|Hypotheses: Hypotheses are assumptions made before performing experiments to compare the results of these experiments with them. Hypotheses must be fed by a given theory that is at least rudimentary, which determines what belongs to the domain of the objects involved, the concepts used and the possible consequences, and what cannot belong to it. In the course of the theory formation there is a mutual correction of assumptions and test results and the set of concepts and sentences of the theory. See also theories, methods, verification.|
Books on Amazon
Hypotheses/Duhem: A considerable group of experimental laws was established by observers. Theorists reduce them to a small number of hypotheses. But one can draw an unlimited number of conclusions from these hypotheses.
If, in theory, we are more likely to see a natural classification, we are inclined to anticipate and bet on their experience.
A theory, which we regard as a purely artificial system, will be seen as refuted by new emerging facts.
Hypotheses/Criterion/Duhem: Certain fundamental hypotheses cannot be refuted by any experiment because they form in reality definitions, and certain expressions used by the physicist only receive meaning through them.
E.g. free fall cannot be refuted (like Putnam's robot cats) When a heavy body falls free, its acceleration is constant. This cannot be disproved in the experiment because it is the definition of what one understands under free fall.
Le Roy: "Laws are not verifiable if they are taken strictly, because they themselves constitute the criterion."
Duhem: the words "free fall of a heavy body" now have a twofold meaning: for those who do not know the theories, they have a real meaning, for the physicist they have a symbolic meaning.
Duhem: the theory would not have fulfilled its task if the second sense was not the expression of the first. Only that the symbolic sense does not have the accuracy of the real and concrete case we observe.
Ziel und Struktur der physikalischen Theorien Hamburg 1998