Alexander’s Thesis
Scott Alexander posted an essay recently on the “debate” or “argument” between people who are Rational versus those who aren’t. He frames it around a back and forth between Steven Pinker and Howard Gardner. Alexander starts by debunking a comment Pinker posted on twitter “slamming” Gardner. It could be paraphrased as “I am in favor of rationality. Howard Gardner, who is arguing against me, uses reason and logic. Therefore, he is just proving my point that rationality is right!” This is absurdist. If I declared myself in favor of truth, then am I the only one who gets to use truth, and can I claim victory whenever someone argues with me says something true? If I declared myself in favor of freedom, then am I the only one who gets to act freely, and can I claim victory whenever someone uses their free will?
Alexander gets that this is a very not smart position, and so he starts trying to reason through what might really be going on here. Maybe “Team Reason” is about doing careful, rational calculations, while the other team is actually “Team Heuristics,” who want to use rules of thumb more often? Or maybe it’s “Team Reason” against “Team Intuition”? Or maybe Pinker really thinks that “Team Reason” means being in favor of whatever is true and works, in which case… what’s the debate? He could just declare himself as being on “Team Right” and say he’s in favor of whatever turns out to be right on every issue. All of these are kind of dumb debates to have, in that no one (on either side!) is always doing long, explicit calculations or always using intuition or rules of thumb. Everyone uses a mixture, and the only interesting point of contention is when it’s helpful to do one during the other.
The real debate, Alexander concludes, must therefore be about people who want to study ways that people can think more effectively, versus those who don’t. As he puts it “rationality is the study of study.” If you’re on “Team Reason,” then you want to have particular rules and processes laid out that you can tinker with and improve over time. You want a theory that can be stated explicitly. It’s not that you don’t know that intuition can’t be useful–and even better than current theories on some problems! It’s that you think it’s more important to have an explicit rule laid out that anyone can follow than to just get a bunch of people trained to have great intuition.
Don’t Stop Thinking There!
The problem is that this, too, is not a normal or reasonable debate. Howard Gardner is a psychologist who has devoted his life to studying how people think. Is he not interested in learning how to think more effectively?
Let’s use Alexander’s own example here. He suggests that doctors who believed in blood letting and the four humours were better–more in line with “Team Reason”--than wise women who passed on traditional healing advice that was actually more effective than blood letting. (Blood letting is actually bad for people. Special herbal tea can actually be helpful, and at least won’t hasten your death.) That’s because these doctors had some kind of explicit theory that they used to guide their thinking, while wise women (presumably) didn’t.
The most obvious point here is that this is historically incorrect. There are lots of traditional healing practices that have general principles that guide treatment, and people could talk about them explicitly with at least as much depth as you could talk about the four humours. Five minutes googling easily turns up Medieval European practices of keeping herbal gardens, large books with illustrations and recommendations for the use of these plants like the Pseudo-Apuleius, and systematic medical writings on the topic based on observation like those by Hildegard of Bingen. So it wasn’t the case that there was a historical distinction between some people who believed in developing a formal and systematic understanding of medicine, versus another group of herbalists who didn’t develop any theories or explanations for how to treat people. There certainly may have been some herbalists who just carried out specific remedies because that’s what they learned from the local expert (i.e., they weren’t reasoning it out), but there are surely modern medical professionals who do the same thing. Not every doctor is operating from first principles when they provide recommendations; they’re typically following clinical guidelines based on the research of others whose recommendations they trust.
It’s my contention that anywhere in human history where there’s a regular practice of medicine, there are some people who are the experts testing ideas and systematizing knowledge, and some group of people a level lower who will talk with them and try to implement their ideas as best they can–even if they can’t derive that advice themselves from first principles. One interesting example of this is in Tara Westover’s autobiography, Educated. Westover’s mother was an herbalist, and she used systems to test her ideas about how to make proper remedies that would not make sense to almost anyone else. Muscle pains she got and special ways of crossing her fingers were interpreted as divine guidance when she was making recipes. That’s not how I’d evaluate a remedy, but even this woman working alone on her own system of herbal remedies came up with something as logical and consistent as the early doctors talking about the four humours. There’s some system there, it’s just not one we believe meets our current standard for evidence.
The final position that Alexander comes to is just as absurd as the ones that he dismisses above. Being on “Team Reason” doesn’t mean you’re going up against people who don’t want to think systematically. Everyone tries to draw general principles about their experiences and refine their thinking through a combination of observation and reasoning about these principles. Certainly Gardner, a scholar who has written hundreds of scientific articles and over thirty books, isn’t against systematic thinking. What’s in all that writing if not theories that attempt to apply some principles and systems to how to think about specific topics? They don’t have to be convincing to Alexander, but that doesn’t mean they are “anti-reason.”
