I subscribe to Adelstein's Substack, so I feel a bit shortchanged to have to give up some Huemer to get more Adelstein. Adelstein is so prolific I can’t keep up with his output. Huemer is so interesting that I would like to have more than I can handle.
I think critics of utilitarianism can quibble with these sorts of thought experiments and intuition pumps on their own terms. But they always silently side-step my major objection, which has to do with the epistemic aspect of the assumptions. An agent who is perfectly well informed about the consequences of their actions would be justified in at least being tempted by a version of utilitarianism. The situation is not clear for ordinary agents, or if it is clear, it is clearly in opposition to even fairly complicated and sophisticated utilitarianism. In a slightly more realistic scenario, there is a non-zero probability that pushing the fat man will make things worse, perhaps resulting in his death together with the five persons in danger. With that slight reframing of the thought experiment, the conclusion is reversed. Perhaps the basic concept remains, but only when agents have sufficiently strong confidence in their understanding of a situation. Such situations might exist, but tend to be of small importance.
Still working through your undoing and suitcase cases, but I’m not persuaded by the arguments preceding those cases.
Background: I am myself unsure about the normative theoretical truth, but I find it plausible that there are (moderate) deontological constraints against certain sorts of actions. That’s to say, it seems to me there are cases in which an action would bring about more utility (or whatever constitutes value simpliciter) than refraining from it, but is nevertheless wrong. That’s mostly a matter of intuition, not inference. I have that intuition about actions like murder, including actions like pushing the person in the footbridge case. That’s why I hesitate to affirm the morality of that action.
That hesitation isn’t reduced by reflecting on the principle that “if it would be good if X happened, you should do X (assuming it doesn’t trade off against doing other things)”, nor by reflecting on claims like “it can’t be a bad thing to put a perfectly moral person in charge of a process” and “Putting people who do the right thing in charge shouldn’t make things worse”. I’ll explain why.
Re the principle that “if it would be good if X happened, you should do X (assuming it doesn’t trade off against doing other things)”. The sorts of deontological intuitions I have about cases of murder – including in the footbridge case – seem to undermine it. E.g., suppose an elderly tenant in a lodging house is not a happy person, and is somewhat ornery to boot, slightly decreasing the happiness of the other people in the house. Two possibilities: (a) he lives another ten years, or (b) a bottle of poison accidentally falls over and spills into his evening tea, killing him tonight. Suppose total utility is slightly higher if (b) than (a). If I understand the principle correctly, it follows that – putting aside irrelevancies, and assuming an accident isn’t in the cards – you should poison him. But that evokes the very sort of anti-murder intuitions that make me hesitate to affirm pushing in the footbridge case. Maybe they’re inaccurate, but my point is that the principle doesn’t seem to add to the case for pushing.
Re the claims that “it can’t be a bad thing to put a perfectly moral person in charge of a process” and “Putting people who do the right thing in charge shouldn’t make things worse”. Insofar as I find constraints against murder plausible, I find it plausible that perfectly moral people would refrain from murder in some cases in which murder would maximize utility. Indeed, the intuition that perfectly moral people would do that is identical to the very intuitions favoring the anti-murder constraint in the first place, with the additional stipulation that the agent involved is morally perfect. So again, it doesn’t seem like anything has been added that might bolster the case for pushing the person in the footbridge case.
Yes, you can always Moorean shift and say "because you shouldn't kill people in the footbridge case, the fact that it's good for something to happen doesn't always mean you should do it." But insofar as that judgment sounds independently plausible, you have some reason to think it's right, which should lower your confidence in the footbridge intuition.
