Decoupling as a Moral Decision

You may have heard of the dichotomy drawn between ‘high decouplers’ and ‘low decouplers’. High decouplers are people who are able to isolate an idea from its context: as Tom Chivers writes, they can perform a ritual of ‘If we accept X, then we might think Y’, and then explore the implications of X being true without knowing whether it is true. For instance, a high-decoupler might be happy with a thought experiment that starts ‘If we accept that IQ is heritable, then…’, whereas a low-decoupler wouldn’t be happy just assuming that IQ is heritable - they don’t regard this ritual as sufficient in moving on to the next part of the conversation, they might reply to the argumentation that followed with ‘But I’m not convinced that IQ is heritable…’, and feel the need to continue with that part of the conversation. Then a back-and-forth may follow where the high-decoupler says ‘But let’s just assume for the sake of argument that it is true’, and the low-decoupler replies ‘But I’m not sure we can assume that’, and you can probably imagine how the conversation proceeds from there.

But the way that the decoupling dichotomy is often presented is that you are either a ‘high decoupler’ or a ‘low decoupler’, and it is a fairly immutable characteristic. Either, the ritual ‘If we assume X, then Y’ works for you, or it doesn’t. But I’m not really convinced this is true, at least not when being used during the political discussions in which it’s often brought up. Much of the original research on cognitive decoupling comes from Keith Stanovich - the rough idea is that the ability to understand counterfactuals and imagine a world that is basically the same as the real world but differs on some small number of dimensions is useful and important for making causal inferences. This ability is correlated with IQ - Stanovich writes that:

The raw ability to sustain such mental simulations while keeping the relevant representations decoupled is likely the key aspect of the brain’s computational power that is being assessed by measures of fluid intelligence.

Consider the following question, taken from Levesque and given as an example of a decoupling question by Stanovich: Jack is looking at Anne, and Anne is looking at George. Jack is married, and George is unmarried. Is a married person looking at an unmarried person?

A. Yes.

B. No.

C. Unable to be determined.

While most people answer C, the correct answer is A. If Anne is married, then Anne is looking at George, which fulfils the criterion. If Anne is unmarried, then Jack is looking at Anne, which also fulfils the criterion, so the answer must be A. This is an example of cognitive decoupling as understood by Stanovich (and as it is often used in the psychological literature), and requires using thought and imagining hypothetical worlds to figure out the answer. But this seems fairly different from how decoupling is being used by Chivers, and by other people who talk about ‘high decouplers’ and ‘low decouplers’. Decoupling in the academic literature is like a cognitive ability that is correlated with IQ, whereas decoupling as understood by Chivers and the blog Everything Studies (which popularised the concept in the aftermath of a debate about the relationship between race and IQ between Sam Harris and Ezra Klein) is more like a willingness to consider controversial hypotheticals as a way of thinking about the world.

The former (let’s call it A-decoupling, for Academic Decoupling) is likely to be a fairly immutable cognitive skill, broadly part of general intelligence, but the latter (let’s call it D-Decoupling, for ‘Discourse Decoupling’) seems almost like a moral (and possibly political?) decision about whether decoupling is appropriate in a particular circumstance, rather than some personality trait that some people have and others don’t. For instance, Chivers discusses in his article the fact that the question of whether it would be possible to genetically alter the population through sterilising some people is completely different to the question of whether it would be morally right to sterilise people in order to genetically alter the population through sterilisation (i.e. eugenics).

But whether I would be willing to D-Decouple here is totally dependent on the context in which the conversation was taking place. Suppose that someone begins a discussion with ‘If we forget about the moral concerns surrounding eugenics for a moment…’, would I be willing to D-Decouple and indulge in his ‘If X then Y’? Well, if it’s a friend and we’re having a casual conversation, I would be willing to see where he was going. If I’m in a chat with Richard Dawkins on the Radio discussing his claim that ‘Eugenics would work’, I would definitely consider insisting that we talk about the moral implications. If I’m having a debate with a well-known eugenicist who is gaining public support for his suggestion that we sterilise people with mental health problems, I would certainly reject the D-Decouple, because I think there would be a moral obligation not to let him move on to the ‘would it work to eliminate disabilities?’ without discussing the horrors of eugenics in the past.

Let’s briefly touch on the Harris-Klein debate: in the debate, Harris is happy to talk about potential racial differences in IQ and D-Decouple that from the history of people advocating for racist policies on the basis that ethnic minorities are intellectually inferior. Klein, on the other hand, rejects the D-Decouple, insisting on talking about the history of racism linked with the idea that ethnic minorities have lower IQs. I’m unconvinced that the Harris-Klein example is one of a ‘high-decoupler’ coming up against a ‘low-decoupler’, simply a cognitive mismatch that, if pointed out, would lead to them saying ‘Ah, I see we just have a different cognitive tendency towards decoupling, how foolish we were to get so heated over this!’. In fact, Klein is almost certainly capable of accepting the D-Decouple (and probably has a high ability to A-Decouple), but he believes there is a moral duty not to let the conversation go that way, in the same way that I believe there is a moral duty not to let a eugenicist D-Decouple the moral argument against eugenics from the question of whether eugenics is possible in a public debate.

So, the ‘If X then Y’ ritual that Chivers identifies is not, in my view, simply a ritual that some people are naturally inclined to accept and others are inclined to decline (although there is probably some variance in how much people enjoy and are able to partake in hypotheticals), but often (especially during public conversations) an offer from one party to another that will be completely contingent on context. And there are examples of D-Decoupling offers that I think almost everyone who identifies as a high-decoupler is likely to reject (although I’m willing to accept there might be some people who would accept): if you got in a taxi with a taxi-driver you had never met, and he posed the question ‘If you agree with me that [insert ethnic group here] are genetically inferior and unable to learn how to drive safely, would you accept my proposal that they ought to be banned from driving taxis?’, and posed this question completely seriously, I think most people (if they were willing to engage with the taxi-driver at all) would reject the D-Decouple rather than accepting just because the ritual was performed. But maybe I’m just a low-decoupler and there are people out there who would be willing to go ahead with the hypothetical proposed!

Leave a comment or message me on Twitter with your view!