[Top] [All Lists]

Re: [ontolog-forum] on a new list of categories

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: richard murphy <rick@xxxxxxxxxxxxxx>
Date: Sat, 14 Jul 2007 13:02:42 -0400
Message-id: <469901B2.9040308@xxxxxxxxxxxxxx>
Many thanks for your comments John. You'll be pleased to know, a few 
weeks ago in a W3C workshop when Tim BL positioned RDF/OWL as THE WAY, I 
gave him more than an earful. He was very gracious about it. When we 
talked later, he was a bit defensive, but I recognized his sincere 
interest in the discussion.    (01)

more below ...    (02)

John F. Sowa wrote:
> Rick,
>  > I've done enough on the owl representation to stimulate
>  > discussion and hope you, and all members of the ontolog forum,
>  > will have time to comment on this work in progress.
> There is certainly a lot to discuss.  That paper by Peirce,
> which he published in 1868, is his first stab at a problem
> that which he continued to discuss, elaborate, and clarify
> over the next 46 years.
> Peirce certainly made a good stab in his first swing, but
> as I replied to Gary, OWL is a very simple language that
> disguises a lot of underlying assumptions in a verbose
> notation.    (03)

I hope the CL/KIF representation of the manuscript will serve to 
illustrate just that. One of my points to Tim was the importance of 
differentiating the reasoning services associated with OWL ( DL 
reasoners typically provide classification, realization, satisfaction 
and consistency) from other reasoning services (unification, resolution, 
skolemization and GMP). In addition to your point above, it seems to me 
the mismatch between what standard DL reasoning services have to offer 
and some naive expectations about THE WAY, is pretty clear.    (04)

Representing this manuscript in OWL and CL/KIF can serve to better 
illustrate these important issues.    (05)

> The COE diagram that Pat derived shows the bare
> bones more explicitly, and it makes those uninterpreted
> strings stand out more clearly.
> To illustrate the kinds of problems involved, following is
> a note that I sent to the conceptual graph mailing list.
> It illustrates the amount of discussion that can be generated
> by one word, such as 'threaten' or 'to be', or by a seemingly
> simple feature, such as the plural marker '-s'.
> I don't want to shut down discussion, but I just want to point
> out that doing justice to the topic would require an immense
> amount.
>     (06)

Considering the timeless nature of the challenge, the ontolog forum can 
help make our small contribution sustainable by preserving the record of 
the debate, in addition to the work products. Looking forward to your 
leadership over the next few decades ...    (07)

> John
> -------- Original Message --------
> Subject: CogniTextes - Inaugural issue now online
> Date: Sat, 14 Jul 2007 09:02:13 -0400
> From: John F. Sowa <sowa@xxxxxxxxxxx>
> To: cg@xxxxxxxxxxxxxxxxxxxx
> Following is an announcement of a journal on cognitive linguistics,
> the first issue of which is available for a free download:
> http://aflico.asso.univ-lille3.fr/cognitextes/volumes/CogniTextes_2007_1.pdf 
> This issue has three articles.  The first one, which is written
> in English, is about the German word 'drohen' (to threaten),
> its word senses, and how they evolved over time.  It illustrates
> the enormous amount of complexity that underlies just a single
> word, how that complexity evolves, and how it can be expressed
> in another language (such as English).
> The second article, written in French, is about the verb 'to be',
> and the third, also in French, is about issues of singulars and
> plurals of nouns.  It begins with an English quotation that
> summarizes the following 20+ pages:
>    "Number is the most underestimated of the grammatical categories.
>    It is deceptively simple, and is much more interesting and varied
>    than most linguists realise."
> If anyone still needs more evidence about the complexity of translating
> natural languages to logic or to one another, read articles like these.
> Although I have been arguing in favor of using controlled natural
> languages, I have emphasized that they are *formal* languages that
> abstract from the very much more complex NLs.  Translating from
> English to CLCE or ACE requires a reanalysis and reconceptualization
> of the subject that is *identical* to the analysis required for a
> translation to any version of logic or any KR language.  It will be
> a long, long time before such a translation can be fully automated.
> The difficulty of translating poetry from one NL to another has led
> many people to observe that doing a "poetic" translation requires a
> translator who is a poet.  The translation is actually a recreation
> of a new poem in the target language on the same theme as the poem
> in the source language.
> The same is true of the task of translating any natural language
> to any version of logic (including controlled NLs).  The translator
> must have a knowledge of logic that is sufficient to recreate a
> new text in the target language that is on the same theme as the
> original -- even when translating English to controlled English.
> John Sowa
>     (08)

Best wishes,    (09)

Rick    (010)

email:  rick@xxxxxxxxxxxxxx
web:    http://www.rickmurphy.org
blog:   http://spout.rickmurphy.org
cell:   703-201-9129    (011)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (012)

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