Abstract
Replication of a finding is a sign – for some, the only sign – of scientific truth. Evidential discordance compromises truth, because discordance in scientific evidence means that a finding has not been reliably replicated. We distinguish between different types of evidential discordance, and single out persistent evidential discordance as a particularly serious problem for the epistemology of science. Building on Boyd’s (2018) notion of enriched lines of evidence, we propose a strategy for addressing persistent evidential discordance.
