[Top] [All Lists]

Re: [ontolog-forum] Semantics of Natural Languages

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Mon, 29 Aug 2011 22:30:08 -0400
Message-id: <4E5C4B30.8060107@xxxxxxxxxxx>
On 8/29/2011 2:29 PM, Rich Cooper wrote:
> But I also have trouble with the VivoMind example because
> it is closed - you have code and methods that have not been
> well defined and described for public consumption.    (01)

I was not mentioning any of those examples as subjects for developing
an ontology.  I was using them to illustrate the kinds of technology that
I have been talking about for years:    (02)

 1. Sparsely axiomatized lexical resources.    (03)

 2. Microtheories with detailed axioms for low-level problem-oriented
ontologies.    (04)

 3. A hierarchy for organizing and relating any or all theories.    (05)

 4. Guidelines for developing technologies that support #1, #2, and #3.    (06)

That is what I have been discussing in talks such as
     http://www.jfsowa.com/talks/iss.pdf    (07)

and in published articles such as
     http://www.jfsowa.com/pubs/rolelog.pdf    (08)

These slides present the ideas in terms of which I discussed the
VivoMind examples.  I was addressing a way of solving problems,
not specific technology.    (09)

> We need a SMALL example, a SIMPLE example, and a
> WELL DOCUMENTED world, IMHO, to make progress on
> Doug's already well developed nucleus of
> microtheory, which is a great starting point.    (010)

A starting point for what?    (011)

All small examples are toys.  The only kinds of examples that are not
toys are problems that somebody is willing to pay somebody to solve.    (012)

Is there any problem that anybody would be willing to pay somebody
to solve for which Doug's ontology would be useful?    (013)

> What can we choose that will be acceptable to all of us,
> small enough to make progress with, and yet able to clear
> up our early diversity of viewpoints?     (014)

I doubt that any such problems exist.  Any problem that can be
described simply won't illustrate the difficulties that exist in real
problems that people who build real applications need to solve.    (015)

Just look at the 50 years of research on toy problems in linguistics.
Chomsky started with "Colorless green ideas sleep furiously"
and Montague chose "John seeks a unicorn."  They developed
very impressive theories that are useless for NLP.    (016)

Hans Kamp was a student of Montague's, who got a summer job at
the Rand Corp. to translate an article from the _Scientific American_
into logic.  That job forced him to address serious problems that
neither Chomsky nor Montague considered.  Kamp's discourse
representation structures are more useful than any notation
developed by Chomsky or Montague -- but more is needed.    (017)

> At one point, you wanted to use biosemiotics and apply Peirce's
> thoughts to the problem.  So I tried to respond with Use Case 1,    (018)

Biosemiotics is a metalevel framework for organizing how to think
about problems.  It establishes criteria for organizing categories,
not a specific set of categories.    (019)

> We could map Use Case 2 onto Use Case 1 and perhaps
> deal with the self-interest versus interest issue.    (020)

The kinds of problems that can be clarified by semiotics or
biosemiotics arise in the development of a complete framework.
They don't show up in little examples.    (021)

In my previous note, I cited the debate between Barry Smith and
John Searle on issues related to Searle's book _Social Reality_.
That is the level at which such issues become critical.    (022)

Another area where such issues become critical is in the choice
of the upper level ontology.  Lenat said that the upper levels
weren't critical for solving most problems, but I believe that
a more satisfactory upper level could provide better guidance
about how to address the more detailed low-level tasks.    (023)

John    (024)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (025)

<Prev in Thread] Current Thread [Next in Thread>