Theories are judged not one test at a time but by whether their research programme keeps predicting novel facts or merely absorbs anomalies.
Lakatos's 1970 framework distinguishes a programme that predicts new facts from one that only accommodates old ones after the event. It is the standard reconciliation of Popper with Kuhn.
In practice
- A theory that predicted a new particle and then found it is progressing; one that has explained every failed prediction after the fact is not.
- A management framework that keeps generating testable claims is worth more than one whose defenders explain every failure as poor implementation.
- Standing
- The standard reconciliation of Popper's falsificationism with Kuhn's account of how science actually proceeds.
- Subject
- Science & Epistemology
- Essay
- How knowledge grows, and how inquiry fools itself