Home > philosophy > Would You Press the Button?

Would You Press the Button?

You are riding on a trolley which is moving quickly down a track and you have no brakes. In front of you the track splits to the left and right. On the left track, which is the direction you will take at the split, are five people who cannot get off the track. If you hit them they will die. On the right track is one person in a similar situation. By pressing a button you can switch your path from the left to the right and kill the one person instead of 5. Should you press the button?

Later that day you are riding your trolley again and find the brakes still don’t work. On the track in front of you (there is no split this time) are 5 people. If you hit them they will die. Luckily riding on the trolley with you is a fat friend. If you push him off the front of the trolley he will die but will stop the trolley and avoid the five on the track being hit. Should you push him?

After surviving the two incidents on the trolley you go to work. You are a surgeon and have 5 patients dying from various diseases. The only way to save them is with transplants. In the next room is a healthy person from whom you can harvest the 5 organs needed to save the other five. But you need to kill him to do this. Should you?

Driving home from work that night you come across a person injured in a motorbike accident. He is bleeding and needs to be taken to the hospital or he will die from blood loss. You have just bought a brand new Mercedes (you’re a rich surgeon, remember) and you know it will cost $500 to have the blood stains removed but you can’t see any other traffic on the road. Do you take him to the hospital?

Finally arriving home you relax to unwind after a hard day coping with making unusual moral decisions. But the phone rings. It is a charity wanting $500 to save the life of a child dying from starvation in Central Africa. Do you give them the money?

These are all examples of questions used by researchers trying to study the nature of human morality. The responses people give are quite consistent. Generally people will change the direction of the trolley to save 5 but kill 1. Most won’t push the fat person off the front to save 5 but kill 1. And few will think it’s OK to kill one person to save 5 with transplants. And almost everyone is prepared to spend $500 to repair the car after transporting the injured motorcyclist but most won’t donate $500 to save a starving African child.

Let’s look at these responses and see if they make sense.

In the first three cases the person has to make a conscious decision to kill one person to save 5. When the details are stripped away there is no significant difference between the three cases, but the vast majority of people will press the button but won’t push the fat person or harvest the organs of the patient. Why not?

In every case they must perform a deliberate action knowing fully that one person will die as a result but that 5 will be saved. What is the real difference in these cases? The main one seems to be the directness of the action and existing social norms. Pressing a button is a less direct way of killing someone than pushing a person off a trolley is. And harvesting organs of a living person is against all medical ethical rules despite the fact that there is no real difference between that and pushing the button.

In the second two cases a similar argument applies. In both of these situation spending $500 is required to save a person’s life, but only when the person is lying on the road, bleeding right in front of them will most people act. When the person to be saved lives thousands of miles away the general level of interest is much less.

Interestingly some research has indicated that rich people are often less generous in both cases: some wouldn’t even stop their cars, and many certainly wouldn’t even consider the donation. And groups who claim a greater level of morality are no more likely to actually make the donation than anyone else, although most would at least stop the car.

So what does this all mean?

I think this supports a social evolutionary explanation of the origin of moral behaviour. We would expect that people would gain most from helping their friends and family because their friends are most likely to help them in return, and their family share their genes.

If there was a more absolute, objective source of morality then surely there would be a more universal trend to apply it. After all, if helping other people is considered a moral act should it depend on how close that person is located to you? And if there is an absolute rule that you should not kill does that still apply if you kill one person by a deliberate action but by doing that fail to kill 5 through inaction?

If morality is dictated by a traditional source such as the Bible we would expect to see a major difference in the responses of believers and non-believers. But we don’t. Either morality is specified in these sources but ignored, or it is specified in a way very similar to what most people believe anyway, or non-believers (perhaps subconsciously) follow those rules without acknowledging it.

Of course my answer regarding religious morality would be that religious texts often reiterate the rules which we all know anyway, plus add a few more specific to the religion which are largely irrelevant to most people and in many cases run counter to normal moral beliefs (whatever that means) anyway.

For example, in the Ten Commandments there are various rules which are clearly religiously motivated and completely irrelevant to having a moral life. And the remainder are the sort of thing that any reasonable person would believe anyway. But even then they cannot be absolute. One of those commandments is “thou shalt not kill” yet many Christians will kill, either by joining the armed forces, or in the theoretical case of being on that out of control trolley!

Morality is nothing special. It is just a set of expedient guidelines which we have accumulated over the millennia living as a social species. And note that I said guidelines, not rules or laws. My recent theme continues: nothing should be absolute, and there is no good or bad. I think that makes the world a far more interesting place!

