[hist-analytic] My response
danny.frederick at tiscali.co.uk
Wed Aug 26 09:51:08 EDT 2009
I do not think the addition of "or probably true" makes any difference.
There is no method or procedure by which any proposition can be known to be
probably true. The probability calculus is like a logical calculus. Just as
in the latter we can derive truths from other truths, PROVIDED WE ARE GIVEN
SOME TRUTHS TO START WITH, so in the former we can calculate probabilities
from other probabilities, PROVIDED WE ARE GIVEN SOME (PRIOR) PROBABILITIES
TO START WITH. But in each case, the starting point is problematic. Whatever
axioms we adopt can be questioned; and it seems odds-on that every proposed
axiom will eventually be questioned by some competent people, if they have
not all been impugned already. And the prior probabilities of Bayesian
progic are assigned arbitrarily. Thus the best we have are methods of
ascribing truth or probability to propositions ASSUMING the truth or
probability of others. The regress can be stopped only arbitrarily.
This does not mean that rational debate is impossible. For, we can evaluate
rival theories according to how well they solve theoretical problems. In the
sciences, an essential part of this evaluation is consistency with accepted
observation statements and prediction of novel facts. In philosophy, this
often plays a role too, at least insofar as consistency with accepted
scientific theory can count in favour of a philosophical theory. But even in
the sciences, explanation and prediction of observation statements is not
the only means of assessment: considerations such as inherent simplicity and
coherence with other parts of our knowledge play an important role, and
these can play a role in the assessment of philosophical problem-solutions
too. Also, as I have said before, the dividing line between philosophy and
science is a shifting one.
You ask: 'if a philosopher makes a determinate claim that such and such is
the case, I would want to know why we should believe him or her. Why should
we take such a claim seriously?'
I have no interest in the question of belief: I am not inclined to believe
any philosophical or scientific theory. They are all guesses to be improved
upon. But we can take the philosopher's claim seriously if it is proposed as
a solution to a problem. For then we can evaluate it, compared to rival
solutions, in terms of how good a solution it is: does it solve the problem
or leave it unsolved? does it give a merely verbal solution that provides no
illumination? is it consistent with itself and with other theories that we
currently accept? And so on.
For further discussion of this, see Popper, 'On the Status of Science and of
Metaphysics' in his 'Conjectures and Refutations.'
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the hist-analytic