I feel like there must be a misunderstanding. I was just giving an intuitive counterexample to the principle that if it's good (simpliciter) for X to happen then it's good (morally) to deliberately bring about X. And I was pointing out that the intuition seems not interestingly different than the intuition which makes me think pushing is wrong in the footbridge case, so the principle gives me no more reason to think it's right than simply urging that it's right in the first place. I'm a bit unsure whether I find the principle intuitive. It certainly has an air of plausibility. But I'm not sure that's the same as intuitiveness. "Killing is always (at least prima facie) wrong" has an air of plausibility, but that may because we interpret it charitably and don't think about the proposition it literally expresses (which implies that, e.g., killing grass when you mow the lawn and killing skin cells when you scratch an itch are wrong). Maybe we're doing that in this case too -- just thinking about the cases in which it is true. But even if it is intuitive, I'm not sure how much that adds to the case for the morality of pushing the man in the footbridge case. I already find it somewhat intuitive that it's right on the more straightforward grounds that it saves five lives. Adding the principle that if it's good for something to happen then it's good to bring it about doesn't add much beyond that, especially when the same thing which counters the original intuition that saving five lives makes it right also counters the principle.
Okay I think we have different intuitions. When I think about the case, I'm pulled in two directions: the principle seems right, but the counterexample seems (at least pre reflection) right. But this should serve to undercut our confidence in both.
Is the fat man under an obligation to jump off the bridge to stop the train in light of the utilitarian arguments advanced here? I assume there isn’t a utilitarian reason why he shouldn’t jump.
Can you say more about what we’re supposed to imagine in Suitcases? I am specifically interested in the role of consent in that case.
Here’s one way of imagining what’s going on. You’re on the bridge trying to decide whether to push the suitcase. Somehow all six people know the situation and are all yelling “go for it! push it! I give my consent, even if it turns out to be me on the bridge!” (You can hear all six, but can’t distinguish whose voice is coming from where.) In this scenario, I definitely find it intuitively morally better to push. But that seems deeply different than the original bridge scenario. It seems analogous to a bridge case where the big buy recognizes what’s happening and consents to being pushed (suppose he either can’t or can’t bring himself to jump). I also find it intuitively much better to push him in that case. But the more interesting case (at least so far as utilitarianism vs. moderate deontology goes) is one in which you don’t have his consent.
So it seems to me a better suitcase case is one in which all six people in the suitcases are unconscious. That does seem analogous to the original bridge case. In fact, it seems like it basically is the original case, with the uninteresting stipulations that everyone is unconscious and in suitcases. Those stipulations don’t change my intuitions about the morality of pushing – it still seems wrong.
Indeed, imagine a scenario where all six are instead yelling the following from their suitcases: “I don’t know where I am, but I do not consent to being pushed if I’m the one on the bridge – I don’t want to die, even if it would save five people. If I’m not the one on the bridge, then I strongly prefer that you push the one on the bridge, but there’s no interesting sense in which I can consent to you pushing them, because the question of consent doesn’t arise when someone else’s rights are at stake!”. That seems analogous to a version of the original case where the five people are yelling “push him!” but the big guy is yelling “don’t push me!”. And my intuition about both is that it’s not right to push.
But I’m not sure if I’m thinking about Suitcases correctly. Can you say more about what we’re supposed to imagine is going on?
Can you explain what you mean by “decisive” moral reasons? Does it mean the fat man commits a serious moral wrong by declining to jump? Can he rightly be sanctioned or punished for this failure? Does the person who pushed him deserve praise or even reward for saving lives?
Thanks for your very clear and stimulating post. I enjoyed learning from it. Your last answers were more vague and make it less clear what you think. It sounds like you’re hedging. Maybe that’s okay. The conclusion that someone who killed another to increase overall utility should be applauded looks like a reductio. JJC Smart was famous for being a utilitarian always ready to embrace the reductio, but few have been willing to follow him in this regard. They start saying things like, it all depends, or pragmatically it may not be a good idea. Again, thanks for the post.
