History is the science of things which are not repeated.

Valéry’s paradox turns on what “science” usually implies: repeatable phenomena, controlled comparison, and general laws. History, by contrast, deals with unique, unrepeatable events—wars, revolutions, decisions, accidents—whose exact conditions cannot be recreated. Calling history “the science of things which are not repeated” is both ironic and cautionary: it suggests that historical knowledge is necessarily interpretive and probabilistic, not experimentally verifiable in the way physics or chemistry is. The line also warns against treating historical patterns as reliable predictions; at best, history offers analogies and insights into human behavior, not repeatable proofs.