Saturday, May 19, 2012

Using thought experiments to investigate your reactions and motivations


2.bp.blogspot.com
by Julia Galef

Philosophers often incur derisive sniggers at the idea that they can figure out the world “without leaving their armchair!” In some respects, I agree with that criticism. The idea of being able to reason a priori about the finitude of the universe or the existence of a god is pretty absurd.

But there is one tool in the philosophical toolkit that I absolutely believe allows you to learn about the world from the comfort of your favorite armchair: the thought experiment. In a regular experiment, you intervene in the world in some way, observe the results, and make an inference from those results about how the world works. Thought experiments use that same template, but the “data” you’re observing is the output of your own brain, and you use it to make inferences about the workings of a very specific part of the world: you.

This is one of the most useful tools I know for introspecting about my own motivations, values, emotional reactions, and preferences. I’ll start with a simple example, to illustrate the idea. Let’s say I’m considering applying to graduate school, but I’m reluctant, and I want to have more insight into the reasons for my reluctance. I might tell myself that it’s because I believe it wouldn’t be worth the time and difficulty, and that I didn’t expect to find a good enough job upon graduation to make that investment worth it. But I also suspect that I might be loath to apply because I’m afraid of getting rejected. How can I test which factor is really motivating me?

Well, what I would do in that case is hone in on one variable that I hypothesize might be important — e.g., my fear of rejection — try varying it, and observe how that changes my outcome variable (motivation to apply to grad school). There are plenty of ways I could vary “fear of rejection.” For example, I could imagine myself applying and knowing, somehow, that I would get in to all the schools to which I applied. That’s not very realistic, though. It’s hard to convincingly imagine something that would never happen in real life, and in real life, we can’t just “know” what’s going to happen ahead of time with certainty. It’s much better, when possible, to make your thought experiments believable so that your reactions to them constitute reliable data.

So instead, the thought experiment I would do in that case is to ask myself: “Imagine that your top choice grad school department invited you to come visit, and after interviewing with the professors there, they were so impressed with you that the head of the department told you flat-out: We want you here and we’re just going to skip the admission process with you. Let us know by tomorrow if you want to join our program.” Would I, in that scenario, still feel hesitant about going to grad school? If so, that result gives a lot of credence to the hypothesis that my hesitation stems from doubts about whether grad school is a worthwhile investment for me. If not, that result gives more credence to the hypothesis that my hesitation stems from a fear of rejection.

Here’s a related example of using thought experiments to figure out my own motivations. Once I was in graduate school, I had to figure out whether I really wanted to be a professor. I had a few hypotheses about my motivations, but to keep things simple, let’s focus on two. Hypothesis 1: I want to be a professor because I predict that I will really enjoy research and teaching. Hypothesis 2: I want to be a professor because it’s prestigious and I like the idea of having a prestigious job.

What kind of thought experiment could I do to test Hypothesis 2? As in the last experiment, there are many ways I could vary the “prestige” variable. I could say, “Imagine being a professor wasn’t prestigious...” But that’s the kind of change that’s hard to imagine. (In general, thought experiments of the sort “Imagine that our world was different in the following way(s)...” produce less reliable data than those that begin, “Imagine that tomorrow, X happens...” In other words, counterfactuals are harder to imagine than likely future changes.)

So the thought experiment I did instead was to ask myself: “Imagine that you’re offered a job that allows you to do all the research and/or teaching you want, but it’s not officially a professorship. It’s... I don’t know, let’s say, a ‘Lecturer’ position or something else that doesn’t sound very prestigious. In other words, you’re not allowed to call yourself a professor. Now how much do you want the job?” And if my answer is “Eh, not much,” then that suggests Hypothesis 2 was on the mark.

In addition to investigating your motivations, you can also use thought experiments to investigate your emotional reactions to people or situations. For example, I was once planning a potential vacation but feeling kind of blah about it when I envisioned myself on the trip. But was that “blah” reaction an accurate simulation of how much I would enjoy the trip once I was on it? Or might it be that I was just tired or in a funk for some other reason and therefore unable to imagine myself having fun? The  thought experiment I did in this case is somewhat different: I imagined myself doing something that I knew, from repeated past experience, I always enjoyed (a friend’s annual Halloween bash). The result: I couldn’t imagine myself enjoying that either. Which indicated to me that the original hypothesis (“I can’t simulate myself enjoying this trip because this trip isn’t the sort of thing I would enjoy”) was probably false, and that the true hypothesis was more likely to be: “I can’t simulate myself enjoying this trip because my simulation module is temporarily broken.”

