From: Lee Corbin (lcorbin@tsoft.com)
Date: Sat Jul 27 2002 - 12:34:53 MDT
Amara writes
> In a nutshell, I am convinced that Popper's falsification concepts and
> the Bayesian logic perspective don't overlap. They are really two
> different ways of looking at the world, and of formulating scientific
> problems. I've even heard some use the phrase: "Bayesianism makes
> falsification obsolete."
Yes, I'm somewhat acquainted with criticism of falsification per se,
and suspect that Popper tried to push the idea too far. Not all
good scientific theories are falsifiable IMO, and one of the chief
purposes of theories is to provide explanation, a topic most masterly
dealt with by David Deutsch in "The Fabric of Reality".
> ...a chapter exposing many (many) holes in Popper's falsification
> logical framework. Because of that, I've dropped giving weight to
> Popper's logical ideas.
I think that Popper still deserves enormous credit for contributing
to and popularizing the claims, such as that our minds work mainly
by means of conjecture and refutation, which lie at the basis of
Pan Critical Rationalism, which, IIRC, evolved from Popper's ideas.
> Besides the Jaynes link I posted previously about probability as a logic
> of science, here is something new (to me) that Serafino gave me on the
> topic that could be helpful:
>
> 'Bayesian view of everything' by Carlton Caves
> http://info.phys.unm.edu/~caves/
That's a fascinating essay, and quite readable for those with
some familiarity with thermodynamics and who have an interest
in probability theory. Until someone convinces me otherwise,
however, I find his claims like the one that we should regard
entropy as subjective, quite misleading if not downright wrong.
Taking his view, one would conclude either that "subjectivity"
was not something limited to the operations of minds, e.g.,
the higher animals especially humankind, ...or that the operation
of the entire universe depends on human consciousness or the
consciousness of some creature (which is a similar notion
popular among some physicists such as Wheeler, in his notorious
"participatory universe").
Perhaps one ought not aspire to a Bayesian view of *everything*.
Lee
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:15:42 MST