Named after Pierre Duhem and Quine:

It is impossible to test a scientific hypothesis in isolation, because an empirical test of the hypothesis requires one or more background assumptions.