A "prior" denotes that which precedes, serving as an initial state or pre-existing condition. It represents the foundational knowledge or belief held before any new evidence or observation. In Bayesian inference, a prior distribution quantifies this initial belief.