Abstract
A population can be ideologically or affectively polarized. Ideological polarization relates to people’s political beliefs, while affective polarization deals with people’s feelings toward the ingroup and the outgroup. Both types of mental states, beliefs and feelings, are typically measured through direct self-report surveys. One philosophical assumption underlying this way of measuring polarization is a concrete version of the first-person authority thesis: the speaker’s sincerity guarantees the truth of their mental self-ascriptions. Based on various empirical studies, the first part of this paper argues that we are particularly bad at spotting our own mental states regarding political issues. This, in turn, raises doubts about the accuracy of direct self-report surveys in measuring polarization. In the second part, I introduce Michael Lynch’s notion of political meaning to argue that traditional surveys can still provide valuable information for detecting polarization. However, I suggest that this information pertains not to participants’ beliefs and feelings, but rather to their level of commitment to the core beliefs of the political groups they identify with, which is a relevant aspect of pernicious polarization.