ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Fwd: Re: Using controlled natural languages for onto

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Adrian Walker <adriandwalker@xxxxxxxxx>
Date: Wed, 23 Mar 2011 12:17:49 -0400
Message-id: <AANLkTinM+irfc0DGe=NRaiSKkt3_tbBcRL5AiJ_ZH6Qi@xxxxxxxxxxxxxx>
Hi Leo, John & All --

Leo wrote:

> The power of formalization is that it does make things more precise,
> which lends itself to explicitly developing hypotheses/theories for
> refutation, convergence on disputed terminology, and amenability
> to computational implementation.

Unfortunately, the devil is in the details of "precise".

Consider the statement      

                      p("J","B",N) & N>5        [1]

It appears to be more precise than

                 John has written many fine books    [2]

However, different humans looking at [1] could think it says that Jane has more than 5 boys, and so on.

There's also the problem of non-unique models for FOL.

                             -- Adrian

Internet Business Logic
A Wiki and SOA Endpoint for Executable Open Vocabulary English Q/A over SQL and RDF
Online at www.reengineeringllc.com   
Shared use is free, and there are no advertisements

Adrian Walker
Reengineering
On Tue, Mar 22, 2011 at 10:55 PM, John F. Sowa <sowa@xxxxxxxxxxx> wrote:
Leo,

> The power of formalization is that it does make things more precise,
> which lends itself to explicitly developing hypotheses/theories for
> refutation, convergence on disputed terminology, and amenability
> to computational implementation.

On that point, we are in total agreement.  I have been preaching
that in my logic courses and in my research for over 40 years.

I was not attacking the use of logic in NLP or the use of model theory
as the semantic foundation for logic.  What I was criticizing is the
claim that Montague's particular version was useful.  In fact, I would
say that the only things in it that were novel weren't very useful.

> I think that the model-theoretic semantics of NL largely ignited by
> Montague's work has greatly advanced the science of the semantics of NL,
> and thereby made semantics eminently implementable for practical NLP.

It's a tragedy that Montague died before he could develop his system
further, but the person who did more than anyone else to promote it
was Barbara Partee.  But note that even Hans Kamp abandoned the novel
aspects of Montague's system.  There were some people who pursued
intensional logics along Montagovian lines, but none of their work
has been useful.  I believe it's a dead end, but I'd be happy to be
proved wrong.

In my previous note, I also said that several people had *implemented*
translators from subsets of English to FOL *before* Montague published
his version.  Among them were Bill Woods, Terry Winograd, and some
of my colleagues at IBM.  Their work was independent of Montague,
it was just as formal, and it had the advantage of being useful.

As I also pointed out, William of Ockham developed a model-theoretic
foundation for Latin (circa 1323) that was not a formal as Tarski's,
but it had broader coverage.  Any graduate student who knew FOL,
who studied Ockham's theory of propositions, and who knew some
grammar for any NL could write a model theory for that NL.

Please look at the following book ($35 at Amazon):

Ockham, William of.  _Ockham's Theory of Propositions_, translation
of Part II of Ockham's _Summa Logicae_ by A. J. Freddoso &
H. Schuurman, University of Notre Dame Press, Notre Dame, IN, 1980.

Peirce had not only read Ockham's _Summa Logicae_, he lectured on
it at Harvard, he translated parts of it for his lectures, and it
was probably the inspiration for his model theory for existential
graphs.

> Do I think that the model-theoretic semantics of NL as currently
> expressed represents all of the semantics-pragmatics of NL?
> No, obviously not.

On that point, we agree.  And so does Barbara Partee.  She admitted
that lexical semantics has a lot to say that hasn't been adequately
incorporated into the formal work.  In fact, she divorced Emmon Bach
and married a Russian lexical semanticist.

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
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (01)

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