Re: Philosophy underlying LO? LO260

Jim Michmerhuizen (jamzen@world.std.com)
Sun, 26 Feb 1995 18:12:47 +0001 (EST)

On Mon, 20 Feb 1995, John Conover wrote in LO179:
[...everything snipped up to a sentence of Conover's...]
> be complete, (but some can, however.) The information theorist have
> been looking at scientific induction, recently, and there is some
> evidence (heaven forbid,) that scientific induction is logically
> inconsistent with itself, (I wouldn't even attempt to speculate on the

Love it. Classical scientific research _DEPENDS_ critically on bad logic.
I'm simplifying here, but they did it first:

We have a hypothesis

p implies q

and we do an experiment and find q to be true. And so we conclude that
p must be true. And of course that is bad logic.

I'll allow that I've simplified. The "hypothesis" that experiments are
supposed to test is - or really, logically, _must_ be - a comparative one:
if theory A is right, our experiment will show X results, but if theory A
is not right, then our experiment will show Y. But that doesn't help out
the research at all; it only means that we can disprove theories with a
lot more logical rigor than we can prove them.

I wouldn't be the least bit surprised to see a proof turn up one of these
days that scientific induction is in some way self-contradictory.

In an earlier post today I alluded to how every tiniest fact participates
or plays a role in an infinity of hierarchies and structures - not just
one. Some of these structures, for example, are "all of the _FALSE_
theories about the nature of the universe" (due credit here to Jorge Luis
Borges even if I _can't_ pronounce his name). There's no experiment
_within_ the facts that can force, by deductive logic, some interpretation
_of_ the facts.

Regards
jamzen@world.std.com
There are more different kinds of people in the world than there are people...