popper-logic-scientific-discovery

246
some structural components of a theory of experience
establish indeterminism reveal a mode of thought which can only be
described as determinist, in the metaphysical sense. (Heisenberg for
instance tries to give a causal explanation why causal explanations are
impossible.* 5 ) I may just remind the reader of the attempts to demonstrate
that the uncertainty relations close some avenues of possible
research, as does the principle of the constancy of light velocity: the
analogy between the two constants c and h, the velocity of light and
Planck’s constant, was interpreted by saying that both set a limit, in
principle, to the possibilities of research. Questions raised in the
attempt to grope beyond these barriers were dismissed by the wellknown
method of dismissing unpalatable problems as ‘pseudo’. In my
view there is indeed an analogy between the two constants c and h; one
which, incidentally, ensures that the constant h is no more a barrier to
research than the constant c. The principle of the constancy of light
velocity (and of the impossibility of exceeding this velocity) does not
forbid us to search for velocities which are greater than that of light; for
it only asserts that we shall not find any; that is to say, that we shall be
unable to produce signals that travel faster than light. And similarly, the
Heisenberg formulae ought not to be interpreted as forbidding the
search for ‘super-pure’ cases; for they only assert that we shall not find
any; and, in particular, that we cannot produce any. The laws forbidding
velocities greater than that of light and ‘super-pure’ cases challenge
the investigator, as do other empirical statements, to search for
the forbidden. For he can test empirical statements only by trying to
falsify them.
From an historical point of view, the emergence of indeterminist
metaphysics is understandable enough. For a long time, physicists
believed in determinist metaphysics. And because the logical situation
was not fully understood, the failure of the various attempts to deduce
the light spectra—which are statistical effects—from a mechanical
model of the atom was bound to produce a crisis for determinism.
Today we see clearly that this failure was inevitable, since it is impossible
to deduce statistical laws from a non-statistical (mechanical)
model of the atom. But at that time (about 1924, the time of the theory
* 5 His argument is, in brief, that causality breaks down owing to our interference with
the observed object, i.e. owing to a certain causal interaction.