I’ll close with one more example of investigating your emotional reactions with thought experiments. A while back my roommate offered me a cookie from the box she had just bought. I felt a (mild, fleeting) pulse of irritation at her offer, because I was on a diet and she knew it — wasn’t it insensitive of her to offer me something she knew I wasn’t allowed to eat? At least, that was the explanation my brain gave me for why I felt irritated. Then I did a thought experiment: what if she had not offered me the cookie? How would I have reacted then? To my surprise and amusement, when I simulated that scenario, I still felt mildly irritated: “Is she not offering me a cookie because she knows I’m on a diet? Huh, that seems kind of paternalistic of her. Can’t she let me make my own choices about what I eat??” The fact that I would have felt irritation no matter how my roommate had handled that situation indicates that the irritation had more to do with the situation itself (want cookies, can’t have ‘em) than with anything my roommate did.

Next time: Using thought experiments to explore your values.

7 comments:

  1. Love the article, but what I REALLY want to know now is: are you a professor for the prestige?!

    ReplyDelete
  2. Everyone practises self deception ~ so how do you know for sure that you are irritated at having a want denied? What if you hate your room mate, but you are hiding that fact from yourself? Perhaps your apparent feelings about the two cookie scenarios are a cover for a deeper resentment that you're harbouring.

    The results of thought experiments about the physical world can be put to the test later by designing a physical test & collecting data. Is it possible to test YOUR OWN reactions & motivations with acceptable certainty given that our brains are the most complex objects known in the universe? I'm leaning towards "No".

    ReplyDelete
    Replies
    1. Well, self-deception is usually at least in rather introspectively predictable directions (e.g., self-interest, ideological stance, ingroup/outgroup). Ceteris paribus (i.e., unless her roomate is a professional competitor or a republican or something like that), there is no reason to think that Julia would be *in denial* about resenting her roommate, and therefore no particular reason to worry about that source of bias.

      Delete
    2. The "hate" option is only one of many possibilities ~ the difficulty is identifying ALL of them rather than just the two that Julia is conscious of. Then one has to determine what personal biases might be operating below ones awareness for each option. Then when one has conducted the thought experiments the question arises ~ "how does one know that ones analysis is correctly unbiased?"

      The biases can stretch back to childhood [Rosebud]. The biases could be cultural.

      The trouble with thought experiments of this kind is there usually isn't a way of confirming ones conclusions independently ~ it's all opinion & just so stories designed to accommodate a coherent narrative. Read any autobiography around events with which you have some direct knowledge & outside of the dates & places it's striking how much of what's left is purely post hoc rationalisation.

      A prime example of this effect is to be found in police witness reports. We all walk through our lives almost completely unaware of the world we are in & we weave the story of our lives from the few bare threads that we pick up & retain.

      Self-deception is a central process in our lives ~ we don't usually notice the huge gaps, mis-rememberings & misinterpretations of our own [& other peoples] motivations in our life stories.

      Delete
  3. After finding out (with probability)what your motivations or emotional reactions are - could you use thought experiments to change them, supposing you don't like what you've found out about yourself?

    ReplyDelete
  4. Believe in an objective reality, do we? Then paragraphs 1 and 2 do not disagree with each other and flow. Think there are problems with an objective reality? Then the thought experiments are of value when wrestling with any question you like, from the best working definition of a god, the best way to get from here to there, or the best approximation of what we used to call facts.

    Sorry, Julia, but claiming the idea of a god using thought experiments alone as absurd strikes me as short-sighted and backed up by little if none of the achievements in science and information technology this past century.

    The argument by physicist/philosophers who have made the study of reality their daily business is that a case can be made for what goes on in the imagination to be fundamentally no different than what goes on outside the imagination. Why? <

    - The difference between the two is the level of agreement about the data. You and I might 'agree' the sidewalk and the rain, but we usually do not agree about the things we conjure.

    - A growing consensus among physicists that objects separated by 'physical' time and space would then be separated by human or imaginary constructs, given the imaginary nature of time and space, given time and space as devices constructed by humans as to better deal with their environment.

    - Every 'physical' object built by humans can be said to have originated in one or more human minds. It sort of flows that 'natural' objects, or those not known to originate in human minds came from some other sort of mind.

    The rebuttals to these arguments are often of a common-sense " Dr. Johnson refuting it thus" nature. But the simulation argument says that the foot kicking a rock at the forward-thinking bishop-philosopher is as imaginary as the pain caused. And if you do not buy simulation, there is simply the informational aspect of all things, and if you do not buy information as the basis of all things, then replace it with energy - it doesn't matter. What matters is that real, solid objects that make up our world are at heart, at a minimum, nothing but fields.

    The foot is making a statement - of sorts. It is saying, "Stop imagining by yourself, partake in my imagination too, sent to you via the kicked rock, and build the future via committee, not fiat".

    Can we not agree that reality is simply the things we agree, nothing more, nothing less?

    Like Facebook, what most call reality seems more likely a social network.

    ReplyDelete
  5. http://en.wikipedia.org/wiki/Introspection

    ReplyDelete

Note: Only a member of this blog may post a comment.