Reading Recommendations for Rationalists

Here are a couple of somewhat congruent interactions between the deeply idealistic rationalist community and myself in the role of cynical old bastard.

(1) I was at a LessWrong event the other night where Ethan Dickinson spoke on social interactions and the problems that can arise and various ways of dealing with them. Here՚s the blurb:

Imagine a world where people cannot truly know each other. Miscommunication abounds. Well-intentioned remarks offend, while obvious truths are hidden or denied. Certain actions are deemed admirable by some, yet annoying or abhorrent by others. Entire communities enter into downward spirals of evaporative cooling, unproductive bickering, and bitter tribalized feuding.

This is the dystopian mindscape we find ourselves in when we fall prey to the biases and heuristics surrounding mental modeling and communication. What are these flaws in our understanding of each other? Is there a realistic path of self-improvement that can lead us to become better predictors and communicators? Can individual-level skills be parlayed into community-wide improvement?

My very immediate reaction to the first sentence was, wait, what other kind of world is there? Maybe the failure of imagination is mine, but I cannot imagine a world where we “truly know each other”, or even truly know ourselves for that matter. One thing you can say for humans, they are complex, and even for the people we know very well, there are always untapped depths.

But let՚s assume that we don՚t take “truly” overly literally. In fact, delete that sentence and I can՚t find anything objectionable at all. Still, I am not sure the idea of accurate representation of other people is the right model for human communication, any more than accurate physical cosmology is the right model for religion. Religion is about participating in ritual community; the truth of the words of a prayer have very little to do with it. And ordinary human communication also has a ritual quality to it, it is about expressing emotions; satisfying needs; finding, signalling, and reinforcing political/tribal realities; passing moral judgement; or simply enacting social roles. People interacting are only incidentally building more accurate models of each other; there is usually something else going on, and that something else has to be recognized and acknowledged.

My reading recommendation: All of Erving Goffman՚s work on strategic social interaction. Start with “On Cooling the Mark Out” and for more, see books like Interaction Ritual and Stigma: Notes on the Management of Spoiled Identity.

(2) Here՚s Scott Alexander making a typically lengthy and tightly-written fret about the fact that people use all sorts of irrational techniques to reinforce their beliefs, that tribalism is both a cause and a consequence of all sorts of questionable kinds of reasoning, and that lot of what should be rational debate is actually just mindless cheerleading for your side or hatred of the other side.

I hate to argue with him because he is so accurate in his diagnoses and so high-minded in his solutions. But essentially he wants human nature to change, for everybody to put down their emotion-based alliances and think objectively, which in cases like these mean among other things being able to empathetically take the viewpoint of an opponent.

As in the first case, this seems like a great idea but in some respects it misses the point of political speech, which is not primarily about reasoned debate and more about forming coalitions of power.

My reading recommendation: All of Bruno Latour՚s work on the politics of knowledge, starting with Science in Action. Also his bit on Socrates and Callicles in the more recent Pandora՚s Hope.

(3) Both rationalists that I am riffing off of are disturbed by the negative effects of emotion and power in human affairs. Rationality is seen as a corrective, a way of thinking (and being) that is at least in part insulated from such destructive forces. As a goal, that seems hard to argue with. But it may be that emotion and power are too fundamental to human behavior to be papered over by the rather thin layer of rationality available to us.

One could make an argument that these authors are actually more in tune with a certain kind of rationalism than positing a more naive sort of default agreeability. After all, there is no a priori reason why one agent in a conversation or other social relation should have the same goals as any other agent. Presumably they are out for themselves, and if they can find common cause, that՚s wonderful, but it՚s an achievement that has to be accounted for. And in all likelihood not as stable an achievement as we might like. So a properly rich rationalism would treat human interaction as more goal-oriented than accuracy-oriented.

The authors I am recommending have something in common: they both, in quite different ways, try to deal with the reality that humans are power- and status-seeking creatures well before they are truth- and comity-seeking. And whatever success they have at the latter is built using the machinery developed for the former. Latour and Goffmann both have developed a rich set of methods for describing the relationship between power and knowledge: Goffman applies the vocabulary of drama to ordinary life, and Latour merrily dispenses with normal ontological distinctions so he can describe power alliances between people, ideas, machinery, and nature. Conflict is an essential part of their world-pictures: people and other things have their own interests, the world is a chaos of competing interests, not a well-behaved unity.

This point of view may seem superficially cynical. But the depth of these authors՚ intellectual humanism elevates their work above mere cynicism. It is clear that despite seeing the often dirty and ugly machinery that underlies cognition and society, they still retain fondness for humans and their complex processes. And my argument for reading them is not based on either cynicism or idealism, but realism. They add a layer of depth to our understanding of human social processes.