[Top] [All Lists]

Re: [ontolog-forum] What do ontologies have to do with meaning?

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Leo Obrst <lobrst@xxxxxxxxx>
Date: Mon, 14 Jun 2004 10:54:18 -0400
Message-id: <40CDBC1A.D6E2EA8E@xxxxxxxxx>
A bit late to this discussion, but I'd say that some of the confusion is
due to the following points:    (01)

1) The original workshop on the relations between meaning and
ontologies: this is a cognitivist workshop. Starting about 20 years or
so ago, cognitive linguistics arose that tried to address
meaning/semantics from a non-formal point of view, meaning grounded
nearly purely in human-experiential and cognitive terms. Gardenfors
recently (Conceptual Spaces) tried to show  that much of "meaning" is
underlying cognitively derived and results in perhaps our formal
symbols. This is sometimes called the "symbol-grounding" problem in AI,
meaning, how do our symbols bottom-out? I think very many scientists
would not deny that our formal representations (conclusions, symbols,
etc.) are grounded in neural and cognitive events. Are our logical,
mathematical, scientific artifacts, generalizations, etc., including
everyday generalizations/patterns, just "spikes" in our neural
substrate, i.e., emergent order, etc.? I think that is probably so, but
those "spikes" arise as they do because they represent something outside
of ourselves that is recognized by our mind-bodies.     (02)

2) Meaning (semantics), even in the formal tradition, really rests on
two poles (originally due to Frege): sense and reference. Sense refers
typically to the "conceptual structures" or mental models, i.e., the
formal structures of science, logic, mathematics, etc., that we use to
construct ontologies with. Reference refers to the relations between
these formal (mental, conceptual) constructs and real (or possible or
impossible) objects in the world. A good theory of reference is still
lacking, since it's a very hard problem.     (03)

3) "Literal" meaning is sometimes what is meant by "semantics" formally,
as opposed to "pragmatics", which deals with semantics in context, its
use and the intent of the producer (of the utterance, text, etc.) But
really, semantics-pragmatics can be viewed just the same notion, a
continuous spectrum, and we just call out different "halves" of that
spectrum when we want to make specific points. I think that
"semantics-pragmatics" in principle, formally (using logic, math, etc.)
can spell out unambiguously what we mean unless what we mean is
inherently ambiguous or vague. Vagueness is about whether an
instance/individual falls in the positive or negative extension of a
predicate, i.e., the boundaries of the predicates are vague. Example:
"close to New York City" or "greater London" or "part of Mt. Everest" or
"red", etc. Is Newark "close to New York City"? Usually, but not always,
borderline examples can be made to fit in either the positive extension
(yes, it is in X) or the  negative extension (no, it is not an instance
of X), if greater refinement is made. I.e., additional assertions,
conditions, constraints, are adopted. And vagueness can be formally
addressed.    (04)

Indeterminacy and vagueness are not the same thing.     (05)

In recent years, I've found Achille Varzi's papers on the topic to be
especially of interest. See:
http://www.columbia.edu/~av72/Vagueness/syllabus.pdf for his course in
98 for the range of issues. There have also been a few recent books on
vagueness.     (06)

To return to the original point, semantics-pragmatics does have to
address "meaning" in context, and it does.    (07)

Leo    (08)

Kurt Conrad wrote:
> Hmmmm.....
> I don't recall the meeting where we assigned Adam the task of
> deciding what topics are relevant to this forum.
> 1) If a topic generates discussion, that seems to be prima facie
> evidence of relevance.  I found the exchanges to be both very
> interesting and (from my admittedly biased position) quite useful.
> 2) Our charter defined the scope of this forum using the phrase
> "ontologies used in business".
> Until all business is done entirely by machines and we are all
> hopelessly unemployed, I expect that business will have to continue
> to rely on agents that learn from experience (regardless of whether
> or how well they communicate).
> Speaking from my own, personal experience, I have trained a number of
> the agents that I have worked with _not_ to communicate with me,
> directly, demonstrating an interesting ontological mix of both
> learning and lack of communication.  I have also found that the most
> critical business issues routinely involve ontological breakdowns
> that occur well outside of communication, formalization, and even
> automation.
> Mike, please don't hesitate to post any questions to this forum,
> including (or even, especially) those that deal with
> "non-communication semantics".  I believe that a better shared
> understanding of literal meaning, true meaning, and a variety of
> other types of meaning can only help businesses, and all other forms
> of human organization, to operate more effectively.
> /s/ kwc 2004.06.13 23:11
> At 2004.06.09 13:58, Mike Brenner:
> >Hi Adam,
> >
> >Okay, I'll not post non-communication semantics questions
> >to "ontolog" any more.
> >
> >Mike
> >
> >
> >Adam Pease wrote:
> > >
> > > Mike,
> > >    In your original post you commented:
> > >
> > > "Thus, I don't see meaning as related to conceptual forms in the
> > mind of
> > > the reader, but rather to physical forms which are the context in
> > which the
> > > communication takes place."
> > >
> > > Are you now addressing agents which don't communicate, and
> > somehow learn an
> > > ontology from experiences?  If so, then that doesn't seem
> > relevant to this
> > > forum.
> > >
> > > Adam
> ___________________________________________________________________
> Kurt Conrad
> 2994 Salem Dr.                     408-247-0454
> Santa Clara, CA 95051-5502         408-247-0457 (data/fax)
> http://www.SagebrushGroup.com     mailto:conrad@xxxxxxxxxxxxxxxxxx
> _________________________________________________________________
> Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
> Subscribe/Unsubscribe/Config: 
> Shared Files: http://ontolog.cim3.net/file/
> Community Wiki: http://ontolog.cim3.net/wiki/
> To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>     (09)

Dr. Leo Obrst       The MITRE Corporation, Information Semantics
lobrst@xxxxxxxxx    Center for Innovative Computing & Informatics 
Voice: 703-883-6770 7515 Colshire Drive, M/S H305
Fax: 703-883-1379   McLean, VA 22102-7508, USA    (010)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (011)

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