History is littered with scientific errors: bodily humors, atomic affinities, mislabeled planets... This contrasts sharply with the image of science as "self-correcting" and as providing a systematic method that reliably yields trustworthy knowledge.
Toward a Philosophy of Error in Science seeks to resolve this puzzle.
By carefully dissecting cases from history, we can catalog a vast inventory of the sources of error. The sensational claim that chronic fatigue syndrome was caused by the XMRV virus? A contaminated commercial reagent. Newton's faulty formula for the speed of sound, which even the greatest physicists could not fix for over a hundred years? Lack of collateral knowledge about adiabatic phenomena. How did Boyle's law become a universal "law," if it has so many exceptions and conditions? Overgeneralization.
History also allows us to document how each error was corrected. How did the OPERA team track down the loose cable that led to international news headlines about a faster-than-light neutrino? How did the Mesmer Commission debunk animal magnetism? How did happenstance help unravel the entrenched view that stress and diet caused ulcers? How did the craniological practice of measuring intelligence, sustained for decades by racist and sexist ideology, eventually unravel in just a few years?
We can develop more effective methods for regulating errors in the future. We can also rethink our views of progress in science, the role of "bias," and even our very concept of knowledge.
