Experimental Philosophy

Europa Malynicz pointed out to me that the BBC currently has a discussion of famous thought experiments in ethics, including Judith Jarvis Thomson’s violinist case, and a few variants on runaway trolley cars. As of this writing, over 12000 people had sent in their votes on the moral status of actions in the examples, and it is interesting to see what this (self-selected, non-random) sample of the folk think. I’ve got some comments on the results below the fold, but I’d rather everyone here went and voted before seeing the votes, so I’ve put them below the fold.

The votes on three of the cases are fairly unsurprising. In Thomson’s violinist case, the vast majority thinks that you aren’t obliged to stay hooked up to the violinist. In the original trolley case, a vast majority thinks it is OK to flip the switch and thereby kill the one in order to save the five. (75.6% say yes.) In the fat man case, where the only way to save the five is to push the fat man off the bridge, only a small minority say it is OK to push the guy and thereby kill the one in order to save the five. (24.43%)

The first thing to note this is that although the group of voters taken collectively has very different opinions about the two cases, it seems opinion is split fairly evenly on whether the cases are alike or unlike. Assuming that only a handful of people say that it is wrong to flick the switch, but right to push the fat man, then only a whisker over 50% of people think the cases are unlike. A lot of philosophical energy goes into explaining just why the cases are unlike, but if this poll is right (big ‘if’) that view is one that is only barely a majority view. That’s even among BBC magazine readers, who are presumably paradigmatic of the kind of folk that we have in mind when we talk of the folk.

The big surprise for me was the result of the fat man stuck in the cave example. A vast majority say that it is OK to dynamite away the fat man who threatens the lives of the people in the cave. In that case 76.1% of people say it is OK. That’s an even larger percentage than in the original trolley case, though the difference is not significant. I’d have thought the cave case went with the fat man on the bridge case. Yet more ignorance of what the folk are like on my part I guess.

What could be driving the differences? Here are three hypotheses, in order of descending charity to the voters. (Descending that is from a pretty low starting point!)

Hypothesis one: By having the hypothetical voter imagine the cave from the inside, you get the voter to empathise with the potential victim. This view holds that people are basically utilitarian, and what is to be explained away is the anomolous result concerning the Fat Man case. The anomoly is explained by noting that making people empathise more with one side (the Fat Man standing right in front of you) than the other (the nameless faceless five stuck out of site in a tunnel) you can draw them away from their natural utilitarian inclinations. So it is all a framing effect.

Hypothesis two: It matters that the voter is to imagine themselves in the cave. Many people have the view that it isn’t OK to kill an innocent person to save a third party, but it is OK, morally OK, to kill an innocent person to save their own life. This view says that people hold that the right to self-defence, even against innocents, to be a strong moral principle, even though there is no such right to defend others.

Hypothesis three: If the guy in the cave hadn’t been so fat, there wouldn’t have been a problem. So he isn’t really an innocent. So rules against killing the innocent don’t count in the minds of some voters in this case.

If those are the choices, I’d choose one. But I’d rather think that I’d missed a more charitable explanation of the voting patterns.

10 Replies to “Experimental Philosophy”

  1. The question in the Trolley case is “Should you flip the switch?”, which is different from “Is it morally ok to flip the swtich?”

  2. Part of the explanation of the difference in response in the two fat man cases is that, in the trolley/bridge case you’re to imagine that you’re a third party, whereas in the cave case you’re to imagine that you’re one of the five who will die unless the fat man is killed. But another part of the explanation of the difference in response is that the fat man is an innocent threat to the five in the cave case, whereas he’s merely an innocent bystander in the trolley/bridge case.

    Let’s factor out the first consideration and focus on the second:

    People tend to think it’s okay to kill an innocent threat to save yourself, but they’re much less inclined to say it’s okay to kill an innocent bystander to save yourself. Nozick has a case in which you’re at the bottom of a well and an innocent person has been thrown down the well and will land on you and kill you if you do nothing (but will survive the fall, cushioned by your body). Here people think it’s permissible for you to vaporize that person with your ray gun in self-defense. But now suppose that a javelin is fast approaching you and will impale and kill you unless you grab an innocent bystander and use him as a shield. People are much less inclined to declare it permissible for you to grab this innocent bystander in order to save your life.

