From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!zaphod.mps.ohio-state.edu!samsung!emory!iccdev!gwinnett!depsych!rc Tue Nov 19 11:09:42 EST 1991
Article 1274 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!zaphod.mps.ohio-state.edu!samsung!emory!iccdev!gwinnett!depsych!rc
>From: rc@depsych.Gwinnett.COM (Richard Carlson)
Newsgroups: comp.ai.philosophy
Subject: Is "logic" important?
Message-ID: <Pg99aB3w164w@depsych.Gwinnett.COM>
Date: 11 Nov 91 15:49:12 GMT
Lines: 91

I have been following the thread on "informal logic" with
considerable interest.  It seems to me that this relatively early
post raises some questions which haven't been explicitly followed:

(Mikail Zeleny writes:)
>I don't even have a general objection to inter-disciplinary studies of this
>sort.  My concern, if you wish, is of a conservative nature: I am quite
>comfortable with the traditional definition of logic as the study of
>arguments valid in virtue of their form, rather than merely plausible in
>virtue of the closed-world assumption.  The subject matter of logic is
>normative, rather than descriptive; perhaps this is why it fails to satisfy
>most people working in AI.  If you wish to study the patterns of actual
>human reasoning, you certainly don't need my or anyone else's permission to
>do so, but please call the resulting theory proairetics, hypoleptics,
>phronetics, or what have you, -- anything but logic.  There's plenty of
>room for argument over what constitutes a logical theory, viz.
>intuitionistic, modal, deontic, and perhaps even certain kinds of doxastic
>systems; however practical reasoning, regardless of its sterling virtues,
>simply doesn't make the grade due to its lack of validity and reliance on
>the subject matter.  By the way, I am not a Doctor of any sort, so perhaps
>this plea doesn't matter after all.

Mr. Zeleny's point is that he is defending a purist notion of
"logic" that restricts the meaning of the term to processes that
extract valid conclusions from premises.  He excludes the kind of
"logical empiricism" which generates knowledge that Kant would
have called synthetic and a posteriori or that Aristotle would
have called probable (e.g., Castro will be overthrown) or true
(e.g., Communist rulers in Eastern Europe were recently
overthrown) and retains only what Aristotle would have called
necessary.

Mr. McCarthy makes the point that informal reasoning can be
formalized and indeed is currently being formalized be workers
within the AI framework, and he gives quite a few examples.  He
proves, to my satisfaction, that reasoning can indeed be
automated.  Actually we don't even need AI and computers to do it.
Any explicit algorithm or set of algorithms that can be taught to
an uncreative person in a cookbook more or less proves that
reasoning can be automated, with some remaining issues of how much
implicit or "intuitive" knowledge the uncreative person following
the cookbook brings to the process.  (Picture a hack chef in an
overpriced New York restaurant recreating the latest masterpieces
from Paris.)

If that is so, just how important is "logic?"  How meaningful is
it that some thought processes contain the complete load of
information in their premises?  Are they more meaningful in some
ontological sense?  Is Mr. Zeleny reifying or hypostatizing some
eternal real of Platonic-like ideas when the necessity involved in
the inference may in fact be a trivial artifact of the definitions
employed?  (I suspect that the kind of spatial metaphor embodied
in Venn's diagrams may be the conscious or unconscious model
Aristotle had in mind when he developed his "logic," which he
called analytics.)

How important is either pure logic, i.e., in Mr. Zeleny's sense of
dealing with valid (necessary) conclusions, or even automated
informal reasoning (phronetics or whatever) to:

1.) the development of hard science.  I.e., could the thought
processes of a creative nuclear physicist be automated?

2.) ordinary living.  This realm, often identified with
"_Lebensphilosophie_," seems to require skills and "judgments"
that have only a minimal relevance to either logic or reason.

3.) social and ethical issues.  How compelling is that starving
baby in Africa which emerged later in the thread, and can we use
either pure logic or automated reason to generate normative
prescriptions about the baby?

4.) the dialectic.  Aristotle saw his analytics as a formalization
of a small part of the dialectical movement of thought.  Why we
are so fascinated by automated thought taking place outside the
human head on a piece of paper or in a machine is unclear.  A
moment's reflection is all that is required to realize that the
artificial thought is no more likely to arrive at metaphysically
true conclusions than the same process occurring inside a head,
but we are all, including me, fascinated by  both logic and AI.
But it seems to me that the most important aspects of human
thought are dialectical and therefore cannot be either formalized
or automated.  How can we devise a formula to give us the one and
only contrary to "hot" or "white?"  Is the contrary to white
"black?"  What's the contrary to "red?"

--
Richard Carlson        |    rc@depsych.gwinnett.com
Midtown Medical Center |    gatech!emory!gwinnett!depsych!rc
Atlanta, Georgia       |
(404) 881-6877         |


