I study philosophy of science and fail to grasp the idea of conventionalism (Poincare, Duhem) that came to replace inductivism. What is the idea of conventionalism and why was it offered? What is its strength? Is it sort of instrumentalism?


Disclaimer: what follows is greatly simplified. Conventionalism in the philosophy of science can be understood as a rejection of foundationalism - the idea that statements about the world can be deduced (or even induced) directly from observations that are infallible. Conventionalists point out that observations are potentially ambiguous and require interpretation, and in any event, theories are under-determined by data so one cannot have a single theory or set of statements about the world that are uniquely deducible from one's observations.

More extreme versions of conventionalism (LeRoy, Ajdukiewicz) reject objectivity entirely and hold that conventions of language, measurement and methodology mean that one cannot have any kind of objective facts, statements or theories.

Conventionalism is different from instrumentalism, which in effect holds that scientific theories are not true at all, just instruments of prediction. Most conventionalists, at least of the moderate kind, hold that while we can have more or less good theories about the world, the best we can hope for is for these theories to be coherent and to do a good job of making sense of observations by their own lights. As such, conventionalism is often associated with a coherentist account of truth.

| improve this answer | |
  • 1
    Thanks. What is then the difference between conventionalism and falsificationism? (maybe it is a separate question to raise in site) – L.M. Student Nov 28 '15 at 0:07
  • 2
    The arch-falsificationist, Popper, was opposed to conventionalism, because he saw it as abandoning objectivity. But we might interpret falsificationism as a kind of meta-level principle. It holds that theories are scientific to the extent that they make falsifiable predictions. This could be consistent with moderate conventionalism, because one could maintain that while individual statements and even theories are not falsifiable in isolation, one's web of belief is falsifiable as a whole. – Bumble Nov 28 '15 at 1:20
  • 2
    ...This is a familiar idea from Quine's holism, under which when we discover data that contradicts what we believe we make adjustments on the basis of minimum mutilation, but we cannot say of a particular sentence that it is testable in isolation from all the other theories and hypotheses we hold to be true. – Bumble Nov 28 '15 at 1:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.