    The fat man in the cave is an innocent threat to the five: he’s (innocently) blocking the five from getting out of the cave.

    Now suppose that, in the trolley/bridge case, you’re to imagine that you’re one of the five whom the trolley will run over unless you shoot the fat man so that he falls off the bridge and stops the trolley. I’m fairly confident that, even if voters were placed in the position of being one of the five who will die if the fat man isn’t killed, far fewer would say that it’s okay to shoot the fat man in this case than would say that it’s okay to blow up the fat man in the cave case. This is because the fat man on the bridge is an innocent bystander rather than someone the position of whose body is endangering the five.

  3. About whether the cases are like or unlike: I think asking the questions one right after another may push people to answer the same way. I should say might push them, since I have no data at all. But it doesn’t seem unlikely, since this presentation does highlight their similarities (by pointing out the paradox).

    Mike’s innocent bystander/innocent threat point sounds convincing. I also wonder whether it makes a difference that Big Jack is one of the people who’s originally endangered. Suppose that the person stuck in the hole is someone who stumbled in from above, got stuck, and is waiting for a crane to come and pull him out (and let’s stipulate that this happened long before you all got trapped by the rockfall). In this case, I feel that it’s somewhat less acceptable to use the dynamite.

    But suppose, in the trolley case, the fat man is originally on the track, and crane lifts him off, but the crane breaks as he’s in midair — is it then acceptable to lasso him back down to block the train? I think not, even though he was originally in danger. So maybe it’s threat vs. bystander that does the work.

  4. I think Mike’s point sounds convincing, at least as a summary of what the intuitions are, but I’m not entirely sure that it is ultimately defensible. I have the intuitions Mike suggests, but I don’t have the ‘theoretical’ intuition that this should matter.

    And the point Matt makes is certainly right as a summary of the intuitions. (I think Unger made this point, and I should have remembered it when writing the post.) If the one is regarded as being part of the ‘same problem’ as the five, then we are much more willing to kill the one to save the five. Again, I don’t think this sounds particularly morally sound, but I agree those are the intuitions.

  5. “In Thomson’s violinist case, the vast majority thinks that you aren’t obliged to stay hooked up to the violinist.”

    It’s too bad that the question is phrased this way. Suppose it were asked—as in fact Thomson asks—whether it would violate his right to life were you not to stay hooked up. The answer might be different; in any case I’d like to see what is said.
    It would be even more interesting to see what is said in (non-coercive) cases where (i) you allow violinists to attach themselves to you, (ii) any violinist that does attach to you is there for 9 months or dead, (iii) you put up a sign that you know is incoherent to violinists that says “please, violinists don’t attach yourself to me”. I wonder whether the general view is Thomson’s here: viz. that it would be nice to stay attached to a violinist that attaches himself to you, but it would violate no rights to detach yourself.

  6. Brian, I agree about the defensibility (though I might give a little more weight to the intuitions than you do); maybe I should clarify that when I said “I feel that it’s more acceptable” I was reporting an intuition, not expressing a considered judgment.

  7. “I have the intuitions Mike suggests, but I don’t have the ‘theoretical’ intuition that this should matter.”

    I agree with Brian and have defended the claim that, notwithstanding intuitions to the contrary, there’s no moral difference between killing an innocent threat and killing an innocent bystander. (Details in “Killing the Innocent in Self-Defense,” Philosophy & Public Affairs, 1994.)

  8. notwithstanding intuitions to the contrary, there’s no moral difference between killing an innocent threat and killing an innocent bystander


    Do you mean that, in the sorts of situations where one’s life is endangered, they are both generally wrong or that they are both generally permissible? The latter seems less plausible, especially in view of Sorensen’s nice “ducking harm” cases.

  9. Mike,

    By ‘no moral difference’ in my comment above I just meant that the killing of an innocent threat and the killing of an innocent bystander are on a par as far as permissibility is concerned.

    But I also argue in the P&PA piece that both types of action are impermissible.

Comments are closed.