General forum:

Back    All discussions

2009-12-09
Switching the Trolley
I find these results interesting:

Trolley problem (five straight ahead, one on side track, turn requires switching, what ought one do?): switch or don't switch?

Accept or lean toward: switch  635 / 931 (68.2%)

Other  225 / 931 (24.1%)

Accept or lean toward: don't switch  71 / 931 (7.6%)


In Fiery Cushman's and my survey of philosophers' attitudes about moral dilemmas, we asked about this case and our results don't line up very well with yours.  Here's the prompt:

You are standing by the railroad tracks when you notice an empty boxcar rolling out of control.  It is moving so fast that anyone it hits will die.  Ahead on the main track are five people.  There is one person standing on a side track that doesn't rejoin the main track.  If you do nothing, the boxcar will hit the five people on the main track, but not the one person on the side track.  If you flip a switch next to you, it will divert the boxcar to the side track where it will hit the one person, and not hit the five people on the main track.

The response scale ran from 1, labeled "extremely morally good", to 7, labeled "extremely morally bad", with 4 labeled "neither good nor bad".


Looking just at the responses when this was the first question in our survey (and thus not subject to order effects -- which we did find, even for philosophers with PhDs and an AOS or AOC in ethics) we found the following:

Rating toward the "good" end of the scale: 57.5% (96/167)
Rating "neither good nor bad": 19.2% (32/167)
Rating toward the "bad" end of the scale: 23.4% (39/167)

"Don't switch" is obviously the big difference here (chi-square p < .001).

One possibility is that the response options just don't map very well onto each other: "Don't switch" isn't quite the same as saying that it would be "morally bad" to switch.  Another possibility is that our respondents are demographically different in some key respect.  Still another possibility -- what I lean toward -- is that the more vivid the presentation of the case, the more one finds don't / bad responses.  Your presentation is minimalist; ours is bare but has at least some narrative structure.  Imagine a Hollywood movie with the hero standing by the switch and close-ups of the victims on the tracks.  Conjecture: A large minority of viewers, maybe even a majority, will think the hero is doing something wrong if she flips the switch.  This might also fit with Greene's findings that people with more activation in emotion-related brain areas tend to give more deontological judgments about moral dilemmas.


2009-12-09
Switching the Trolley
Hi Eric.  You say: " "Don't switch" isn't quite the same as saying that it would be "morally bad" to switch. "  But the key part of the question was "what ought one do?".  I probably would switch but I don't think one ought to switch.  I think you are right that the differences in narrative structure partially explain the different results.  However, that can't be the whole explanation.  The target group is likely to have been exposed to the narrative on several occasions in the past.  Hence, the trolley scenario should be quite vivid to them.

2009-12-10
Switching the Trolley
Yes, I agree.  But exposure to the narrative might not be quite the same has having the scenario vividly in mind.  Past exposure to the narrative didn't seem, in Fiery's and my data, to reduce order effects, for example.

2009-12-10
Switching the Trolley
In the trolley case, I’m the conductor, doing my job as best I can. I have a responsibility to make a decision. The decision to switch tracks, or not to, is mine to make, within my authority, one I can make. Even if people disagree with my decision, nobody can ask me why I got involved .
I was in charge.

In the box car case, as you describe it, I’m a bystander who happens to be standing by the tracks. I have no authority or responsibility in this situation; indeed it’s probably against the law for me to flip that switch. Consequently I’m much less inclined to get involved. All things being equal, I may be less inclined to switch the tracks than I would be if I were the trolley’s conductor.

2009-12-10
Switching the Trolley
Reply to Jim Stone
It's possible, then, that different people read the original question differently.  The original question was:  "Trolley problem (five straight ahead, one on side track, turn requires switching, what ought one do?): switch or don't switch?".  I think I was mostly inclined to read it in the way described by Eric.  But I would have given the same answer on the other reading.

2009-12-10
Switching the Trolley
Looking online at descriptions of the trolley case, I can’t find anything placing me on the trolley. I think my version comes from Philippa Foot and I don’t remember how she says it. It’s been a while. I believe there was an accompanying case where one is piloting an airliner crashing into the center of a populous city. One has just enough control to steer for a suburb.   Perhaps I just ran the two together in my mind over the years. 

2009-12-11
Switching the Trolley
Reply to Jim Stone
It would be interesting to see if the airliner case would yield different results.  However, I still think you made a good point.  It's possible that different people read the question differently, which may explain why the numbers in this case differed from those obtained earlier by Eric and Cushman.

