Re: [ontolog-forum] "I don't believe in word senses." Sue Atkins

[Gary Berg-Cross <gbergcross@xxxxxxxxx>]

John B

I might largely agree with this process view - we use a process that in part accessed our background and linguistic knowledge to understand a word meaning in the context of it communications usage.  

Unfortunately we are proabably still a reasonable distance from using computers in this way. 

The interesting questions is how  to begin to go about this.  The Cyc project thought that a large concept/word KB and logical reasoning would do it in 10 years or so of effort, but alas it hasn't quite.  There are various other cognition systems and their parts involved in understanding.

Gary Berg-Cross, Ph.D.  

gbergcross@xxxxxxxxx     

http://ontolog.cim3.net/cgi-bin/wiki.pl?GaryBergCross

NSF INTEROP Project  

http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=0955816

SOCoP Executive Secretary

Knowledge Strategies    

Potomac, MD

240-426-0770

On Fri, Oct 18, 2013 at 6:34 AM, John Black <JohnBlack@xxxxxxxxxxx> wrote:

I would paraphrase this as: "I don't believe in static word senses."  As I see it, a word invokes a process.  In the past we couldn't record processes, and so we used static tools such as dictionaries.  Today, we have computers and can begin to index the processes that make sense of a word by the words that invoke them. So I agree with that article and John S. that the word senses recorded in static dictionaries are abstractions. They abstract away from the process of making sense of a word in a situation. 

 But there is another abstraction involved in the historical approach to making word senses.  That is the use of corpus citations collected in static corpora as the basis of forming word senses.  What this amounts to is the creation of multiple snapshots of the process of making sense of a word. The corpus citations used for this purpose represent and abstract from people speaking and listening. Thus I would go one further than Sue and Adam and article cited and say that individual utterances, not the corpus citations, are the basic objects in forming ontologies. 

However, I agree more with John's response that with Rich here that: 

RC

It would seem to be a reasonable conclusion that ontologies are
personal systems of what each individual believes exists.

Different people will interpret the same text in different ways.
But the author who wrote or spoke the text did have something in mind
at the moment of writing or speaking it.  Some readers or listeners
will have a better or worse interpretation of what the author meant.

Because, to the extent that common knowledge among populations of word users of the optimal process of making sense of a particular word in a particular situation as been achieved, the abstractions work. So our ability to communicate, using a particular word, is proportional to the common knowledge we share of the process to make sense of that word in any given situation. 

John Black

On Oct 15, 2013, at 5:19 PM, John F Sowa <sowa@xxxxxxxxxxx> wrote:

Gary, Rich, David, and Melvin,

GBC

Still would we agree that "definitional" ideas  term/wordsenses as
assembled in dictionaries, glossaries and the like are useful places
to look when trying to develop ontologies?

Yes.  The point that Sue & Adam were making is that those definitions
are abstractions from the way people use language.  Sue devoted her
entire career to writing, editing, and analyzing such definitions.
She definitely considers them useful, and so do I.

But Sue, Adam, Wittgenstein, and I do not believe that you can reliably
assign a predefined "word sense" to each and every word that is found
in any particular document.

On the other hand, those definitions are valuable starting points
for anybody who is going to specify an ontology for a controlled NL.
They are also useful starting points for analyzing a document, either
by a well-informed human or by a suitably designed computer.

The main caveat, however, is that you shouldn't expect any finite
set of predefined definitions to be adequate for specifying all
the "word senses" of any arbitrary text.

RC

It would seem to be a reasonable conclusion that ontologies are
personal systems of what each individual believes exists.

Different people will interpret the same text in different ways.
But the author who wrote or spoke the text did have something in mind
at the moment of writing or speaking it.  Some readers or listeners
will have a better or worse interpretation of what the author meant.

DE

Language is a moving target.

Yes.  That's a good summary.  But it's important to recognize
how language moves, why it moves, and what we can do to derive
an interpretation that is adequate for our purpose.

MC, quoting James Joyce

And perhaps it is madness to grind up words in order to extract
their substance... and to attach them to the feelers of expressions
which grope for definitions of the undefined.

I wouldn't say that it's complete madness.  It's better to say that
those definitions are useful starting points, not absolute truth.

For the final word,

Nora Barnacle, wife of Mr. Joyce

James, why don't you write books people can read?

John

_________________________________________________________________
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

_________________________________________________________________
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
 

_________________________________________________________________
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    (01)