Tuesday, February 10, 2009

The Railway Tunnel

I've got sort of an odd new trolley problem, and I'm curious to hear what people think. Post your answers in comments.
You look uphill into a very long railway tunnel and see five men working in the middle of it. You see two of them stand up, hearing something at the far end of the tunnel. “It’s a train!” one of them shouts. “Run!”

The train appears in the distance, outside the far entrance of the tunnel. Next to you, there is a button on the wall that you can push to collapse the scaffolding that is over the far entrance. You can’t see the scaffolding, since it’s on the other side of the tunnel, but an indicator beside the button tells you that one man is working on it. You know that if you push the button, that man will fall to his death and his body will stop the train from going into the tunnel. Whether you push the button or not, you’re safe, since you’re outside the bottom of the tunnel and you can easily move aside.

You see the five workmen, now running down the tunnel as fast as they can. You know that they cannot get out. They are too much far from you, and the train will speed up as it goes down the steep slope.

There was an accident like this many years ago. The bodies of the men in that accident were crushed so badly that they were unrecognizable. You know that each of the five men you see in the tunnel will meet the same fate unless you push the button.

Do you push the button?


Justin said...

I get an extremely strong intuitive response of "boy, I'm glad I don't have to make that decision!" That's a response which I don't usually get to trolley cases.

Btw, what do you think the relevantly different feature of the case is? It seems like pushing the man off of a bridge to stop the train, but without the up close and personal violence.

John Casey said...

This is a tarted-up version of the question, "Would you kill one man if you had sure and certain knowledge that it would save the lives of two others." To which I would respond, no knowledge is sure and certain enough that I would take a life. Give my own, perhaps, but not take one unconsenting.

Dan said...

I think you should push the button, but then again I'd say the same thing about the fatman.

Are you familiar with the work of Josh Greene, Harvard psychologist and one of our nation's foremost trolleyologists? He & his colleagues collected data on a bunch of trolley variations, and found that one of the two features that influenced people's judgments was whether the scenario involved the use of physical force (the kind of violence that our evolutionary ancestors could've used). Most people don't want to physically push the man off the bridge, even if they're pushing him with a long pole (so it's not about direct contact). People are more willing to cause his fall without physical force, like by pulling a lever for a trapdoor, even if they & the lever are up on the bridge next to him (so it's not about proximity). (The other feature that matters is related to the death being a means vs. a side effect.)

Neil Sinhababu said...

For those of you who are curious, here's what I was doing with the case. And yes, it is fat man all over again, except with a different camera angle.

Yes, Dan, I actually heard Greene's stuff at a conference at Dartmouth some time ago. I do think it's relevant to my paper -- you could say I'm positing a mechanism to explain his results.

Dan said...

Interesting idea. Are you planning on re-running this as more of a proper experiment, with people getting randomly assigned to either this version or another version that is as similar as possible except for the balance of vividness? From this scenario alone, we can't tell if vividness is having an independent effect or if you're just replicating Greene's physical force result.

Anonymous said...

My first thought was "What idiot has a work crew in a tunnel that's in use? Why wasn't the train diverted?"

But since my choices are limited, along with my knowledge of the situation, I'd push the button.

Or simply push the guy off the scaffold. The personal use of physical violence doesn't bother me.

My moral compass is probably skewed, but I can't tell.

I'm not anonymous. I'm zak822

Tom said...

It would seem that the choice is to (1) push the button, in which case one person dies, or (2) not push the button, in which case 5 people die.
So if I choose (2) I am in effect responsible for the death of 4 people.
Given that I know nothing about any of the 6 people involved here, it would seem that I would have an obligation to push the button. Of course, in practice the situation is one where one would not have much of a chance to think, and furthermore the nature of the choice may not be very clear. For example, one might think it possible that the 5 men in the tunnel have an alcove they are going to dive into out of the train's path (why else would they be putting themselves in such an obviously risky situation on a track that is being used?). Hence I would not blame any person who picked (1), thinking that it must be the case that there is some such alcove.
Needless to say, knowledge of the nature of the 6 people might rightly affect one's choice. For example, the 5 inside the tunnel might be murderers, and the person outside my daughter. If I knew this to be the case, I would choose (2).
Another consideration is that the 5inside knowingly took the risk that the train might come, whereas the person outside seems to have accepted a job where he could avoid the train's path. Which would imply that it might not be right to push the button.
I am glad I do not have to make this choice--or maybe I should say I hope I do not ever have to make it. I would be tormented by guilt afterwards no matter which option I chose.
One important slant on this hypothesized situation is that I would not consider a person faced with this sudden choice as being guilty for picking either (1) or (2). Perhaps this is as significant a question as "what would you do?". That is, would you consider a person guilty for choosing option (2)?

Neil Sinhababu said...

Good point, Dan, I don't know about that. I'd like to do a proper experiment on this (or at least see somebody do one, I've never run a large-scale one before, and out here in Singapore I can imagine some concern about cross-cultural variation in intuitions).

Dan said...

I thought that someone (Hauser?) did some cross-cultural trolley research without finding much variation, so Singapore might not be a problem. You could also do a properly designed experiment online. You just need to have more than one version of the scenario (to isolate the variable that you care about), randomly (or quasi-randomly) assign subjects to the different versions, and keep some other things in order (for instance, don't let subjects see other people's answers & ask them if they're already familiar with the trolley problem).

The biggest potential hurdle is that you might need to get your study approved by your university's human subjects review board, and go through all of the procedures that they require.

bunny mcintosh said...

Do I love any of these men? That totally dictates my answer.

If not, I don't think I'd kill the scaffolding man. He's the smart one who bothered to build scaffolding instead of making the decision to work willy nilly in a tunnel with an active train in it.

Let one of the five men throw themselves at the train to stop it, or be a little ingenuitive. A train that can be stopped with one man's body can be stopped fairly easily.

Neil Sinhababu said...

Yeah, I tried to put in the thing about the train picking up speed as it goes downhill to deal with the one self-sacrificing guy solution.

Jeremy said...

I'd say no.