2009-12-11
Switching the Trolley
Thank you, yes it's possible they are being read differently.

FWIW, I do think that, if I'm the pilot of that airplane, I ought to steer for a less populated area.
Supposing I come down in the center of the city and miraculously survive, and people
ask me 'Why didn't you steer for a sparsely populated suburb? You killed five times more
people than you had to!' I honestly don't know what I would say.

2009-12-11
Switching the Trolley
Reply to Jim Stone
In that scenario I would not know what to say either.  Perhaps I would point out that I didn't actually do anything, hence I didn't actively kill anyone.  But I agree, it's tricky.  But if I did steer for a sparsely populated suburb, and the mother of a dead child asked me, "why did you decide to aim at our suburb, I lost my husband and my little child", I would not feel good about saying "If I had done nothing, five times more people would have been killed. So, I aimed at your suburb".

2009-12-11
Switching the Trolley
Thanks for those comments, Jim and Berit!  I'm inclined to agree with you.  Unfortunately, Fiery and I don't have an airline-like case among our stimuli.

2009-12-13
Switching the Trolley
I find the airline case very interesting, because I've always been inclined to let the trolley roll on and kill the five, but here I feel a strong push towards steering the plane for the suburb. It always seemed to me that in switching the trolley track you rope someone innocent into a pre-existing (albeit horrid) situation. The real fact is that a trolley is heading for five people. It is the person who contemplates switching the tracks that brings the single bystander into the situational web. But why doesn't this intuition transfer over to the airline case, encouraging me to plow into the city center? It might be this: The plane isn't, as it were, on rails that head towards the people in the city center. There's a sense that it could crash anywhere, it hasn't been set up to head that way in particular. So I feel more comfortable in trying to minimize damage by diverting it. Is that what gets rid of the sensation of dragging innocent people into the equation? Help needed!

2009-12-13
Switching the Trolley
Reply to Sam Coleman
I wrote this on another thread and, as I went to the trouble of doing it, I thought I would copy it here, in the hope that some people might find it helpful. Apologies for telling you what you already know.
        .................................
The Trolley Case (what's that?):

A. I’m the conductor on a runaway trolley which is on a track such that it will kill five innocent workmen unless I switch to a sidetrack, where it will kill only one workman.

This is often coupled with another case,

B. The same scenario but there is no sidetrack. However there is a very fat man on the trolley and if I throw them over the front, his body will brake the trolley, so that it stops before it kills the five.

Many people feel that, though both A and B have the same cost benefit analysis, one dies, five live, A is permissible but B is not. In the first case I’m not trying to hit the lone workman on the sidetrack. He death isn’t part of my plan. His death is the foreseen but undesired side effect of what I do to save the five, namely, switch tracks.

Jeremy Bentham said that a consequence of an act is ‘obliquely intended’ when it is foreseen but not desired. The workman’s death is obliquely intended.

In case B., I’m trying to run over the fat man with the trolley, as a means of saving the five. Hitting him with the trolley is part of my plan. Bentham said that an effect of a voluntary act is ‘directly intended’ when it is foreseen and desired both. The harm I do the fat man is directly intended, not for its own sake, but as a means of saving the five.

This goes to the Doctrine Of the Double Effect, a medieval Roman Catholic doctrine, according to which it is sometimes permissible to bring about by oblique intention what it would be wrong to bring about by direct intention.

A consequence of the doctrine is that A. is permissible, B is not. The doctrine, which is quite controversial, is applied in many practical circumstances, including medical ethics and the Catholic doctrine of The Just War.

The cases also go to the ‘consequentialist’ doctrine that the consequences of actions are all that determine their moral value. Utilitarianism is such a doctrine.  As both A and B have the same Consequences, one dies, five live, if we decide that they have different moral worth, we reject consequentialism.

...............................................

I add this:

The force of the doctrine in the trolley case is that switching tracks is permissible. Especially I am not trying to hit the lone workman with the trolley. If somehow he escapes, I saved the five in precisely the way I meant to. There was nothing I was trying to do that I failed to do. His death is desired neither as an end nor as a means. This helps explain why switching tracks is permissible, according to the doctrine. The intentions with which I act make a moral difference.

