I’ve recently read this paper by Bryan Caplan, which argues that many neurodivergences can be understood, from a microeconomic standpoint, as very unusual sets of preferences.
I think Caplan does seriously overstate his case. In particular, his argument for why delusions should be viewed as extreme preferences fails. There is no reason to believe that people with fixed false beliefs should not be responsive to incentives. I have a fixed true belief that two plus two is four, but if I were involuntarily imprisoned and drugged unless I said that two plus two is five, I would certainly do so. That doesn’t mean that I “really” believe that two plus two is five, or that my belief in basic addition is any less fixed. In addition, the fact that some people can reason their way out of delusions does not make their impairment any less real. It is possible to believe things on one level but not another. Most people discard the hypothesis “all my friends are secretly conspiring to hurt me” without even thinking about it; having to carefully gather the pros and cons of that belief and laboriously work out whether it’s true is, in fact, believing “all my friends are secretly conspiring to hurt me” far more than is warranted.
Similarly, his “gun to the head test” for telling apart impairments and preferences does not make very much sense. If you put a gun to the head of someone with a migraine and said “dance or I’ll kill you,” most people with migraines would manage to dance. But it would be very strange to think that migraines are actually the desire to lie under a blanket with a bottle of painkillers and moan “owwwwwwww”. Being able to overcome one’s constraints if one’s life is at stake is not the same thing as not having constraints.
However, I think it’s interesting to think about how many neurodivergences are actually unusual preferences.
Some neurodivergences are just odd preference sets. Transness, for instance, is the unusual preference to be a different gender than the one you were assigned at birth. Bodily identity integrity disorder is the unusual preference to have one of your limbs removed. Paraphilias (which are still in the DSM!) are unusual sexual preferences.
Some neurodivergences seem to be pure impairment. Caplan’s protests aside, delusions seem to be clearly in this territory. Similarly, depression is perhaps the most crippling impairment that exists; other impairments are bad when they make it harder to be happy, but depression is literally an impairment in one’s ability to feel happiness.
On the other hand, a lot of neurodivergences seem to be a combination of impairment and unusual preferences. Consider autism. An autistic person has certain impairments: for instance, they may lose the ability to speak under stress, or not be able to tell people’s emotions from their faces. An autistic person also has unusual preferences: they might only want to eat certain foods, or like flapping their hands when they’re happy, or collect lots of information about a special interest, or play by lining their cars up in a row.
Similarly, borderline personality disorder comes with certain impairments: for instance, I do not have an instinctive understanding that when people leave the house they continue to exist, and I’ve been known to have fixed false beliefs (although more “cognitive distortions” than “delusions”). BPD has less unusual preferences and more unusual strength of preferences: as best I can figure out, I’m some sort of mild utility monster who gets much, much happier from some stimuli and much, much sadder from others.
Probably most of you have figured out by this point that yesterday’s post was not entirely motivated by abstract concern about tiling the universe with things. Basically: many solutions to the tiling problem– average utilitarianism aside– make eliminating unusual preferences at least morally problematic. On the other hand, I think this lines up pretty well with my intuitions about eugenics: eliminating BPD and autism– much less paraphilias and transness– seems to be wrong in a way that eliminating depression or delusions does not. So that is useful.