On the Value of Thought Experiments
I am fond of thought experiments, though many people are not—or so I infer from the fact that every time I propose one the most common response I get is a refusal of its terms. So a number of people who have responded to my recent little exercise have said something like “But that’s not the situation we’re in”—Yes it is, in this thought experiment that I am totally making up—or “I would not vote for either party”—but in this thought experiment you have to choose one.
There’s some of this even in the response from my friend Noah Millman, as when he wonders whether there really are threats to religious liberty. In my thought experiment there damn well are, because I say there are! Against Noah, I say that the premises of my thought experiment are not and indeed cannot be “debatable premises,” because they are the ones I posit simply for the sake of the experiment: thus my insistence at the outset on the term “hypothetical.”
I can’t help being reminded of one of my favorite scenes from Wodehouse, in which the pathologically diffident Gussie Fink-Nottle discusses with Bertie Wooster whether he should follow Jeeves’s advice to build his self-assurance by wearing a Mephistopheles outfit to a costume party:
‘And you can’t get away from it that, fundamentally, Jeeves’s idea is sound. In a striking costume like Mephistopheles, I might quite easily pull off something pretty impressive. Colour does make a difference. Look at newts. During the courting season the male newt is brilliantly coloured. It helps him a lot.’
‘But you aren’t a male newt.’
‘I wish I were. Do you know how a male newt proposes, Bertie? He just stands in front of the female newt vibrating his tail and bending his body in a semi-circle. I could do that on my head. No, you wouldn’t find me grousing if I were a male newt.’
‘But if you were a male newt, Madeline Bassett wouldn’t look at you. Not with the eye of love, I mean.’
‘She would, if she were a female newt.’
‘But she isn’t a female newt.’
‘No, but suppose she was.’
‘Well, if she was, you wouldn’t be in love with her.’
‘Yes, I would, if I were a male newt.’
A slight throbbing about the temples told me that this discussion had reached saturation point.
I continue to believe that a thought experiment like the one I suggested is valuable in the same way that A/B testing is valuable. When someone asks you which of two shades of blue you prefer, you can, I suppose, say “Why just two? Why not fifty shades of blue?” or “Why not green, and red, and burnt umber, and all the other colors?” But maybe we would all learn something, even if something small, if you just picked one of the damned shades of blue. And then we can move on to other experiments after that, and gradually, incrementally, build up a more reliable understanding of our own values and preferences.
To those who would say that A/B testing, and thought experiments, are simple in comparison to real-life decisions, I reply: Precisely. That’s just the point of them. Politics is hard because it’s so outrageously complicated. It’s easy to get lost in all the overlapping questions and competing priorities. If you agree with a political party about seven of its official platform positions, but disagree about only one, but the one is something you care passionately about while the seven are, for you, relatively insignificant—how are you supposed to weigh those things? It’s impossible to say off the cuff. More thinking is required. It helps to break the situation down into its component parts. That’s what a thought experiment like the one I proposed is for.
More about the substance of the matter later; right now, I have teaching to do.