A Roman Catholic doctor, treating a dying patient whose death is imminent and who is in serious pain, may treat aggressively for pain, even to the point, if it is necessary for pain control, of suppressing breathing and killing the patient. The idea is that as the medical mission of saving life is justly abandoned, because it cannot be accomplished, the mission of relieving pain is all that remains. So she can knowingly kill her dying patient as the foreseen but undesired side effect of giving adequate pain control. But she cannot try to kill him. If she gives the adequate dose of morphine and he doesn’t die but is simply relieved of his pain, there was nothing she was trying to do that she failed to do. His death wasn’t part of her plan though she knew she would kill him.
However the doctor who, in precisely the same circumstances,  gives the same dosage of morphine because he wants to kill the patient to put her out of her misery, is trying to kill the patient, which is forbidden. In short, there may be circumstances where it is permissible to kill the dying patient but there are no circumstances where it is permissible to intentionally kill him.

(As Roman catholic medical ethics became part of professional medical ethics proper, it has long been part of standard medical practice that there can be circumstances where it is permissible to knowingly kill a patient and also to permissibly let a patient die whose life could be saved ( where the treatment would pose a 'grave burden' to the patient), even though killing or letting die with the intention that the patient will die were forbidden. Within the prohibition on euthanasia there has been considerable latitude on killing and letting die.)

In fighting a just war, we may bomb the enemy’s munitions plant which is placed near a day care center. When the munitions plant blows up it will kill the children in the day care center. The deaths are foreseen but not desired, obliquely intended, so permitted. But suppose we learn that the dictator’s child is at the day care center and we know him well enough to know that if the child dies he will lose all will to fight and will sue for peace. We are forbidden to target the day care center itself, because then the death of his child, who is innocent, is directly intended. Knowingly killing innocent people in wartime can be permissible, but intentionally killing innocent people is murder and is strictly forbidden.

In such cases The Doctrine of the Double Effect is combined With the Doctrine of Proportionality. Civilian casualties must not outweigh the value of the military target. The idea is NOT that we can permissibly kill lots of innocent people as long as we do not directly intend their death. Killing must not demonstrate ‘a reckless disregard for human life.’ If it does, killing is murder even if the deaths are not desired. If I use a hydrogen bomb in a crowded city to take out a radar installation, thereby making sure to get it on the first try, not caring one way or another about the million people I kill by oblique intention, I commit murder according to the Doctrine of Double Effect.

Personally I believe it is permissible to switch tracks for precisely the reasons the doctrine captures. I’m not trying to kill anybody, the death is the foreseen but undesired side effect of what I do to save the five. No one is being used as a means, no one is being exploited. I think the intuition to the contrary tends to flow from viewing the killing of the lone workman as if he is being targeted in a way that he isn’t. I'm just trying to get off the main track in the only way I can.

If switching tracks is permitted, I think one can argue that it is what I ought to do, in the following way:
if I’m in a situation in which I will do serious harm no matter what I do, and there is a morally permissible alternative which involves doing far less harm than I will otherwise do, then I ought to take it. All things being equal, I ought to do as little harm as possible.

2009-12-14
Switching the Trolley
FWIW, I believe that the original trolley case came from Foot and was an attempt to do a version of the airline pilot case without uncertainty.  (Rails add certainty to trolley cases.)  So on that understanding of the case, the person switching is in some sense already driving and hence already on track to kill someone.

Subsequent discussions seemed often to turn the person into a bystander by a switch.  Such a bystander is not on track to kill someone unless she switches the trolley. 

I've always thought that is a relevant difference in that the first sort of scenario is kill one vs. kill five.  The second sort is kill one vs. don't prevent deaths of five.

Recently Judy Thomson has espoused a similar view in response to some input from one of her former grad students.  See: JUDITH JARVIS THOMSON, "Turning the Trolley," Philosophy&Public Affairs, 36 NO: 4 (2008): 359-374.

This was why I gave "need more info" or one of the "other" options as my answer to this case.

2009-12-16
Switching the Trolley
Interesting!  However, if by 'kill' you mean 'intentionally kill', I don't see the first sort of scenario as a kill one vs. kill five scenario.  If you flip, then you will kill one.  If you don't do anything, then you will let five people die.  Perhaps letting five people die is worse than intentionally killing one, but, as I see it, it's not a kill one vs. kill five sort of scenario.

2009-12-16
Switching the Trolley

Hi Berit,  I'm not sure we're in disagreement, but no I don't mean "intentionally kill."  There are two cross-cutting distinctions, at least on one interpretation of the killing/letting die distinction and the intended/unintended distinction, employed by the doctrine of doing and allowing and the doctrine of double effect respectively. 

