It must be possible for an empirical scientific system to be refuted by experience. Karl Popper o F p REDICTION AND p ROOF
By the early 1980s, the importance of impact in the solar system was established as a fact, as was the presence of high iridium concentrations in at least a few K-T boundary clay sites. That the Cretaceous had ended with a great mass extinction was also a fact, though the suddenness of that extinction was disputed. The Alvarezes invented a theory that tied these facts together. To explain the observational facts is merely the first obligation of a theory; often several do a good job of explaining at least some of the observations. The theories that prove to have lasting value go further: They predict new facts that have yet to be discovered. If these predicted facts are subsequently found, the theory gains strength. Curiously, however, a theory is never completely proven. The possibility always exists that some new evidence will come to light to discredit the theory, or that some clever scientist will come up with an alternative theory that explains more of the facts. Luis Alvarez never went so far as to claim that the meteorite impact theory had been proven, though he came perilously close. Typically he would assert only that the theory had met a large number of its predictions (and "postdictions," which are reasonable predictions that happened not to be thought of until later). Being human, however, he was not above, in Tennyson's phrase, "believing where we cannot prove."2
German philosopher Karl Popper has done more than anyone to advance the notion that scientific theories may be disproven, but never proven. He argued that for a theory to be called scientific, it must be possible to disprove, or falsify, the theory.3 For a theory to qualify as part of science, it must be possible to devise tests that, if a theory is wrong, will reveal it as wrong. If no such tests can be devised, then the theory is not useful, at least for the time being. This is one reason why premature theories languish: No one can think of anything useful to do with them. Popper did allow that theories could be "corroborated"; that is, they could prove their mettle by standing up to a succession of severe tests. Corroborate is a good word—it means "to strengthen or support with other evidence; make more certain." Corroboration falls short of proof, but shows that research is heading in the right direction.
Though philosophers and historians of science debate the utility of Popper's formulation and are apt to go on doing so, it jibes with our common sense to say that science advances not by proving theories right but by weakening them until they are falsified. Looking back at the history of science, it is clear that this is the way it works. Yet if one were randomly to select a scientist at work and ask, "What are you doing?" one would be apt to get the answer: "I am confirming such and such a theory." In their daily lives, most scientists try to confirm or extend theories, not to falsify them. In part this is because scientists are rewarded for breakthroughs, not for falsification. Rewards aside, however, human beings will not spend long hours and entire careers searching for falsity. Thus a contradiction exists between the way individual scientists behave and the way science as a whole evolves—as the cumulative result of the work of all scientists. A host of them, each trying to shore up their favorite theories, will in time lead to the falsification of the weakest, to the great disappointment of its proponents but to the advancement of science overall.
Was this article helpful?