[Top] [All Lists]

Re: [ontolog-forum] Semantics of Natural Languages

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Wed, 31 Aug 2011 21:19:44 -0400
Message-id: <4E5EDDB0.8050105@xxxxxxxxxxx>
Rich, David, and Azamat,    (01)

> Let me propose that it is SUBJECTIVITY i.e. a proper understanding
> of self interest, that I think is missing.    (02)

You can have a formal theory of anything, including subjectivity.
That is, in fact, why I mentioned semiotics because Peirce's
category of Thirdness includes subjectivity as a special case.    (03)

But just sitting down and writing a bunch of axioms is not going
to do anything useful, unless and until you have a fairly good
*informal* theory to begin with.    (04)

If you haven't done so already, please read the debate between
Barry Smith and John Searle.  You can also start by reading the
Wikipedia article about Social Reality, which has pointers to
Searle's book and several others:    (05)

    http://en.wikipedia.org/wiki/Social_reality    (06)

The subject is filled with complex issues, but one of my favorites
is the following quotation from the Wikipedia article:    (07)

> One aspect of social reality is the principle of the "the big lie",
> which states that an outrageous untruth is easier to convince people
> of than a less outrageous truth. Many examples from politics and
> theology, e.g. the claim that the Roman Emperor was in fact a "god",
> demonstrate that this principle was known by effective propagandists
> from early times, and continues to be applied to this day, e.g.
> the propaganda model of Noam Chomsky and Edward S. Herman, which
> supports the 'big lie' thesis with more specifics.    (08)

We've all seen major whoppers that politicians and theologians
have foisted on the public since antiquity.  And when politicians
work in collusion with theologians, the combination can become
explosive -- literally.    (09)

>> Why DON'T huge hunks of deduced, induced, abduced and reduced
>> knowledge suffice?  What is still lacking?  Why don't gobs of
>> special purpose functionality, coupled with gobs of knowledge,
>> do the trick?    (010)

The first thing that's lacking is an adequate amount of informal
knowledge.  Any mathematician will tell you that if you can't explain
your ideas adequately in your native language, you'll never be able
to state them in any version of mathematics or logic.    (011)

> #1 - much of life isn't subject to mathematical logic (e.g. much of
> business activity is highly illogical)
> #2 - life (particularly as expressed with language) is a constantly
> moving target, based on a poorly defined "foundation."    (012)

I agree more with David than with Rich.    (013)

That is why I recommended that we do something we know how to do:    (014)

  1. Use informal lexical resources (such as WordNet, for example)
     as a basis for the NLP.    (015)

  2. Develop narrow, special purpose theories for specific problems
     that we know how to solve.    (016)

  3. Develop NLP systems that use the resources of point #1 to relate
     natural languages to the theories of point #2.    (017)

> Personally I believe the good news is that the business thingys are
> not all that numerous.  I think there's some room to argue—definitely
> ARGUE—that organizations run on between 1500 & 6000 concepts.  But
> then it gets ugly since there are many, many, many synonyms for core
> concepts.  Remember my oft repeated: in 1980 a life insurance company
> found in its software systems 70 different names for the "policy
> number" concept.    (018)

> But I would still like to find ways to shrink the diversity of
> terms; I just don’t think it will shrink to that small a number.
> A thesaurus, by expression instead of by word or concept, might
> be a better goal than a minimalist number of utterances.    (019)

It's possible that if we develop computer systems that can understand
natural languages, we'll be able to use them for many useful projects.    (020)

But I believe that we're going to achieve the goal of a modest level
of understanding by points #1, #2, and #3 above rather than by writing
millions of axioms along the lines of Cyc.  As I said, if you want
to solve a real problem that some paying customer needs to solve,
focus your attention on that problem.  Don't write a bunch of axioms
with the hope that they might someday be useful.    (021)

John    (022)

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    (023)

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