Pseudoscience

Popper and the Demarcation Problem

Popper began working on the problem of demarcation between science and pseudoscience (as well as more generally nonscience) as early as 1919. He was particularly concerned with David Hume's famous problem of induction, the idea that there does not seem to be a logically independent way to justify inductive reasoning, the basis for the scientific method. Popper thought he had arrived at a single idea that represented both a solution to Hume's problem as well as a clear-cut demarcation criterion: falsificationism. He proposed that science is in the business of advancing falsifiable (i.e., refutable in principle) theories about how the world works. This appeared to have bypassed Hume's issue about induction because falsificationism can be thought of as an application of modus tollens, therefore relying on deductive, not inductive, reasoning. At the same time, it seemed to Popper that pseudosciences (among which he counted various schools of psychoanalysis as well as Marxist theories of history) made statements that were not falsifiable, and were thereby unscientific.

While Popper's contribution to both issues remains a fundamental starting point for any discussion of demarcation and induction, there are good reasons to believe that he was a bit too quick in [Page 767]declaring victory on both fronts. This entry directs the reader to two comprehensive articles concerning the problem of induction and focuses instead on falsification as a demarcation criterion.

It is easy to show that falsification leaves much good science out and allows a significant amount of pseudoscience in. For instance, the history of science is riddled with examples of scientific hypotheses that—when first proposed—were apparently falsified by the data and yet scientists kept them alive because they seemed promising enough. The initial version of the Copernican system, with its circular planetary orbits, was doing no better empirically than the Ptolemaic system it was supposed to replace, and it was not until Kepler realized that the planets move along elliptical orbits that the theory was vindicated. Copernicus's book was published in 1543, but it was not until 1609 that Kepler put out his fundamentally revised version of the theory.

On the other side of the divide, so-called scientific creationism does make perfectly falsifiable predictions, such as that the earth is only a few 1,000 years old. These predictions have indeed been amply falsified by modern geology, physics, chemistry, and biology, and yet there does seem to be a strong sense that we should not simply consider creationism a science, even a failed one (for one thing, because of its appeal to supernatural, by definition inscrutable, forces, which are themselves outside the purview of science).

...