1970년, 반증을 만난 과학은 후퇴를 발명했다

A single failed prediction, we are often told, is enough to bring down a scientific theory. But Imre Lakatos, writing from the London School of Economics in 1970, argued that the history of science rarely cooperates with such a tidy picture, and he proposed that what scientists actually defend is not an isolated law but a research programme. Around its hard core of fundamental commitments lies a protective belt of auxiliary hypotheses, which absorbs anomalies and is steadily revised as evidence accumulates. When Uranus drifted from its calculated path in the 1840s, John Couch Adams and Urbain Le Verrier did not abandon Newtonian gravitation; they posited an unseen planet, and in 1846 Johann Galle observed Neptune almost exactly where Le Verrier's calculations had placed it. The same maneuver, repeated for Mercury's anomalous perihelion through the hypothetical planet Vulcan, yielded nothing, and the puzzle dissolved only with Einstein's general relativity in 1915. Lakatos drew from this contrast a criterion that orthodox falsificationism had missed. A research programme deserves continued allegiance, on his account, when its repairs yield novel predictions that later observation confirms, while a degenerating one merely patches its protective belt with ad hoc adjustments designed to shield the core from refutation.