The common version of the thought experiment involving the option to push a heavy man off a bridge to his death on a train track to stop a train that would otherwise run over and kill several other people is a poor one for at least four reasons, First, even the largest man who ever existed would not be sufficiently massive to stop a train. An oncoming train will knock a truck off the tracks and keep moving. Second, the strongest man who ever existed wouldn't have the strength to push off a bridge anything sufficiently heavy to stop a train. Third, any fool who pushes a man off a bridge in a vain attempt to stop a train would be immediately responsible for a death for which he'd probably be arrested and imprisoned if not executed. Fourth, even if a heavy man falling on the tracks could stop the train and it's morally desirable for that to occur, the morally correct action is for the man to sacrifice his own life by jumping off the bridge onto the tracks. That would save the man who might otherwise push him from being a murderer. Some similar thought experiments are more realistic, but deliberately causing an unconsenting innocent person's death to save any number of other human lives is an injustice.
There can be situations in which one must choose between committing an unjust act to avoid a very bad consequence and declining to commit the act and prevent the bad outcome. Anyone who isn't a Kantian would probably prefer some injustice to a very great evil, though people differ about the amount of injustice that is acceptable to avoid a given amount of evil. The commission of an injustice to prevent a greater evil is called a moral tragedy.
But thought experiments can be unrealistic and legitimate. We can imagine that you have really big muscles, you'll face no legal consequences, and the guy really is large neough to stop the train.
I know thought experiments can be unrealistic and legitimate. However, I object to using unrealistic ones when a realistic one can be used instead. I can easily create a realistic thought experiment to make the same point. Even better, I can describe actual events in which innocent people were killed to prevent a great evil. Don't you think realistic examples are more persuasive?
I’m no expert on philosophy, so here’s a theoretical one that I’ll throw to the experts. Someone is dying of cancer and their quality of life isn’t very good. However, they wish to live—just barely. They’ll be dead in six months. This person is a terrible human being and viciously mistreats everyone around them. They have great power over many people and delight in ruining the lives of others. Is it morally permissible to kill them?
I personally follow Christian philosophy and would say no. But wouldn’t a utilitarian have to say yes?
I subscribe to Adelstein's Substack, so I feel a bit shortchanged to have to give up some Huemer to get more Adelstein. Adelstein is so prolific I can’t keep up with his output. Huemer is so interesting that I would like to have more than I can handle.
I think critics of utilitarianism can quibble with these sorts of thought experiments and intuition pumps on their own terms. But they always silently side-step my major objection, which has to do with the epistemic aspect of the assumptions. An agent who is perfectly well informed about the consequences of their actions would be justified in at least being tempted by a version of utilitarianism. The situation is not clear for ordinary agents, or if it is clear, it is clearly in opposition to even fairly complicated and sophisticated utilitarianism. In a slightly more realistic scenario, there is a non-zero probability that pushing the fat man will make things worse, perhaps resulting in his death together with the five persons in danger. With that slight reframing of the thought experiment, the conclusion is reversed. Perhaps the basic concept remains, but only when agents have sufficiently strong confidence in their understanding of a situation. Such situations might exist, but tend to be of small importance.
Still working through your undoing and suitcase cases, but I’m not persuaded by the arguments preceding those cases.
Background: I am myself unsure about the normative theoretical truth, but I find it plausible that there are (moderate) deontological constraints against certain sorts of actions. That’s to say, it seems to me there are cases in which an action would bring about more utility (or whatever constitutes value simpliciter) than refraining from it, but is nevertheless wrong. That’s mostly a matter of intuition, not inference. I have that intuition about actions like murder, including actions like pushing the person in the footbridge case. That’s why I hesitate to affirm the morality of that action.
That hesitation isn’t reduced by reflecting on the principle that “if it would be good if X happened, you should do X (assuming it doesn’t trade off against doing other things)”, nor by reflecting on claims like “it can’t be a bad thing to put a perfectly moral person in charge of a process” and “Putting people who do the right thing in charge shouldn’t make things worse”. I’ll explain why.
