Class 4.1: Belief Polarization
As we’ve seen, If we want to understand what brings people into conflict with one another, one promising way is to start by looking for rationalizing explanations of their conduct. Rationalizing explanations appeal to the beliefs and goals or values of the individuals, and show that these combinations of beliefs and values would make some combination of choices or actions rational from their individual perspectives, and so just what we should expect people to do.
So far, we’ve been focusing on the “goals” or “values” part of this equation. When people have different or competing values, as Hobbes observed, that can put them into conflict. And it is very easy - inevitable given the relative equality of the state of nature - for people to have such competing goals. We also took these ideas further and explored them from other directions using Gauthier and Nguyen.
Now this week it’s time to turn to consider the “beliefs” part of this equation. Even when two people have exactly the same goals, they can be put into conflict with one another if they have different beliefs about how to achieve those goals. We begin with an article by the philosophy Tom Kelly, before turning on Wednesday to recent work once more from Thi Nguyen.
Tom Kelly
Tom Kelly is a philosopher who teaches at Princeton University. He is most recently the author of Bias: A Philosophical Study, and has consistently been interested throughout his career in how to understand issues about how knowledge and rational belief are affected by the acknowledgement that we live in a world of difference.
Disagreement, Dogmatism, and Belief Polarization
In the article that we are reading for today, Kelly explores the phenomenon of what he calls “belief polarization”. Kelly doesn’t define this term exactly, and there are actually several different things that we could mean by it, that have been studied by other people. For example, on one meaning of “belief polarization”, beliefs count as polarized if there are people who believe very different things. Another way in which beliefs can be “polarized” is if there is a strong correlation between the beliefs of two different groups of people, so that there is a lot of disagreement between groups about a lot of different topics but not a lot of disagreement within groups. We observe both of these things in the real world in, for example, the contemporary United States. But in his paper Kelly is focusing on something more more specific: the empirical evidence that when people who start with different beliefs encounter new evidence, they end up more entrenched in their original beliefs, rather than less.
The fact that people end up sticking to their original beliefs - or even more confident in them - after reviewing mixed bodies of new evidence about complex questions can help to explain why disagreements between groups of people will be hard to resolve. And so long as it is a psychologically real phenomenon, it can help to explain this even if it is an irrational process. But Kelly is interested in the extent to which it is a rational process. He is interested in whether it might actually be rational for people to end up sticking to their original beliefs after encountering new evidence.
The answer that Kelly comes to is relatively sophisticated and subtle. He neither says that it is straightforwardly rational to stick with your beliefs when you encounter new evidence, nor that this is never rational. Try to write down in 1-2 sentences what you think Kelly’s takeaway is. When we start class on Monday, we will begin by trying to compare each of our descriptions of the takeaway moral from his paper, as an exercise in careful reading and understanding.