An Outside Perspective
What Alexander is missing in his attempt to define the conflict is how this looks to someone outside “Team Rationality.” What Pinker and Alexander appear to be doing is trying to define in-groups and out-groups using tools instead of ideas. They are suggesting that what defines the debate is not a specific set of ideas, but a way of evaluating ideas. That will never work, because ways of thinking are tools that are accessible to everyone. What defines groups are ideas and values. As Alexander walks through his own logic, he keeps finding that both people he considers on his (and Pinker’s) team and those against his team are using the same logical tools. This will always be the case, because all of the tools he’s describing are broad level. (If he were to get into the very specific, for example Bayesian versus Frequentist statistics, he might find a real debate over tools–but that would lose all the big culture war energy the piece brings.)
To get more insight, let’s look at the specific points Gardner makes that Alexander objects to. Here are some quotes where Alexander summarizes and comments on Gardner’s arguments:
“The criticism was facile, a bunch of stuff like ‘rationality is important, but relationships are also important, so there.’”
“He has an inane paragraph about how respect is more important than rationality”
“One of the most common arguments against rationality is ‘something something white males’ ”
“Gardner and his supporters believe you should often retreat to heuristics like “don’t do anything that violates human rights” or “live a holy and god-fearing life” or “don’t do drugs” or “try to favor black women”.”
Not to put too fine a point on it, but this is some bitchy bullshit that does not use any of the careful rational thinking that’s supposed to be at the heart of the debate. This makes everything Gardner says sound stupid, even when the points are obviously true–and not even necessarily “in conflict” with rationality. The point that “rationality is important, but relationships are also important” is obviously true. Adding the “so there” is juvenile, and just serves to dismiss a rational argument.
The argument goes like this: “when you are out in the world, there are times when you might want friends more than you want to be right. So sometimes you might choose to believe in something that is not purely rational, to make a friend.” In Gardner’s essay, he says that someone moving to a new area might join a local church to feel like part of a community, even if some of what the church teaches is irrational. This is what Alexander considers “facile criticism”? I don’t even see it as an indictment of rationality, it’s just saying that there are multiple values people can hold, and sometimes one of them might override rationality. It’s not that rationality is wrong, it’s just that other things also matter. And this is a criticism that makes Alexander angry, something he feels the need to refute?
Gardner’s “inane paragraph” about respect says that… drumroll… we should all respect each other, even if we disagree. By implication, we should treat people who aren’t “as rational” as us with respect, even if they disagree with our arguments. It seems like pretty boilerplate “don’t stoke the online fires of hate” stuff, the type of stuff that you shouldn’t need to say to intelligent, rational people. And yet here we are, watching a leading light of “Team Reason” get all pissy about having to be respectful to someone he disagrees with. He certainly does not treat that person’s ideas with respect. So maybe Gardner did need to make the point that “reason isn’t everything” and you shouldn’t be disrespectful of someone just because they don’t agree with your line of arguing.
The line that an argument against rationality is “something something white men” is just red meat for fans of the culture wars. It literally has no content except to signal that any argument that identifies white men specifically should be completely dismissed. We don’t even need to know what it is. Maybe Alexander does see a lot of poorly reasoned arguments made about how rationality is bad because of the way white men (in general) use it to dismiss other people’s arguments. It might be legitimately frustrating. But it is definitely not rational to present an argument as “something something white men.” If you wanted to have a Rational Conversation about whether there’s any relationship between being a big fan of “Team Reason” and being a white man, or whether white men who are fans of “Team Reason” actually tend to behave badly (and often not rationally!), that may be worth having. But you’d want to actually include some reasons on both sides. (Also, consider reading Kate Manne’s Entitled first.)
Finally, Alexander says Gardner-type, non-rational people “retreat” to heuristics like “don’t do anything that violates human rights” alongside “try to favor black women.” What is this supposed to do? Should we think that both of these are silly, bad rules of thumb to use? Because not violating human rights seems like a good rule of thumb to have and use whenever you’re in a new situation. In fact, it seems like a moral rule that should never be violated, particularly because we have come up with human rights through a process of rational deliberation.
The rule “try to favor black women” seems like a culture war point to me. Gardner doesn’t mention black women (or black vs. white differences) at all. It sounds like it’s being set up to ask people to give a special advantage to black women in decision-making processes. Another formulation, which is more in line with how non-libertarian, non-right wingers understand the argument, is that people in the U.S. tend to have an intuitive bias against black women, based on the representations and stereotypes they’ve absorbed from society. They should try and counteract this bias when making decisions. One simple way to do this is to make their normal evaluations, and then remember at the end “I usually don’t give black women enough credit, so I’ll bump up my evaluation of them slightly.” When you describe it that way it sounds positively rational.
