🚀 go-pugleaf

Initial hypotheses (axioms, postulates), according to Karl Popper,
should be "bold", which means that the prior probability that they are
false is relatively high. Then Popper seems to see no way of getting
rid of false postulates unless the theory as a whole is rejected
because of conflict with the so-called basic statements "whose
acceptance or rejection, in its turn, depends upon our decisions".
Even in the presence of such a conflict however false postulates could
still be saved: Popper: "On the other hand, I also realized that we
must not exclude immunizations, not even all which introduce ad hoc
auxiliary hypothesis....All this shows not only that some degree of
dogmatism is fruitful, even in science, but also that logically
speaking falsifiability or testability cannot be regarded as a very
sharp criterion."

In fact, unless science is to become a religion, postulates can be
checked and rejected if false independently of the development of the
theory. Consider the following scenario. As an observer moves towards
a light source, he finds that the frequency of coming light increases
with his speed (Doppler effect). The observer infers, in accordance
with the formula:

frequency = (speed of light)/(wavelength)

that the speed of light (relative to him) also increases with his
speed. So the observer discovers that Einstein's 1905 light postulate
is false and rejects it. However, if science is to become a religion,
the observer may not wish to reject Einstein's 1905 false light
postulate. So he considers the only alternative the formula above
offers:

The wavelength decreases as the speed of the observer increases.

If the observer is clever, he soon discovers that the varying-
wavelength alternative is just as silly as the statement:

The distance between trees alongside the road decreases as the speed
of the car increases.

From now on the clever observer will know that Einstein's 1905 light
postulate is false but, since science is to become a religion, he will
keep this knowledge in secret.

Pentcho Valev
pvalev@yahoo.com