One way of distinguishing killing from letting die is by whether we are in the causal chain leading to the death or not.  Another is whether we in some sense initiate the threat -- as opposed to merely diverting a threat away from one person to harming another.  We could do either one of those without intending a death.  And relevant to this discussion, the first way but probably not the second will count a bystander diverting a trolley as killing someone.  Both would count the driver of the Trolley as killing someone whichever way she drove in a normal trolley case. (There is a large literature on how to draw the killing/letting die distinction, and I'm no expert on it.)

On one interpretation of double effect, what we intend is what we aim at, either as a means or as an end.  On this interpretation it is possible to aim at a death without killing.  The standard example from a Foot paper is letting someone go untreated to use his/her organs for other purposes.  (In a first paper she counts it as a killing but later she says that was a mistake and this was not a killing but a death that double effect would call intentional.)

My point was precisely that the way one should classify the cases might depend on whether you are a bystander or the driver of the trolley.  The driver kills someone either way (on the reading where being in the causal chain leading to a death counts as a killing), whereas a bystander only kill if she puts herself in the causal chain by flipping the switch.  On some moral views this difference matters.  And if you thought it mattered you would have chosen one of the "other" options in the poll.  Eric's polling question, if I recall correctly (I can't now see the whole thread) included information to make clear which of the two sorts of cases were being asked about and hence offered a scenario in which some of the needed info was filled in.

Maybe that's too much explanation of what I meant . . .


2009-12-16
Switching the Trolley
This is all very helpful and interesting.

I think Philippa says somewhere that killing involves originating a lethal sequence. That is, I start a causal chain that ends in someone’s death. Paradigmatically, I push the boulder down the hill as my enemy passes beneath or  point the gun at him and pull the trigger. But it doesn’t need to be intentional.

If I’m driving the trolley I may have switched onto that track in the first place and accelerated,
thereby creating a potential danger to the five on the track. When the brakes fail, no fault of mine, I still have originated a causal process that leads to their death, so this is killing.

Of course if I’m standing beside the track, a bystander, I do not accelerate the trolley down that track and so if it kills the five, I originate no lethal sequence so I do not kill them. I merely permit a lethal sequence already in progress to run its course.

If I the driver switch tracks so as not to kill the five, it does seem that I originated the lethal sequence that kills the one, I simply change its direction so that it will kill fewer people.

If I’m the bystander and I switch tracks, I’m not sure exactly how to assimilate this to the talk about ‘lethal sequences’. Surely I killed the one workman on the track that I switch to. Suppose nobody is on the main track, nobody is in danger, and I switch the track because I hate that particular workman and wish him dead. Then arguably, by directing toward him a causal process already in progress, I originate the lethal sequence that kills him. So perhaps we can say that, even in the case where the bystander throws the switch that shifts the trolley to the sidetrack, I originate the lethal sequence that kills him, as he was in no danger before. But I don’t think I like this very well.

We might say that originating a lethal sequence is sufficient, but not necessary, for killing. Anyhow, suppose that, seeing the rock rolling down the hill toward the innocent hiker at the bottom, I throw a log across its route which stops it. Then I remove the log, allowing the rock to kill her. This is an example of letting die. I allow a causal sequence already in progress, that I didn’t originate, to continue. Same if I refrain from interfering in the first place.

So suppose the patient is terminally ill with cancer and goes into pulmonary arrest and I refrain from putting him on a respirator.  Then I let him die. Or suppose I put him on it and finally remove the respirator and allow the causal sequence leading to his death to continue. Then I let him die. The cancer does the killing, not the taking him off the respirator.

Consequently removing Terri Schiavo’s feeding tube was killing, not letting die, because there was no lethal process already in progress that killed her. She didn’t die of PVS but of starvation because we removed the feeding tube. This began the lethal sequence. I believe the court order that permitted withdrawal also forbade efforts to spoonfeed her.

Hence the claim that Schiavo’s death was the foreseen but undesired side effect of relieving her of the burden of the feeding tube is implausible. The feeding tube was no burden. It was seized upon as an opportunity to end her life. If a lethal injection have been legal, they would’ve left in the feeding tube and killed her that way. So removing her feeding tube wasn't just killing, but intentional killing.

