Everything should be made simple as possible, but not simpler.
A brief note on the deceptive complexity of human societies and the consequences for productive dialog.
The title of this story, which I’ve also alluded to in my profile description, is a rather famous quote — you can search if you don’t recognize it and wish to know who it’s ascribed to. The reason I think the quote is particularly insightful is that, when we think deeply about what it entails, it calls us to cut through the distractions and clutter to get at the heart of the matter, but also to be cognizant of throwing out the heart of the matter along with the bathwater, so to speak. That is, it asks us to answer the question: how simple can we actually make this?
The title quote is especially potent when paired with a H. L. Mencken quote: “For every complex problem there is an answer that is clear, simple, and wrong.”
If we’re going to seek simple solutions, it would be wise to understand the complexity of the problem.
In the case of human society, a proof of the potential difficulty can be found in the work of the brilliant mathematician John von Neumann and the economist Oskar Morgenstern, who in 1944 published the book, “Theory of Games and Economic Behavior”. A large part of the fame of another prominent mathematician named John, John Nash, came from advances in the mathematical analysis of game theory.
Let’s consider a simple, contrived scenario.
Suppose there are two children, John and Julia, who don’t really want to practice the piano. So their parents come up with a reward: if you practice the piano, you can have three mini-cupcakes! Unfortunately for the parents, Julia and John only have time to practice after they get home from school but before the parents come home from work. And of course, the kids each insist that they themselves practiced.
So the parents take the kids aside, one at a time and say: “I know you don’t want to get your sibling in trouble, but we need to know if they practiced. Tattling is hard, but if they really didn’t practice and you’re honest, we’ll give you an extra cupcake.”
Aside from this being a rotten parenting technique, what should the kids do, given that of course neither of them practiced one bit? (Note: the analysis that follows actually does not depend on whether they practiced.)
John reasons as follows: I really love cupcakes! If I tattle on Julia, I will get four cupcakes instead of three if she doesn’t tattle. And I will get one cupcake instead of zero if she does tattle. Either way, it’s better for me to tattle, so I will tattle.
Julia, an accomplished mathematician in her own right, reasons likewise, then continues: but if I tattle on John, he’ll be mad and will tattle on me next time, and we have to practice the piano every day, not just today. So, I should only tattle on him on the last day — then he can’t retaliate. So maybe he will be mad, but I should certainly tattle on him on the last day to get a guaranteed extra cupcake.
John realizes this too, then continues: but if she’s going to tattle on me on the last day, I have nothing to lose if I tattle on her on the day before — she already would, so she can’t change her behavior to retaliate. So I should tattle on her the day before the last day.
Working day by day, the children realize that the rational course of action is, regardless of the number of days they’ll need to practice, to tattle this very day.
However, Julia and John, being very sensible children, each keep quiet, and each get three cupcakes, which is more than they would have gotten if they were perfectly rational and each tattled.
Wait, what?!
This silly little problem — which you may have by now recognized under its better-known name, the “Prisoner’s dilemma” (in the multi-day case, iterated prisoner’s dilemma) — appears to have broken rationality itself. Indeed, the technical term for the kids’ sensibility is “superrationality”.
So, let’s review. We have a couple of kids who don’t want to practice piano with an ongoing reward at stake, and we broke (conventional) rationality. There’s no implicit bias, no influence peddling, no dogwhistles, no societal stigma, no patriarchy, no income inequality, no PTSD, nobody was canceled, we weren’t owning the libs, there are no garbage people, and we already have a scenario so complicated that there is no known perfect strategy and people write long papers comparing tactics under particular scenarios.
This tiniest idealized microcosm of human interaction already seems beset by nigh-insoluble complexity. Fundamentally, this is because humans react to each other: the key step that unleashed the complexity was Julia’s realization, “but if I tattle on John, he’ll be mad and will tattle on me next time.”
So, next time you feel compelled to try to solve a social problem in 140 characters — or in 280 — consider that maybe, just maybe, you might not have found such robust simplifying principles that you have escaped Mencken’s curse.
And if you happen to find me commenting on one of your Medium articles with irritating points about sample sizes and data not showing this and psychological safety that and absence of evidence isn’t evidence of absence sans a good statistical power test and good methodology, and think that maybe my profile line is misguided, please think again.
Maybe it really is that complicated. Maybe it’s worse. But maybe, if we work together, and think clearly, we can figure out enough to come up with some pretty good ideas.
Addendum: for a lovely explanation of the history and findings of Prisoner’s Dilemma research specifically, and thoughts on its specific implications for human and animal behavior, check out this video by Veritasium: https://www.youtube.com/watch?v=mScpHTIi-kM