Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!seismo!harpo!eagle!mhuxl!ulysses!unc!mcnc!ecsvax!unbent
From: unbent@ecsvax.UUCP
Newsgroups: net.ai
Subject: Re: Information sciences vs. physical sciences
Message-ID: <1709@ecsvax.UUCP>
Date: Tue, 13-Dec-83 09:14:58 EST
Article-I.D.: ecsvax.1709
Posted: Tue Dec 13 09:14:58 1983
Date-Received: Thu, 15-Dec-83 00:57:52 EST
Lines: 51

By rights, these comments should probably go into
net.philosophy.of.science, but ecsvax doesn't subscribe and
the issue was raised here in net.ai, so here goes.

Of course no-one holds a copyright on the word 'science'.
Practitioners of astrology *call* their discipline a science,
and so do practitioners of cosmetology and hair-styling.  It's
easy to understand why, too:  'science' is a *laudatory* term,
an honorific.  What's important, obviously, is the de facto
methodology, not the de dicto terminology.

The methodology of the "natural" sciences--paradigmatically
physics--is *explanatory*.  That is, the theories of a natural
science aim not merely at *systematizing* phenomena but at
*explaining* them, and progress in natural science is measured
by increases in explanatory scope and explanatory power.  It's
this that licenses the claims of natural science to be telling
the truth about "reality".  Appearance and reality are
logically connected by an explanatory 'because':  Things seem
as they do *because* things are as they are.  "Phenomena" are
"how things *seem*"; a theory wins epistemic credibility
within a natural science by offering an account of an
"underlying reality" which successfully explains *why* things
seem as they do.  (The "logical positivists" and "logical
empiricists" of the first half of this century got this
wrong.)

The "social" and "informational" sciences--sociology,
cognitive psychology, Chomskian linguistics, information
science--are *not* explanatory.  That's the significant
methodological difference.  They have a different criterion
for "getting it right".  (It's, in fact, fairly opaque just
what that criterion is:  some sort of coherent "reflective
equilibrium" between intuitions about cases and espousals of
principles, I think.)  There's nothing wrong with that, except
that lots of sociologists, psychologists, and contemporary
linguists evidently *think* that what they're up to is the
same as what, e.g., physicists are up to--and make a great
fuss insisting upon it.  (That's the battle that's typically
fought in terms of the question of whether such disciplines
are "really" sciences.)  Computer and information scientists
characteristically don't think that they're up to the same
sort of project as, e.g., physicists--and that's a good thing.
More power to them!  ("Getting it right" is often just a
matter of getting it to *run*.)

Sorry this is so long.  I'm a philosopher by trade, and we
tend to be a chatty lot.

                            --Jay Rosenberg  (ecsvax!unbent)
                  Univ. of North Carolina at Chapel Hill