child you saved throws itself back in the water (30 minutes later when you are passing by again, and now late for work). Are you obliged to rescue a second time?

    Singer sort of covers this by saying that you performed a similar rescue only last week and does that matter? But it is insufficient to leave it there. We do not know if there were consequences, for example: Were you late for work that day too? And did you boss mind? If it keeps happening will it be a problem?

    But of course, yes you help again. And you return home, change etc. But what if it happens again? Amazingly, this child throws itself in the water again!! Just how many more times must you rescue before you give up?
    Should you wait by the pond so as to be ready to rescue the child again and again?

    It seems an unlikely scenario to have to rescue children every day, perhaps for the rest of your life. But, if Peter Singer wanted a fair analogy, he should know there are a lot of starving children in the world (much more tha one unlucky child falling in a pond). In fact, the number of starving people seems to be an ever increasing number – it is simply not be possible for one person to feed them all. Where does your moral responsibility end? Is it saving one child or millions? This is why the analogy is unfair, you do not have the capacity to help millions.

    2. Diminished responsibility.
    In the scenario, there are others nearer to the child who do nothing to help. If they witness you helping, they may assume that it is your job to help. What if you are justifying their inaction and absolving them from feeling they have to help in the future. By assisting, you may actually be creating a responsibility for yourself, to assist on behalf of others. This may be ok with you because you will feel morally superior, but over time your personal capacity to assist may be worn down.

    On these grounds, I really must challenge the validity of Mr Singer’s analogy.

    PS I already support Oxfam (and Amnesty International). But I have long wondered about its effectiveness… the problem is far greater than my individual capacity to assist and the problem keeps growing (with an ever increasing world population).

  31. I find a false analogy in the problem. In the first instances, I answered that I have a moral obligation to save the child. Whether that is true or not would depend upon your particular moral/ethical stance. But be that as it may, the second instances remove the first-person obligation to act and substitute an obligation to pay another person to fulfill any obligation. This raises a sticky issue due to the increase in uncertainty as to the outcome when acting through a third party. The question bears pondering. Thanks for raising it.

Leave a Comment

NOTE - You can use these HTML tags and attributes:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>