Answering Alexander’s Question
Now that we’ve seen how Alexander describes his opponents, it should be more clear that this piece is really about defining groups the same way that traditional culture warriors do. By defining himself as part of the “Grey Tribe,” Alexander is able to brand himself a bit differently from a Fox News commentator. But basically what he’s doing is creating an ingroup that rejects liberal or left-wing thought, and brands that decision as principled and “rational.” He can get himself part of the way to understanding the problems with his position–that’s what the first half of his essay is. But he can’t get all the way there, because he has a bias for his ingroup.
Backing out further, I’d summarize my argument with a few core points. This is my answer to Alexander’s question: “What Are We Arguing About When We Argue About Rationality?”
Everyone uses a mixture of careful, systemic thinking and rapid, intuitive thinking. It’s therefore not useful (and a bit pretentious) to say that you and your people are using one and your opponents aren’t. (All people use System 1 and System 2, from Khaneman’s Thinking, Fast and Slow.)
Any time a model is used for careful, systemic thinking, it must be a simplification of the real world that highlights certain core relationships deemed important by the person developing the model. This stems from the principle that the only map that perfectly captures all details of the terrain is one that is exactly the same size as the terrain. This map would fit perfectly on top of the world, and not give any new information beyond you just looking at what’s in front of you (because that’s all of the map you can see!). Put another way, the world cannot be compressed into the form of a model that can be reasoned about without making strategic simplifications. The old George Box quote “all models are wrong, but some are useful” applies here.
Rational thought always emphasizes some aspects of reality, while ignoring and making assumptions about others. Any given rational analysis is therefore always susceptible for criticism about being incorrect. All models are wrong. There are some interesting arguments to be had about how to perform calculations in a model (or family of models) that make a given set of assumptions. But what is almost always more interesting, contentious, and important, is what assumptions, simplifications, and elements to focus on in your model.
Arguments about ways of thinking are therefore almost never about the mechanics of reasoning, but about what inputs to consider, and what decision-making architecture to use, when calculating an output (e.g., making a decision). For example, one person might say that decision-making should take into account only whether a decision will lead to a better outcome, while another person would say we need to also include a weight or punishment so that getting a good outcome is weighed against being disrespectful or damaging a relationship.
If arguments are really about these assumptions and inputs, as opposed to the particular tools we use, then the difference between the Alexander / Pinker tribe and the Gardner tribe is not really about rationality. It’s about what you are willing to treat as important and valid inputs, and what assumptions you are willing to make. In other words, it’s about values.
So what we end up with through this formal argument is that these rationality arguments are just about arguing for one particular set of values: those values held by people who self-identify as members of “Team Reason.” This is, of course, something we could start to get a feel of when we were looking at the kinds of criticisms Alexander was leveling at his imagined opponents. He was angry, dismissive, petty, and decidedly non-rational. We can actually see how that contrasts with Howard Gardner, who says we should include Religion, Relationships, and Respect as important values alongside Reason. Here’s how Gardner talks about Pinker:
“Many persons, including me, look forward to each new book by Steven Pinker.”
“[Pinker’s book, Rationality] does not disappoint. It is a thorough and careful discussion of what rationality is (and what it’s not)”
“I find myself in agreement with much of the book. Yet I also have significant reservations that the book will succeed in its aspiration”
“I’m not certain that Pinker would disagree strongly with what I have written. Certainly, in this master work, he gives attention to the importance of our relations with others, and our sense of obligations to the broader planet. But, on my reading, these points get overwhelmed by the importance that he places on our faculty of rationality.”
So Gardner doesn’t just say that relationships and respect are more important to him, he demonstrates what that looks like. While Alexander provides petty dismissals of people who disagree with him, Gardner makes sure to talk about his respect and appreciation for Pinker’s thought. Reading Gardner, you don’t think Pinker is stupid or even particularly wrong. He just emphasizes the inputs to his decision-making process slightly differently. In fact, he manages to emphasize these values without losing reason. His entire essay is laying out examples and reasons for what he believes. Alexander certainly doesn’t seem to find it convincing–which is fine. It’s a short essay in response to an entire book. I think there’s a lot left here that wasn’t spelled out (as in all models), like what weighing relationships against reason looks like in daily life. But at least Gardner seems willing to have a rational conversation with you. Alexander does not. Ultimately, the big takeaway from this article for me is that arguing about rationality is actually just a way of trying to smuggle one particular set of (“grey tribe”) values into a special, superior cultural position.