|
Popper's
falsifiability is a logical criteria, not an experimental one. That is, the idea
of falsifiability is that, given the assertion in question, there can be
constructed a statement which can potentially logically falsify the assertion.
So, if the assertion is "All X are Y", then falsifiability means that there can
be constructed a statement, such as "There exists an X which is not Y", which
can then - at least in principle - by tested empirically.
He agrees that "it
is never possible to prove conclusively that an empirical scientifc theory is
false." That is, the best we can do is have a consensus on the logical level,
and then establish a refutation at that logical level. But, if we do not agree
on that level, then falsification is not possible. For example, if you have
a theory "All X are Y" and I show an example of where "There exists an X
which is not Y", you could alter your theory such that the case of X I showed
did not belong to your theory, either by changing definitions or using ad-hoc
hypotheses, etc.
See "Introduction
1982" in Popper's Realism and the Aim of Science
(Routledge, 1983).
Similarly, the use
of modelling relations in science relies on having a consensus of
encodings/decodings, since they are creative acts, unentailed from within the
modelling relation. If two people do not agree on the encoding/decoding, then
one may see the modelling relation as commuting and the other may see it as
not.
Regards,
Tim
|