Crucial tests on hypotheses (in the form of modus tollens), where a hypothesis is decisively refuted, are impossible in science because it's possible to save a hypothesis from a false test by proactively revising auxiliary hypotheses or a change in test conditions. The following demonstrates the reliance upon auxiliary hypotheses:
If (H and A), then I
not I
not (H and A)
H stands for the testing hypothesis. I stands for the test implication which is derived from H and A. A stands for the auxiliary hypotheses which are subordinate functions to H usually required in order to derive I from H.
We can use Semmelwies theory to illustrate. Semmelwies thought, if doctors washed their hands between projects, then deaths would be reduced. An auxiliary hypotheses is, if a chlorine/lime washing solution is antiseptic, then hands would be sterile. If it wasn't antiseptic, then washing would have produced a false test.
The importance of auxiliary hypotheses becomes clearer when we consider the conditions under which the testing hypothesis occurs. Let C stand for a set of conditions that is derived from H and A. Now consider the proposition If C then E which is a test implication of H. In the Semmelwies case, if tools are used on cadavers, then they are not used on birthing mothers. Another doctor verifying Semmelwies may not follow this condition and not see a reduction in deaths.
In both scenarios, we suspect H; however, the lack of reduction in deaths doesn't falsely H or A. Auxiliary hypotheses may also have the opposite effect. For example, it could have been the case that the patients experimented on were resistant to infections, thus giving a false positive. Thus, it is impossible to have crucial tests that decisively refute scientific hypotheses.
Material is based on the writing of Hempel in the "Philosophy of Natural Science"
Word Count: 298