Note. I have discussed the trolley experiments before, and gave some of the statistics regarding the actual measured results in an entry titled “More Morality” back on 2007-11-27.

  1. October 20, 2016 at 4:10 pm

    In case 4, the person has an active role in saving the person’s life and can ensure that appropriate actions are taken. The $500 expense is a *side effect* of the decision. In case 5, the person has a passive role. Spending money is the *active* role that purportedly saves the life. Do they have to research the charity’s financial documents or volunteer there to prove this? Is it a scam? Case 5 is very easy and sensible to reject. It doesn’t really compare to the other cases. There is a conflation on the $500. To illustrate, what if the ‘cost’ in case 4 was “getting blood on your suit that will be $25 to dry clean”, you could not have the same ‘cost’ in case 5.

    The difference between case 1 and cases 2 and 3 is agency. In case 2 and 3, we would only be morally justified in doing so if we had consent. In case 2, we would have no need to push the friend, they could just do it themselves. It is impossible to get consent in case 1, so we are forced to make a pure numbers value judgement. The moral issues of agency do not apply in case 1. When you consider agency, there is a real moral difference between the otherwise identical options “let fate decide” and “choose to kill 5 people”, just not for the people who die.

    What is striking is that this is so intuitively obvious (the research agrees), and yet you couldn’t enunciate this difference. It appears that you just assumed people were choosing illogically. The difference is not the “directness of the action and existing social norms.” Indeed, the universality of this intuition is evidence of absolute, objective morality.

    I’m curious: when you wrote this post, did you feel an intuitive niggling that there *must* be a difference in case 1, but you pushed the thought away? Perhaps your rejection of absolute morality has crippled your ability to see its possibility. If there is such a thing as absolute morality, then it would have to be found in intuition: even non-believers “follow the rules without acknowledging it”. While many I know reject intuition as being invalid and “unscientific”, I see no reason this isn’t just an arbitrary metaphysical assumption.

  2. OJB
    October 21, 2016 at 2:38 am

    I think you are missing the point. Remember that these are only supposed to be examples of real situations which try to examine people’s underlying motivations. For example, you say the fat friend should choose to jump himself, but what if he was looking the wrong way and there was no time to get his permission? You see, the situation can be easily changed to get back to what the real point of the question was.

    I think your objection to my perceived disinclination to accept an objective morality stems from the meaning of “absolute” in this case. Many people claim that if such a morality exists it must come from a god of some sort. I say that it exists (in a weaker form) and arises as a meta-phenomenon of shared conscious thought.

    In other words, most people know what is right and wrong because those rules for living together have naturally arisen as a consequence of hundreds of thousands of years of social evolution. No god required.

    • October 21, 2016 at 12:22 pm

      When I wrote my comment, I considered that scenario. It doesn’t apply.

      If there was no time to get his permission, then you would be forced to choose “let fate decide” (rather than “choose to kill 5” or “choose to kill 1”). The amount of time you have available does not change the moral judgment any more than the distance changes it. As long as you have the theoretical ability to get consent, you must attempt to do so. This remains the key difference between 1 and 2.

      If the 5 people in the operating room were on the threshold of death and the healthy person was asleep, you would still have to wake the healthy person up, wait for them to become alert, and ask for their consent. If that took a minute too long and the 5 died, then that would be the unfortunate consequence. You would *never* be justified taking a knife to the healthy person without attempting to gain their consent.

      (If you think the real point of the question was to trick people into answering in a way they didn’t intend to, then we have a problem. We must consider how people really think, not what the researchers were trying to do.)

      My objection is that you are arguing against what is a plain difference between case 1 and 2, and I want to understand why. It is not a difficult distinction to make for most people, they do it without even understanding why. But you arrive at the wrong moral conclusion and you spent a lot of time thinking about it!

      I would like to discuss the concept of “intuition” and how it applies to this discussion, but I can’t if you reject the premises leading up to it. If you accept the possibility of intuition, the whole argument on objective and absolute morality changes. It’s difficult to discuss the difference between your almost-absolute social morality and the religious absolute-absolute god-based morality otherwise.

      You’ve convinced yourself that the researchers have uncovered an inconsistency that supports the idea of social norms determining group morality. You could be right, but not because of this study.

