Falsifiability was proposed as the criterion demarcating empirical science from non-science.
Popper's 1934 criterion. Its use as a demarcation line is contested, with Lakatos's research programmes and the Duhem-Quine thesis both complicating it, and it survives best as a discipline for stating claims that could fail.
In practice
- A consultant whose framework explains every outcome after the fact has offered nothing that could have been wrong beforehand.
- It survives as a discipline rather than a test: a claim stated so that it could fail is more useful than one that could not, whatever its scientific status.
- Possible explanations
- The explanation is asymmetry: no number of confirmations establishes a universal claim, while one clear refutation overturns it, so the capacity to be refuted is what gives a theory empirical content.
- In its favour
- It remains an excellent working heuristic, it names what is wrong with theories that explain every outcome, and it survives in practice as the demand that a claim specify what would count against it.
- Against it
- As a demarcation line it is contested: the Duhem-Quine thesis shows that a refutation can always be deflected onto auxiliary assumptions, and Lakatos's research programmes describe what scientists actually do when an anomaly appears.
- Standing
- Its standing as a demarcation criterion is contested; Lakatos and the Duhem-Quine thesis both complicate it.
- Subject
- Science & Epistemology
- Essay
- Rules for weighing claims