Letting someone innocent die can be permissible, according to the doctrine of the double effect, but letting them die intentionally is always forbidden.  We cannot allow the lethal sequence to continue because we want her dead, either as an end or as a means. The physician who withdraws burdensome treatment because he judges it futile, needn’t intend the patient’s death. The death may be the  foreseen but undesired side effect of what he does to save her the pain and expense of a futile therapy. If somehow she goes on living there is nothing that he tried to do that he failed to do. Similarly killing someone innocent can be permissible, but killing them intentionally is forbidden. As long as their death is part of the plan, it doesn’t matter whether we’re killing or letting die–both are forbidden. As letting die needn't involve the intention that the patient will die, typically letting die isn't 'passive' euthanasia, since euthanasia requires the intention that the patient will die.

And yes, back to the trolley, in neither case is anyone killed intentionally. If I take a stiff drink in order to get tipsy, knowing that I will later have a horrible hangover, the same act can be described as getting tipsy and producing a hangover. But I intend to get tipsy, not to produce a hangover. 

2009-12-16
Switching the Trolley
Reply to Jim Stone
Jim,

If initiating a lethal sequence is the right analysis, then it does look like switching isn't killing. (Although what exactly counts as doing that has to be tricky since diverting a rock onto someone normally counts as killing them even if it was already rolling.)  But that isn't the only analysis out there. 

I think the Schiavo stuff is a red herring, since one could just as well stop connecting the feeding tube to the source of food and get the same result.  In that case you wouldn't be initiating anything. 

More importantly, I take the underlying morally relevant distinction to be one between harming/refraining from helping, where this usually lines up with killing/letting die. Feeding someone against their intentions or the intentions of her guardian acting on his conception of her intentions seems like intitiating something that harms her.   But that is controversial and that controversy precisely involves whether death is a harm to her and whether being force fed is a harm to her. Because there are different views on how and whether it lines up in the Schiavo case, it isn't a good case for getting clearer on the underlying distinction.

Anyway, my points aren't so much on the merits as about the differences between the present survey and Eric's survey.  Eric's seems to have given more detail and hence to have given materials with which people could decide whether the death counted as a killing that were left out of the philpapers survey.  That was my main point in this thread.

2009-12-17
Switching the Trolley


Well, I think the Schiavo case is just the sort of thing that these distinctions help sort out, and indeed were meant to help sort out. This is precisely where we need to get things clear. Suppose I knowingly stop feeding my baby and she dies of hunger and thirst. I have plenty of food and so on. I say that I merely allowed her to die. Obviously that’s false. There was no lethal sequence in progress, like cancer, that I allowed to continue to its conclusion. She died simply of starvation and dehydration because I didn’t feed her. I initiated that lethal sequence by withholding food.
So this is killing, not letting die.

Terri Schiavo wasn’t dying. There was no lethal sequence in progress that was endangering her life. The lethal sequence was initiated by the withholding of food and water, both artificially delivered and by mouth. She died of starvation, like my baby in the example above, not of PVS.
So she was killed. Nothing immediately follows about the morality of killing her. The point is just this: withholding food and water from people until they die as a consequence is killing them.

‘I think the Schiavo stuff is a red herring, since one could just as well stop connecting the feeding tube to the source of food and get the same result.  In that case you wouldn't be initiating anything.’

If you stop connecting the feeding tube to the source of food, you would be initiating the lethal sequence of starvation and dehydration by withholding food and water, just as I do when I withhold food and water from my child.  Of course that way of going about killing Schiavo was legally impossible.

‘More importantly, I take the underlying morally relevant distinction to be one between harming/refraining from helping, where this usually lines up with killing/letting die. Feeding someone against their intentions or the intentions of her guardian acting on his conception of her intentions seems like intitiating something that harms her.   But that is controversial and that controversy precisely involves whether death is a harm to her and whether being force fed is a harm to her. Because there are different views on how and whether it lines up in the Schiavo case, it isn't a good case for getting clearer on the underlying distinction.’

I think it is a good case for getting clear on the distinction. Arguably there are cases where death is in someone’s interest. Suppose there are such cases. Killing these people, therefore, is in their interest. That doesn’t make killing them any less killing or any more letting die. Killing is killing even if death is a blessing. Continuing to deliver hydration and nutrition, either artificially or by mouth, to such a person may harm her interests, and withholding food and water may further her interest, but  starving her to death by withholding food and water is killing her just the same and it’s not merely letting her die. It really is worth being pretty straightforward in such cases, or else everything is going to turn rapidly to fudge and a moral discussion becomes impossible.