  3. OJB
    October 21, 2016 at 10:31 pm

    OK, so you say that the time available makes no difference and you say that you must try to get consent even if there is not time to successfully do that, otherwise let fate decide. But that is not the judgement everyone would make and it is unclear whether that is a morally sound judgement anyway. If you don’t act 4 additional people will be killed. Can that be justified morally? I don’t know.

    The same applies to the operating room scenario. Few people would see taking the one life to save 5 in that situation as morally defensible, yet it is identical to earlier scenarios – which they would allow – in its outcome.

    This is the whole point. Morality doesn’t make rational sense and it is inconsistent. I would say that the fact that moral judgements depend so much on immediacy, and social constructs shows that they are derived through social evolution.

    I think the point of this thought experiment is simply that: to show that morality is inconsistent and irrational. But in the social environment people live in that is probably entirely appropriate.

    • October 22, 2016 at 2:06 am

      There is another possibility: That morality is just complex. I see the following moral issues at play in these scenarios:

      (A) The needs of the many outweigh the needs of the few
      (B) A person has a right to agency.
      (C) Whether or not it is moral to take a life when voluntarily offered, that is, is sacrifice moral.

      In general people treat (B) as the primary of these, so in case 2 and 3, they will usually opt for the “let fate decide” option. However, a minority will decide (A) is more important than (B) and will do as you say and throw the friend under the bus. In the operating room you don’t have the imminence and (C) comes into play.

      When you consider all three moral issues, the differences between the three cases become highly rational and consistent. There is a single rational explanation for differences most people feel.

  4. OJB
    October 22, 2016 at 4:18 am

    Yes, I can broadly agree with those factors. But in the first three cases all three factors are more or less equally valid. For example, A (which is roughly a utilitarian approach) is the same for the first three cases (switch the path, push the friend, sacrifice the person). Apart from a small degree of directness all three factors are the same for B as well. And I don’t see that C is really relevant to any.

    So I still can’t see any real difference between the three cases, yet people’s reactions to them are different. To me it’s a clear sign that morality is a socially defined phenomenon.

    • October 22, 2016 at 5:06 pm

      When I argue online, I argue my position with some rigidity. But I mull over what my opponent says for weeks, months, or longer. And either my view is strengthened or I yield the point and change my mind. I say this because I’ve twice stated why B is different from A. I’ll address it once more and then move on.

      In 1, the trolley operator *cannot* take agency into account and is forced to make a pure utilitarian decision. It is impossible to gain consent, so B cannot come into play. In 2 & 3, it is quite possible to gain consent. That is not “a small degree of directness”, but a fundamental moral difference.

      Yes there may be a time factor involved, but that doesn’t change the moral requirement to gain consent. If you think a person should lose agency on the basis of not having enough time to grant it, then you will push your friend. But most people just don’t think that way.

      Why do some people think 3 is more acceptable than 2? Because that scenario has an implied social expectation of “informed consent”. This satisfies the underlying moral requirement of B. 2 does not include implied consent.

      C has three options: all sacrifice is immoral, sacrifice is only moral if it is self-sacrifice, or all sacrifice is always moral. This explains some variability in the answers to 2 and 3.

  5. OJB
    October 22, 2016 at 8:55 pm

    I commend you for the attitude of stating your opinion but being prepared to change it after sufficient consideration. I would hope that is the way I would act too. The problem is not that I refuse to see the truth in what you say, but that I disagree with your interpretation. I realise these psychology and social science problems are rarely straightforward and often have no obvious answer, so this situation isn’t unusual.

    Anyway, here’s why I disagree with you when you say that 1 is different form 2 and 3. The trolley rider *can* get consent in situation 1 by shouting to the single person “are you prepared to die to save the 5?”. Is it practical to do that? I don’t know, but I could ask the same about asking the fat friend. There is very little real difference.

    Can we get that point out of the way before we deal with anything else?

    • October 22, 2016 at 9:39 pm

      Yes we can. There is certainly enough ambiguity here to come to different conclusions. It will still make it hard to discuss intuition, but I’ll try anyway. Why not?

  6. OJB
    October 22, 2016 at 9:54 pm

    Sure, why wouldn’t we discuss intuition? It’s clearly a factor here.

    • October 24, 2016 at 1:54 am

      I’ve tried to respond and I have to admit that I’m not able to. I’ve read material by other authors on the topic who are probably much more qualified. Maybe I’ll revisit this later after I’ve had time to work out a more convincing viewpoint.

  7. OJB
    October 24, 2016 at 2:48 am

    Sure, you’re welcome any time to discuss any of the issues I blog about. I do appreciate your generally rational approach.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: