Oxford University Press's
Academic Insights for the Thinking World

Thought experiments in philosophy

Philosophers love thought experiments. Many of us deploy them as our version of the scientific method. They isolate some feature of our experience and evoke intuitions about it, and these revealed verdicts enable us to adjust relevant theories in light of what we find.

The method goes back to Plato, who, for example, has a character in The Republic challenge Socrates’s view of morality by appealing to the possibility of a ring that could make its bearer invisible, in order to demonstrate that people are most fundamentally self-interested.

Here are some of the other famous philosophical thought experiments:

  • Suppose there were a planet that was a duplicate of Earth in every way except that the chemical compound of the colorless, odorless drinkable stuff in its lakes and rivers was not H20 but something else, XYZ. Would XYZ be water? (Hilary Putnam)
  • Suppose you knew absolutely nothing about your social status, race, ethnicity, gender, basic life plan, or prospects. Under these conditions, what political conception would you choose to live under? (John Rawls)
  • Suppose you are alone in a room, where sheets of paper with strings of Chinese characters written on them are slipped in, and you are supposed to write different strings of Chinese characters on a different piece of paper according to an extremely detailed set of instructions, which you then slip out the door. It turns out you are fooling those outside the door into believing there is a Chinese speaker in the room answering their questions. Does this room — allegedly akin to a computer — understand Chinese? (John Searle)
  • Suppose we could transplant one half of your brain into each of the skulls of your identical siblings (whose own brains were destroyed). Each resulting person would wake up thinking he or she is you, with all psychological connections to you preserved. Which one would be you? (Derek Parfit)
  • Suppose a goddess designs a zygote, combining its atoms in such a way that it grows into a person (Ernie) who eventually judges it best to perform precisely the action the goddess designed him to perform at that moment in time. Does he perform this action freely? (Al Mele)
  • Suppose a team of brilliant neuroscientists has the ability to manipulate Plum into deciding to kill White by producing in him a strongly egoistic neural state just before he reasons about whether to do so, a state that determines he will decide to do so. Is he morally responsible for killing White? (Derk Pereboom)

There are many worries one might have about this general method. The cases aren’t sufficiently detailed, they yield conflicting intuitions, they generate framing effects, the intuitions garnered for use in the theory are culturally bound, and so on. But the worry I want to focus on instead is that sometimes we appeal to these science fiction cases too quickly when there are plenty of real life cases all around us that are potentially more fruitful.

“sometimes we appeal to these science fiction cases too quickly when there are plenty of real life cases all around us that are potentially more fruitful”.

Consider the last two thought experiments. They are both intended to get us to think ultimately about whether freedom and responsibility are really possible in a fully determined world, as each of the external manipulators dramatically brings to life different features of a deterministic world.

They are also, more importantly, meant to isolate the relevant features in a way that will yield clear intuitions on our part. Neither Ernie nor Plum acts freely or is morally responsible, obviously. And the precisely-drawn nature of the cases, and our clear-cut responses to them, are what is supposed to enable us to see more clearly how to make judgments of agency and responsibility about actual people in the messier real world.

But the very messiness of the real world should perhaps give us some pause about this method. Responsibility, for instance, has attached to it a whole range of emotions and practices; anger, gratitude, admiration, disdain, praise, punishment, reward, criticism, withdrawal of trust or warm feelings, and so on. These are deeply ingrained human emotions and practices, typically with a long evolutionary or cultural history.

Consequently, it is hard to see how we might theorize about responsibility in the first place without taking seriously these deeply ingrained responses. But the fact that our responses are so messy when it comes to many real-life people might just mean that responsibility is messy, that perhaps it can’t be made as precise as the thought-experimental method would have us believe.

Numerous real-life cases illustrate the point. None of us are ideal agents, but we may still achieve insight into the nature of agency by investigating the many different kinds of nonideal agents among us.

For example, what are we to think of the responsibility of people with clinical depression, Alzheimer’s disease, OCD, kleptomania, psychopathy, intellectual disabilities, Tourette syndrome, or those from horrifying or morally deprived childhoods? Many of us either know people with these backgrounds or conditions, or we are such people.

