What is falsifiability, and why did Popper rely on it?

Falsifiability is the requirement that a statement or theory be capable of conflicting with possible or conceivable observations. Karl Popper held that this, rather than confirmation by evidence, is what marks a theory as scientific. A theory that no conceivable observation could contradict, he argued, falls outside empirical science. He aimed this at religion, metaphysics and pseudoscience alike. [1] [2]

According to the Stanford Encyclopedia of Philosophy, Popper proposed that statements must be capable of conflicting with possible observations to be ranked as scientific. The test concerns logical form: a theoretical sentence is falsifiable if it logically contradicts some empirical sentence describing a possible observable event. A statement can be falsifiable even if it cannot be falsified in practice. Popper also said at times that a metaphysical idea of yesterday can become a testable scientific theory tomorrow, which sits uneasily with the purely logical reading. See [plato.stanford.edu](https://plato.stanford.edu/entries/pseudo-science/). [1]

The Internet Encyclopedia of Philosophy describes the idea as a kind of boldness. Scientific theories often clash with common sense, but so do many myths, so that alone is not enough. The second boldness is making testable claims that future observations might show to be false, which is a willingness to risk being wrong. On Popper's view, scientists make repeated honest attempts to refute a theory, while defenders of pseudoscience take steps to make observations fit the theory. See [iep.utm.edu](https://iep.utm.edu/pop-sci/). [2]

Oxford's Pseudoscience: A Very Short Introduction says Popper was struck by Einstein's prediction because of the risk involved: had measurements found Einstein wrong, he would have had to abandon his theory. Popper contrasted this with psychoanalysis and Marxism, which were widely seen as scientific in his Viennese setting under the logical empiricist idea of verification. He noted that these theories could absorb almost any data. A Freudian could claim that a given upbringing explained a man's homosexuality, and also that the same upbringing confirmed the theory when he was not homosexual. See [academic.oup.com](https://academic.oup.com/book/45901/chapter/403598107). [3]

For Popper, then, confirmation is cheap, because anything can be read as confirming a flexibly worded theory. The real question is whether any imaginable observation would make its defenders concede it was false. If not, it is not science. The Oxford author adds that it is not strong evidence against Freud or Marx that they fail this test, since the criterion was designed to exclude them. A Scientific American blog makes a similar point: scientific claims specify which observable outcomes would be impossible if the claim were true, while pseudoscientific claims fit any imaginable outcome. See [scientificamerican.com](https://www.scientificamerican.com/blog/doing-good-science/drawing-the-line-between-science-and-pseudo-science/). [3] [4]

Popper's view differs from verificationism in several ways. He did not call unfalsifiable claims meaningless, and he conceded that metaphysical statements may be far from meaningless. He also held that evidence can never give grounds for believing a theory true or likely true, so he is described as a fallibilist. Science, on his picture, progresses somewhat like evolution, with observations selecting against unfit theories by falsifying them. The Oxford text notes the appeal of this view: it offers a bright line between science and pseudoscience and rewards boldness. See [iep.utm.edu](https://iep.utm.edu/pop-sci/) and [academic.oup.com](https://academic.oup.com/book/45901/chapter/403598107). [2] [1] [3]