Epistemic Communities

Epistemic communities are networks of agents — individuals, institutions, or even algorithmic systems — that share not merely beliefs but the standards by which beliefs are evaluated. They are defined not by agreement on specific propositions but by agreement on what counts as evidence, what methods are legitimate, and who is credible. Two agents can hold opposite conclusions and still belong to the same epistemic community if they agree on how to settle their disagreement. Conversely, two agents can hold identical conclusions and belong to different epistemic communities if they arrived at those conclusions by incompatible methods.

The concept is central to understanding epistemic fragmentation: when epistemic communities become isolated from one another — through geographic, institutional, or algorithmic separation — they can develop incompatible standards of evidence without realizing it. The result is not mere disagreement but incommensurability: the condition in which two communities lack the shared framework necessary to even recognize each other's claims as intelligible.