W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > March 2003

Re: Two questions about bagid

From: pat hayes <phayes@ai.uwf.edu>
Date: Wed, 5 Mar 2003 10:05:46 -0500
Message-Id: <p05111b0cba8bba68e5bc@[64.134.139.17]>
To: Tim Berners-Lee <timbl@w3.org>
Cc: Frank Manola <fmanola@mitre.org>, w3c-rdfcore-wg@w3.org

>(with apologies, a message delayed for a few days, while my laptop 
>was repaired)
>
>
>>I don't necessarily mean to suggest that we go back through all 
>>this again, but the pertinent threads where we discussed this were 
>>places like the following, in case anyone wants to go through the 
>>archives (including de re/de dicto, Superman, and the whole nine 
>>yards):
>>
>>http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Jan/0157.html
>
>A thread which argues philosophically about the identity of URIs but does not
>discuss allowed entailments.
>
>There was a lack of test case.   One test case is whether (writing = 
>for daml:equivelentTo)
>the following is a valid entailment:
>
>:Superman =  :ClarkeKent.
>Lois :believes :s .
>:s  rdf:subject :Superman .
>:s rdf:predicate rdf:type.
>:s rdf:object  :Strongman.
>_________________________
>Lois :believes :s .
>:s  rdf:subject : ClarkeKent .
>:s rdf:predicate rdf:type.
>:s rdf:object  :Strongman.
>
>This follows from a current axiom of = that
>
>{ ?x ?p ?y.  ?y = ?z } => { ?x ?p ?z }.
>
>This settles the question as to whether quoting would matter or not.
>Clearly quoting solves the problem.

It stops the bad inference, indeed. But it stops it utterly. For 
example, consider what Lois eventually finds out, when she discovers 
that ClarkeKent really *is* Superman. Now you want to be able to do 
equality reasoning on this part of her beliefs, but howre you gonna 
do it? You don't want to say that Lois believes that

":Superman" = ":ClarkeKent"

because she hasn't lost her ability to distinguish different 
character strings. She has just come to realize that two different 
strings denote the same thing, a fact that is best described with an 
equation which *uses* the strings:

:Superman = :ClarkeKent

The difference being that although this was always true, now Lois 
actually believes it.

What you really want to be able to do is to describe Lois' beliefs, 
to see the world from her point of view, without getting that muddled 
up with the way the world actually is. This requires some kind of 
scoping mechanism, such as a modality, to separate beliefs from 
facts. Then there is a clear difference between saying CK = Sup  and 
saying that Lois believes( CK = Sup) , and the general point of view 
is that

If you use quotation to express beliefs, then you do indeed have an 
insurance against inappropriate de re inferences, but you are unable 
to express actual de re beliefs properly. So quotation is a very 
blunt tool for handling propositional attitudes.

Pat


-- 
---------------------------------------------------------------------
IHMC					(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola              			(850)202 4440   fax
FL 32501           				(850)291 0667    cell
phayes@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam
Received on Wednesday, 5 March 2003 10:06:04 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:56:11 EDT