Re: THE BIG PICTURE

sowa@west.poly.edu (John F. Sowa)
Date: Wed, 7 Jun 1995 10:30:57 +0500
From: sowa@west.poly.edu (John F. Sowa)
Message-id: <9506071430.AA00590@west>
To: cg@cs.umn.edu, srkb@cs.umbc.edu
Subject: Re: THE BIG PICTURE
Cc: sowa@west.poly.edu
Content-Length: 4406
Sender: owner-srkb@cs.umbc.edu
Precedence: bulk
I agree with most of what Doug Lenat wrote in his last note, and I
agree with even more of what Ed Hovy wrote (perhaps because his note
was shorter).  I would like to quote the following comment by 
Immanuel Kant after he presented his 12 categories in his _Critique
of Pure Reason_:

   From the little that I have said, it will be obvious that a complete
   glossary, with all the requisite explanations, is not only a possible,
   but an easy task.  (A:83, B:109)

I quoted that in Chapter 3 of my forthcoming book and followed it with
a comment of my own:  "Whenever a philosopher or mathematician uses
words like "easy" and "obvious", that is a sure sign of difficulty.
Kant used those words several times in the course of a page or two,
and after two hundred years, his easy task is still unfinished."

While I'm quoting philosophers, I'd like to mention the quote by
Leibniz (from _New Essays on Human Understanding_), which I am using
as the opening quote for my Appendix B, Sample Ontology:

   The art of ranking things in genera and species is of no small
   importance and very much assists our judgment as well as our memory.
   You know how much it matters in botany, not to mention animals and
   other substances, or again moral and notional entities as some may
   call them.  Order largely depends on it, and many good authors write
   in such a way that their whole account could be divided and subdivided
   according to a procedure related to genera and species.  This helps
   one not merely to retain things, but also to find them.  And those who
   have laid out all sorts of notions under certain headings or categories
   have done something very useful.

I also agree with both Doug and Ed that FOL plus metalanguage is essentially
all that one needs for almost all problems of knowledge representation.
And I firmly believe that progress in kn. rep. would be immensely aided by
having people put more effort into using their good old fashioned logics to
represent large volumes of serious work and less effort into inventing new
logics or notations with their own "personal stamp".

That last statement might sound odd coming from a person who has been
pushing conceptual graphs for a long time.  But when I present CGs,
I make it clear that I am NOT inventing a new kind of logic.  I am simply
adopting Peirce's second system of logic (graphs) instead of his first 
attempt (predicate calculus), which still has some appeal to mathematicians
(among which I consider myself), but unfortunately zero or a major negative
appeal to 99% of the rest of humanity.

But even while I use CGs, I make it clear that logic requires some notation,
but the underlying notions and techniques are independent of any notation.
In the ANSI standards work, I have been working with X3T2 and Mike G. in
making sure that CGs and KIF have an identical semantic foundation and that
anything represented in one can be translated to the other without loss or
distortion.  We have also been working very hard to show people that every
one of the popular notations in the database and programming design world
(e.g. SQL, Express, IDEF1X, Entity-Relationship diagrams, etc., etc.) can
also be mapped into logic, as expressed in KIF, CGs, or any other notation.

Re CycL:  I'm happy that Guha wrote a translator from KIF to CycL.
But we also need a translator from CycL to KIF plus some assurance that
there aren't any "minor details" left untranslated that will cause
major problems later on.  If there are features of CycL that are not
currently supported by CGs and KIF, then we must negotiate some resolution.

Re committees, workshops, etc:  I do my best, most productive, and most
creative work while working alone.  But after 30 years at IBM and several 
years of working on ANSI and ISO committees, I realize the need to negotiate
with other people who also prefer to work alone and to find some way to take
each other's points of view into account.  Some of my best ideas have come
as a result of wrestling with other people's problems and trying to
reconcile them with my own approach.  The most serious danger of working
alone is that you tend to concentrate on the things that your particular
system can do well, and you tend to ignore or downplay the hard stuff.
When you work with other people, they keep throwing the hard stuff in 
your face, and you are forced to deal with it.

John Sowa