Friday, October 28, 2011

Help with Thought Experiments

In my experience, it is initially difficult to get students to see the validity of using thought experiments in philosophy. For example, in my ethics courses we discuss the argument from analogy Peter Singer uses to help make the case that we are obligated to donate to famine relief, comparing the costs and benefits of saving a drowning child with those of saving the starving child. Usually some student will try to change the parameters of the experiment. For example, after saving the child a student stated that he would simply ask the child's parents to reimburse him for the cost of his clothes that were ruined by going into the shallow pond to save the child.

One thing I now do to motivate thought experiments and get students to "follow the rules" is compare them to the type of experiments a physicist or chemist might do in the lab. A scientist sets up certain conditions in order to explore reality, test a hypothesis, and so on. In order to get the right data and draw sound conclusions, she will set certain boundaries and engage in particular procedures. Similarly, when we engage in thought experiments we limit options available to agents in the experiment and set up certain rules in order to clarify a concept or test a philosophical theory, even if in so doing we are not talking about something that is likely to happen in "real life".

I'm curious if others have found additional ways to deal with problems in using thought experiments in the philosophy classroom. If you have some helpful tips, please share them in the comments.

Finally, for help on teaching the content of some thought experiments, see these 60 second videos from the Open University. Perhaps a video representation will also help students respect the parameters set up by the experiment.


  1. I too have found the analogy to lab experiments useful. I find it is particularly effective to emphasize that the parameters of a thought experiment are used to control for certain variables: it's not that you CANNOT change the parameters, it's just that if you do, you'll be testing for something else. I think this helps them see why simply changing the parameters of a thought experiment is not an effective response to the intuitions that it generates, and also improves their ability to use thought experiments themselves in their writing and thinking. (We're probably all familiar with the student who designs a thought experiment that is entirely unhelpful in getting at the issue that he or she wants to explore. Once they understand that the sometimes whacky parameters of thought experiments are designed to control for certain variables, they become better at devising experiments themselves).

  2. Nice audio on thought experiments.

  3. Perhaps echoing Erich: It's common for students (and philosophers!) to criticize thought experiments, especially in ethics, for being unrealistic. What I try to emphasize with students is that in trying to investigate a domain of complex phenomena, it can help to think about simplified, even austere, examples of the phenomena and then introduce complexity as we go. So, yes, the classic trolley problem is unrealistic and greatly simplified, but if we have clear reactions to simple cases, we can then understand better why we react differently as cases become more complex or 'realistic.' Sort of the philosopher's equivalent of motion on a frictionless plane: If we can understand something in simple cases, we can better isolate the roles that various complexities then play in other cases.

    Also, I think it's worth pointing out to students that simple cases are abstractions from a wide range of situations that, however much they vary, have the abstract features in common. So no, there aren't too many actual trolley problems, but there are many actual cases where we need to decide whether to harm a smaller number to avoid harm to a larger number, etc. So the simplicity or austerity of the thought experiment serves to isolate features multiply instantiated in the world.

  4. I think these are all good suggestions. Mentioning actual non-philosophical thought experiments can be very helpful. (I use the example of the frictionless plain, too, with good results.) Just mentioning that Galileo, Newton, and Einstein also used them can help. Going into more detail about specific scientific thought experiments can be valuable, too.

    Here are a few more suggestions:

    - start off early in the semester with wildly unrealistic thought experiments such as the ring of Gyges or the Matrix. In my experience these examples appear to be so far fetched students have no problem with them.

    - for ethics, point out that real life sometimes presents us with clear cut dilemmas. For example the use of "fetal reduction", abortion to save the life of the mother, or the fact that we have a greater need for organs than we have organ donors. So we can't always just "change the situation."

    - point out that we can answer and often use such unrealistic hypotheticals outside of a philosophy class, and they are informative: “what if you got an A in every class while in college? What would your GPA be, and would you have a better chance of getting into med school? Of course there are many variables, but...” You can also point out that one reason politicians don't want to answer such questions is that voters would be able to make generalizations about how they think about other similar situations.

    - (this may be stating the obvious) stay clear not only on the process, but the overall point, to help us understand (in Singer's case) our obligations to others, not what you or most people might do in that situation (that would be a psychological point rather than a philosophical one) or what it would be nice if we could do.

    - finally, if some in the class are getting it, I sometimes say to the student who doesn't “you aren't really getting the point of the question, are you? And then ask a student I suspect does to explain why.

  5. Good discussion! One of my peers has a funny anecdote for the situation you describe - that is, when students change the parameters of a thought experiment. He calls it, "MacGyvering the example." MacGyver, remember, often manufactured ways out of sticky situations where there appeared little to no resources to do so.


If you wish to use your name and don't have a blogger profile, please mark Name/URL in the list below. You can of course opt for Anonymous, but please keep in mind that multiple anonymous comments on a post are difficult to follow. Thanks!