Questions of societal import have both normative (what should be done) and descriptive (what is the case) dimensions. In this chapter, we address disagreements about the latter through the lens of political cognition, a research effort spanning cognitive psychology and experimental political science. We consider various explanations for fierce disagreements about "facts on the ground" and find that first, extensive evidence shows that judgments sort by political partisanship. Second, individual differences in threat sensitivity and uncertainty tolerance predict partisan group membership, which aids understanding of group differences but not necessarily differences in descriptive beliefs. Third, no evidence suggests a historically unusual deficit in scientific understanding. Fourth, some evidence implicates well-studied features of thought (confirmation bias, motivated reasoning), but these features appear to be symmetrical across the partisan divide. Finally, emergent evidence implicates people's assessments of information sources, processes that are also susceptible to motivated reasoning and partisan cueing and moreover are challenged by a media environment that is historically unusual. We conclude that the most promising conceptualizations of disagreement over empirical matters will acknowledge that partisan cues are often valid cues indicating the beliefs of people's groups and address difficulties of information source assessment in our current, unstable information environment.