Suppose my down syndrome newborn has a malformed esophagus, he's starving,  and I refuse to allow the routine surgery that would open it, because I want him dead so that I can start again with an able baby. Then I am not killing him, I am letting him die because I want him out of the way. It may be definitely in his interest to go on living, families may be lined up 10 deep wanting to adopt him, but I’m still not killing him. I’m not initiating a lethal process or diverting another lethal process so that it focuses on him. I’m permitting a lethal process already in progress to continue, when I could easily stop it, because I want him to die. I don’t kill him, I do bring about his death intentionally. Bringing about someone’s death intentionally is forbidden by the doctrine of double effect.

The question, I submit, is not whether Terri Schiavo was killed but whether she was killed intentionally, whether her death was part of the plan. Or was she just killed where her death was a foreseen but undesired side effect of removing the feeding tube. In the first case, if she was killed because death was thought to be in her interest, one has a case of euthanasia. Now it is terribly hard to believe that her death wasn’t intended, doubly so as the court prohibited attempts to hydrate her by mouth.

On the face of things she was intentionally killed, probably because death was thought to be in her interest or what she wanted. As a lethal injection wasn’t legally available, the husband seized upon the only legal means of killing her he had, starvation.
In such cases two sorts of arguments tend to be made–first, this isn’t killing, it is merely letting die. Second, the death isn’t intended but is merely the foreseen but undesired side effect of removing the feeding tube. The first is simply false and the second is implausible. While Terri Schiavo had said she didn’t want to be kept alive on machines when she was dying, the feeding tube isn’t a machine and she wasn’t dying.  There is no clear indication that she wanted to be starved to death in this way under these circumstances. It is unlikely that anybody would starve to death someone who isn’t suffering or dying, a process that would take a week or two, by withdrawing  non-burdensome and minimal medical intervention, who didn’t want them to die.

Some may lament that she was killed at all, others may lament that no quicker way of killing her was legal.. Either way we go though, I think what I said above sorts out the facts.

2009-12-17
Switching the Trolley
Reply to Jim Stone
Jim,

Sorry, I shouldn't have said "red herring," since it seems to have been a bit inflammatory.  I don't deny that getting straight on the relevant distinctions and their moral relevance might well be helpful in figuring out what to think about the Schiavo case.  I just deny that it is a helpful case to start with when you're trying to figure out the right things to think because I think there are a number of potentially cross-cutting possibly relevant distinctions that the case makes it easy to run together and that we might wish to keep apart for purposes of thinking about what matters.

best,

Mark

2009-12-17
Switching the Trolley
On this question, (after briefly researching the topic) it occurred to me that not-switching is as much of a choice or decision available to me at the time, as is switching. It doesn't strike me that not-switching absolves me of a responsibility that switching does - and I don't think I would be blameworthy for any deaths in either case. I have to choose to either switch or not switch. This is why I selected "lean toward switch" on the survey.  

It could be that the five on one track are all evil villains while the one on the other track is a brilliant doctor who holds the cure to a deadly widespread disease. But the survey didn't mention this, so I didn't assume it.

2009-12-17
Switching the Trolley
Hi Mark, this makes a lot of sense.  However, wouldn't those who think that being part of the causal chain entails killing hold that it's worse to actively switch than to simply let things proceed?

2009-12-17
Switching the Trolley
Hi Berit,

Yes, if you aren't already part of the causal chain.  That's why on this position it matters if you're driving the trolley (and hence already in the chain) or a bystander (and not in the chain unless you switch).  I've always been inclined to that position myself and Thomson in a recent PPA article takes that position as well if I recall correctly. (I talked with her about it this summer after reading the paper and I believe she would have corrected me if I was wrong, but I urge anyone who cares to look at her article and not take my word for it.)

Again, these issues are difficult and I'm not defending my inclination here.  I'm just saying that if you think this is a relevant difference there is an important difference between Eric's question which fills in the details and the philpapers survey which did not fill in these details.

It is worth noting that even those who think that being in the causal chain matters can also think that at some point the extent to which it matters is swamped by the number of lives saved by switching.

2010-04-14
Switching the Trolley
The original problem has a workman on the side track.  This makes all the difference. A workman assumes the risk of a trolley going out of control, even if caused by human intervention.  The new problem is "sterilized"; it's now just one life against five.

Anthony D'Amato