Rational Polarization

Download Edit this record How to cite View on PhilPapers
Abstract
Predictable polarization is everywhere: we can often predict how people’s opinions—including our own—will shift over time. Empirical studies suggest that this is so when evidence is ambiguous. That fact is often thought to demonstrate human irrationality. It doesn’t. Bayesians will predictably polarize iff their evidence is ambiguous. And ours often is: the process of cognitive search—searching a cognitively-accessible space for an item of a particular profile—yields ambiguous evidence that can predictably polarize beliefs, despite being expected to make them more accurate. A series of such rational updates can lead to polarization that is predictable, profound, and persistent. This process is not only theoretically possible, but also empirically plausible. I present an experiment supporting the polarizing effect of cognitive search, and then use models and simulations to show how such ambiguous evidence can help explain two of the core causes of polarization: confirmation bias and the group polarization effect.
PhilPapers/Archive ID
DORRP-2
Upload history
First archival date: 2021-09-04
Latest version: 4 (2021-10-17)
View other versions
Added to PP index
2021-09-04

Total views
469 ( #13,244 of 2,448,376 )

Recent downloads (6 months)
469 ( #716 of 2,448,376 )

How can I increase my downloads?

Downloads since first upload
This graph includes both downloads from PhilArchive and clicks on external links on PhilPapers.