Re the principle that “if it would be good if X happened, you should do X (assuming it doesn’t trade off against doing other things)”. The sorts of deontological intuitions I have about cases of murder – including in the footbridge case – seem to undermine it. E.g., suppose an elderly tenant in a lodging house is not a happy person, and is somewhat ornery to boot, slightly decreasing the happiness of the other people in the house. Two possibilities: (a) he lives another ten years, or (b) a bottle of poison accidentally falls over and spills into his evening tea, killing him tonight. Suppose total utility is slightly higher if (b) than (a). If I understand the principle correctly, it follows that – putting aside irrelevancies, and assuming an accident isn’t in the cards – you should poison him. But that evokes the very sort of anti-murder intuitions that make me hesitate to affirm pushing in the footbridge case. Maybe they’re inaccurate, but my point is that the principle doesn’t seem to add to the case for pushing.
Re the claims that “it can’t be a bad thing to put a perfectly moral person in charge of a process” and “Putting people who do the right thing in charge shouldn’t make things worse”. Insofar as I find constraints against murder plausible, I find it plausible that perfectly moral people would refrain from murder in some cases in which murder would maximize utility. Indeed, the intuition that perfectly moral people would do that is identical to the very intuitions favoring the anti-murder constraint in the first place, with the additional stipulation that the agent involved is morally perfect. So again, it doesn’t seem like anything has been added that might bolster the case for pushing the person in the footbridge case.
Yes, you can always Moorean shift and say "because you shouldn't kill people in the footbridge case, the fact that it's good for something to happen doesn't always mean you should do it." But insofar as that judgment sounds independently plausible, you have some reason to think it's right, which should lower your confidence in the footbridge intuition.
I feel like there must be a misunderstanding. I was just giving an intuitive counterexample to the principle that if it's good (simpliciter) for X to happen then it's good (morally) to deliberately bring about X. And I was pointing out that the intuition seems not interestingly different than the intuition which makes me think pushing is wrong in the footbridge case, so the principle gives me no more reason to think it's right than simply urging that it's right in the first place. I'm a bit unsure whether I find the principle intuitive. It certainly has an air of plausibility. But I'm not sure that's the same as intuitiveness. "Killing is always (at least prima facie) wrong" has an air of plausibility, but that may because we interpret it charitably and don't think about the proposition it literally expresses (which implies that, e.g., killing grass when you mow the lawn and killing skin cells when you scratch an itch are wrong). Maybe we're doing that in this case too -- just thinking about the cases in which it is true. But even if it is intuitive, I'm not sure how much that adds to the case for the morality of pushing the man in the footbridge case. I already find it somewhat intuitive that it's right on the more straightforward grounds that it saves five lives. Adding the principle that if it's good for something to happen then it's good to bring it about doesn't add much beyond that, especially when the same thing which counters the original intuition that saving five lives makes it right also counters the principle.
Okay I think we have different intuitions. When I think about the case, I'm pulled in two directions: the principle seems right, but the counterexample seems (at least pre reflection) right. But this should serve to undercut our confidence in both.
Is the fat man under an obligation to jump off the bridge to stop the train in light of the utilitarian arguments advanced here? I assume there isn’t a utilitarian reason why he shouldn’t jump.
Yes, or at the very least he has decisive moral reasons to jump.
Can you say more about what we’re supposed to imagine in Suitcases? I am specifically interested in the role of consent in that case.
Here’s one way of imagining what’s going on. You’re on the bridge trying to decide whether to push the suitcase. Somehow all six people know the situation and are all yelling “go for it! push it! I give my consent, even if it turns out to be me on the bridge!” (You can hear all six, but can’t distinguish whose voice is coming from where.) In this scenario, I definitely find it intuitively morally better to push. But that seems deeply different than the original bridge scenario. It seems analogous to a bridge case where the big buy recognizes what’s happening and consents to being pushed (suppose he either can’t or can’t bring himself to jump). I also find it intuitively much better to push him in that case. But the more interesting case (at least so far as utilitarianism vs. moderate deontology goes) is one in which you don’t have his consent.
So it seems to me a better suitcase case is one in which all six people in the suitcases are unconscious. That does seem analogous to the original bridge case. In fact, it seems like it basically is the original case, with the uninteresting stipulations that everyone is unconscious and in suitcases. Those stipulations don’t change my intuitions about the morality of pushing – it still seems wrong.
