RE: Bayesian epistemology versus Geddesian epistemology

From: Ben Goertzel (
Date: Sat May 28 2005 - 21:59:29 MDT

> A word of advice: don't waste your time
> with the "philosophy of science". (1) Bayes already presents a
> solution to the "problem" of induction. (2) Bayes presents a
> quantitative "Occam's Razor", obviating the need for determining
> which hypothesis is simpler.
> The applied Bayesian-statistician daily makes use of these
> solutions, while the philosophers will continue to argue about
> these solved problems for decades to come.
> When you have the calculus of science, of what use is the
> philosophy of it?

I don't agree with these statements at all, for the reasons why see my essay
on philosophy of science from not long ago:

Note that probabilistic considerations do play a significant role in my
thinking on this topic. I agree that probability theory is part of the
solution to the main problems of philosophy of science, but I don't agree
that it's an immediate or complete solution.

The third section of my article is entitled

"The Pragmatic Failures of Probabilism"

and it discusses how the probabilistic view of science, though in a sense
"correct in principle", is not actually all that useful for evaluating real
scientific theories or research programmes. Methods for partially getting
around this problem are discussed.

One issue is that real scientific theories of nontrivial scope are never
fully precisely defined nor fully formalized -- prerequisites for the
application of probabilistic mathematics to the problem of
validating/refuting scientific theories.

As a side comment, I also of course reject Marc Geddes' suggestion that
Bayesian inference and deduction are unrelated. The Probabilistic Term
Logic (PTL) framework used in Novamente integrates probabilistic deduction
with Bayes-rule-based probabilistic induction/abduction. And this
particular aspect of PTL is not unique.

-- Ben

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:51 MDT