Thought Experiments: Pipedreams or Profound Insights?

omni · **Posted:** Fri Feb 08, 2008 2:13 am

The New York Times Magazine, on 1/13/08, featured an article by Steven Pinker, "What Makes Us Want to Be Good?" in which he discusses some recent concepts and researches in the field of evolutionary psychology as they relate to morality. (You may need a log-in to view this article in its entirety.) There is a great deal that is intriguing in the piece, but I want to discuss the "thought experiment" on which much of it is based. Here are Pinker's descriptions of the thought experiments I want to consider:

Quote:

The gap between people’s convictions and their justifications is also on display in the favorite new sandbox for moral psychologists, a thought experiment devised by the philosophers Philippa Foot and Judith Jarvis Thomson called the Trolley Problem. On your morning walk, you see a trolley car hurtling down the track, the conductor slumped over the controls. In the path of the trolley are five men working on the track, oblivious to the danger. You are standing at a fork in the track and can pull a lever that will divert the trolley onto a spur, saving the five men. Unfortunately, the trolley would then run over a single worker who is laboring on the spur. Is it permissible to throw the switch, killing one man to save five? Almost everyone says “yes.”

Consider now a different scene. You are on a bridge overlooking the tracks and have spotted the runaway trolley bearing down on the five workers. Now the only way to stop the trolley is to throw a heavy object in its path. And the only heavy object within reach is a fat man standing next to you. Should you throw the man off the bridge? Both dilemmas present you with the option of sacrificing one life to save five, and so, by the utilitarian standard of what would result in the greatest good for the greatest number, the two dilemmas are morally equivalent. But most people don’t see it that way: though they would pull the switch in the first dilemma, they would not heave the fat man in the second. When pressed for a reason, they can’t come up with anything coherent, though moral philosophers haven’t had an easy time coming up with a relevant difference, either.

When psychologists say “most people” they usually mean “most of the two dozen sophomores who filled out a questionnaire for beer money.” But in this case it means most of the 200,000 people from a hundred countries who shared their intuitions on a Web-based experiment conducted by the psychologists Fiery Cushman and Liane Young and the biologist Marc Hauser. A difference between the acceptability of switch-pulling and man-heaving, and an inability to justify the choice, was found in respondents from Europe, Asia and North and South America; among men and women, blacks and whites, teenagers and octogenarians, Hindus, Muslims, Buddhists, Christians, Jews and atheists; people with elementary-school educations and people with Ph.D.’s.

Pinker then goes on to relate fMRI studies performed on experimental subjects considering this dilemma, endorsing the conclusion that the "hands-off" scenario of trolley switching remained in the realm of rational analysis, while the "hands-on" scenario of tossing the fat man off the bridge triggered the "universal grammar" of emotional conflict/morality which makes us averse to man-handling other people. Note that he has emphasized the rigor of the data collection involved with this thought experiment.

While it is always risky to try to confound science with common sense, in this case I think it is warranted.

When I perform this thought experiment, I find myself challenging the credibility of the scenarios: I have time to switch the trolley to another track--but I don't have time to shout out a warning to the five men in danger's way? If there is time to switch the trolley, there must be some possibility that the five men will become aware of impending doom and jump aside. In other words, their deaths are not certain, and by switching the trolley to another track with only one potential victim, I am not playing god--I am merely managing risk.

In the fat-man-on-the-bridge scenario, there is no doubt about the fat man's fate if he is thrown from the bridge: he will die. Yet there is inescapably some doubt about the outcome for the five men: again, if there is time to hoist the fat man and toss him off the bridge, and time for him to fall and land in front of the train, then there is time to attempt a warning and a possibility that the five men may escape the trolley's path.

So it seems to me that the two scenarios vary in an important, realistic way: in the switch case, you have reduced the possible number of victions, but you have not with certainty doomed anyone.

In the fat man scenario, you have without a doubt condemned someone to die. I would expect a reasonable person to find the former scenario (reducing the number of possible victims, who in any case have some chance of escape) acceptable, and the latter (definitely consigning someone to an unjust death) unacceptable.

Frankly, I also doubt that one fat man could derail a trolley! And if he is big enough to stop a trolley, could I lift him? I don't think these are trifling quibbles. Thought experiments may be useful, but surely only if they are as rigorous as every other experiment.

Thought experiments seem only to work as intended when the thinker accepts the tenets uncritically, agreeing to entertain a certainty of outcome that does not exist in the real world.

I'd like to talk about my perspective on this, and, possibly, other "thought experiments." I don't want to go too far astray into evolutionary pscyhology per se, but rather focus on the issue of thought experiments used to generate results that are then treated as rigorous experimental data which can bear the weight of serious hypotheses.