These agents tend to elicit in us a profound unease. We may feel, for instance, that only some of our responsibility emotions and practices are appropriate for them. Or we may feel that some of these agents deserve very different responses than others. Or we may feel that “there but for the grace of God” considerations suggest we extend our unease about their responsibility to those without such disorders and backgrounds.

In any event, thinking hard about these cases may reveal that our commitments and dispositions with respect to real life agents are surprisingly complex and informative, so much so that our intuitions about thought experiments that abstract away from real life details are misleading or irrelevant.

And a bonus of focusing on real-life cases is that we can get help on their crucial empirical details from various sciences employing the actual scientific method (e.g., psychology, biology, neurophysiology), which enriches our theorizing in interdisciplinary ways. And why shouldn’t we approach the matter in this way if our aim is to understand human agency and responsibility?

Philosophical thought experiments are lots of fun to design and think about. But they may obscure reality or prevent progress when too removed from our actual experience. Perhaps, then, we should be appealing far more often to the wide variety of fascinating and complex real life cases near to hand.

Featured image credit: The Death of Socrates, by Jacques-Louis David (1787). Public domain via Wikiart.

Recent Comments

  1. Patrick Lin

    Nice post, David. I’m curious if you’d say the same thing about science experiments. For instance:

    We’ve given spiders all kinds of drugs, and we’ve sent them into space. (These are two different experiments; spiders aren’t getting high in space, as far as I know, though that’d be awesome). And we do this to see the effects on the spiders’ web-spinning ability.

    The practical point or application isn’t quite clear, but that’s ok. The experiments still seem valuable in that they advance knowledge and basic science, as distinct from applied science and technology.

    These spider experiments don’t replicate actual conditions in the real world, of course. Spiders don’t live in space (or at least I hope not), and they’re not smoking weed in their natural environments (only because they can’t operate a lighter or rub two sticks together). But, like thought experiments, they isolate certain variables (e.g., gravity) in a controlled way — one or two at a time — because experiments must be less messy than the real world, in order to draw clear conclusions from the experiments.

    And these conclusions may need to be synthesized and studied further, examined in particular social contexts,etc. to get to deeper, more practical conclusions — ones that translate into real-world applications. But that’s ok, too. A single experiment rarely is meant to solve a big hypothesis or issue.

    Insofar as philosophy is similar to the scientific method, if what I described is accurate but not a liability for science, wouldn’t that also be true for thought experiments? Or, in the other direction: if this is good methodology for science, why wouldn’t it be the same for philosophy?

    Any thoughts are welcomed, at your convenience.

  2. David Shoemaker

    Thanks a lot for your comment, Patrick. You ask a good question. I intended to be drawing attention primarily to thought experiments in domains where our responses (emotional, e.g.) are often a crucial data point for metaphysical theorizing, and whose conclusions are meant to be revisionary with respect to a network of practical concerns and social practices. This is the domain, in particular, of moral responsibility (I occasionally lamely joke that you can’t spell ‘responsibility’ without ‘respons’, which is just to say that responsibility is so closely tied to our responses that you can’t theorize about the former without essential reference to the latter.) Given our wide-ranging responsibility treatments of one another (anger, blame, sanctions, rewards, gratitude, etc.), and given how closely connected responsibility is to our responses, it’s unclear why theorists want to lean so heavily on cases outside of our experience to inform our actual experience with one another *in these matters*.

    This is not necessarily the case in a variety of other philosophical domains in which thought experiments are regularly deployed. And I also don’t deny that there may be value to doing so even in the responsibility domain. My primary point was to urge us to look more closely to hand to real life cases than we do when we our theorizing is taken to have normative implications for our networks of (deeply human) emotional responses and practices.

    Note that we don’t have any such analogous practices organized around what spiders do when high, or how they spin webs in space (at least that I know of).

  3. […] David Shoemaker在OUPBlog中發表了一篇短文:Thought Experiments in Philosophy,批評了哲學中思想試驗的運用,認為日常生活中的事例更為豐富,他以responsibility為例做說明,值得思考! […]

  4. […] “Here are some of the other famous philosophical thought experiments” written by David Shoemaker at OUP […]

Comments are closed.