The principal assumption made in the transition from Duhem's thesis to the stronger thesis, which I will call the Quine-Duhem thesis, is that when a conjunction of.
Thus the "Quine-Duhem thesis" has come to be the namefor the (presumably true) view that what is actually required for a testis not merely hypothesis, but the conjunction of that hypothesiswith a wide variety of often tacit hypothesesThe occurrence of an observation which refutes a prediction derived fromthe of the theory under test with only shows that at least memberof that conjunction is false, that the false assumption is necessarilythe . Whether the false prediction resulted fromthe falsity of the tested hypothesis or the falsity of some other one ormore of the assumptions is not known. Thus, accordingto the Quine-Duhem thesis, by suitable modification or buttressing of theproper auxiliary hypotheses, any theory can always be "saved" from potentialrefutation. Furthermore, it is often claimed that historical research showsthat scientists do frequently do precisely this. Hence it would seem that, Popper, theories cannot be definitively refuted any more thanthey can be confirmed.
In Defense of the Quine-Duhem Thesis: A Reply to Greenwood
"The Quine-Duhem Thesis," in Campbell, Heidi and Looy, Heather (Eds.) The Science and Religion Primer (Grand Rapids, MI: Baker Academic, 2009).