Indeed, imagine a scenario where all six are instead yelling the following from their suitcases: “I don’t know where I am, but I do not consent to being pushed if I’m the one on the bridge – I don’t want to die, even if it would save five people. If I’m not the one on the bridge, then I strongly prefer that you push the one on the bridge, but there’s no interesting sense in which I can consent to you pushing them, because the question of consent doesn’t arise when someone else’s rights are at stake!”. That seems analogous to a version of the original case where the five people are yelling “push him!” but the big guy is yelling “don’t push me!”. And my intuition about both is that it’s not right to push.
But I’m not sure if I’m thinking about Suitcases correctly. Can you say more about what we’re supposed to imagine is going on?
Can you explain what you mean by “decisive” moral reasons? Does it mean the fat man commits a serious moral wrong by declining to jump? Can he rightly be sanctioned or punished for this failure? Does the person who pushed him deserve praise or even reward for saving lives?
Saying someone has a decisive moral reason to do X means they have strongest moral reason to do X.
//Does the person who pushed him deserve praise or even reward for saving lives?//
Well, there's a pragmatic question of whether it's a good idea to praise them, but their action is praiseworthy.
Thanks for your very clear and stimulating post. I enjoyed learning from it. Your last answers were more vague and make it less clear what you think. It sounds like you’re hedging. Maybe that’s okay. The conclusion that someone who killed another to increase overall utility should be applauded looks like a reductio. JJC Smart was famous for being a utilitarian always ready to embrace the reductio, but few have been willing to follow him in this regard. They start saying things like, it all depends, or pragmatically it may not be a good idea. Again, thanks for the post.
To figure out whether a person who takes an act is praiseworthy, you need to know about more than just the act.
The common version of the thought experiment involving the option to push a heavy man off a bridge to his death on a train track to stop a train that would otherwise run over and kill several other people is a poor one for at least four reasons, First, even the largest man who ever existed would not be sufficiently massive to stop a train. An oncoming train will knock a truck off the tracks and keep moving. Second, the strongest man who ever existed wouldn't have the strength to push off a bridge anything sufficiently heavy to stop a train. Third, any fool who pushes a man off a bridge in a vain attempt to stop a train would be immediately responsible for a death for which he'd probably be arrested and imprisoned if not executed. Fourth, even if a heavy man falling on the tracks could stop the train and it's morally desirable for that to occur, the morally correct action is for the man to sacrifice his own life by jumping off the bridge onto the tracks. That would save the man who might otherwise push him from being a murderer. Some similar thought experiments are more realistic, but deliberately causing an unconsenting innocent person's death to save any number of other human lives is an injustice.
There can be situations in which one must choose between committing an unjust act to avoid a very bad consequence and declining to commit the act and prevent the bad outcome. Anyone who isn't a Kantian would probably prefer some injustice to a very great evil, though people differ about the amount of injustice that is acceptable to avoid a given amount of evil. The commission of an injustice to prevent a greater evil is called a moral tragedy.
But thought experiments can be unrealistic and legitimate. We can imagine that you have really big muscles, you'll face no legal consequences, and the guy really is large neough to stop the train.
I know thought experiments can be unrealistic and legitimate. However, I object to using unrealistic ones when a realistic one can be used instead. I can easily create a realistic thought experiment to make the same point. Even better, I can describe actual events in which innocent people were killed to prevent a great evil. Don't you think realistic examples are more persuasive?
No. How realistic a scenario is doesn't matter at all. If a scenario is like the real world that often biases our judgments.
Have it your way. I don't feel like arguing this issue.
I’m no expert on philosophy, so here’s a theoretical one that I’ll throw to the experts. Someone is dying of cancer and their quality of life isn’t very good. However, they wish to live—just barely. They’ll be dead in six months. This person is a terrible human being and viciously mistreats everyone around them. They have great power over many people and delight in ruining the lives of others. Is it morally permissible to kill them?
I personally follow Christian philosophy and would say no. But wouldn’t a utilitarian have to say yes?