Re: action item on clarification: first installment

Without disagreeing with anything Frank has reported, my view of much of 
the debate is that it is narrowing down to an issue of RDF semantics:

- does/should/can RDF core convey any useful semantics?
-- (a) If so, they should be formally specified.
-- (b) If not, we should be clear that RDF is simply (abstract) syntax.

This begs a definition of semantics, and since the logicians seems to have 
spent much more energy answering this question, it seems reasonable to 
accept their definition (i.e. model theory).  I'll post a separate message 
"Model theory for programmers" which is my take on what this means, which I 
offered to someone else in an offline discussion about this matter.

If RDF does not have any semantics, then reification as a concept is 
meaningless (at this level).

If RDF does have semantics, then what is reification?  Is it quotation, or 
something else?  (I have formed an opinion that the purposes of reification 
are served by something less logically exotic than quotation [1].)

...

Dan Conolloy has posted a summary of some work by Sergey Melnik that 
attempts to define a model theory for RDF [2].  There has been some 
debate.  The direction looks positive to me, but has not reached a clear 
conclusion.

#g
--

[1] http://lists.w3.org/Archives/Public/www-rdf-logic/2001May/0120.html

[2] http://lists.w3.org/Archives/Public/www-rdf-logic/2001May/0114.html




At 02:12 PM 5/21/01 -0400, Frank Manola wrote:
>from the minutes of the 2001-05-18 teleconference:
> > ACTION ITEM: Frank Manola agrees to summarize what the the logicians
> > (www-rdf-logic) like to see in RDF and provide references to issues
> > regarding clarification of RDF semantics, removing the extraneous
> > parts of RDF / modularizing RDF, etc. by early next week.
>
>What I had in mind when I made the comment about there having been prior
>discussion was (mainly) the following messages (which hopefully can
>serve as entry points into some of the more detailed discussion):
>
>http://lists.w3.org/Archives/Public/www-rdf-logic/2001Apr/0016.html
>where Drew McDermott said:
>
> > As I have argued before, I don't think it would be particularly
> > difficult to fix the problems with RDF.  However, fixing them would
> > require thinking of an RDF expression as a self-contained textual
> > object, which seems to be incompatible with thinking of it as a piece
> > of a graph, and therefore incompatible with the triples model.  I
> > think this is where we're stuck.
>
>http://lists.w3.org/Archives/Public/www-rdf-logic/2001Apr/0047.htm
>where Pat Hayes said:
>
> > The way to fix RDF is to admit that logical content requires the use
> > of some - maybe not many, but some - nontrivial syntactic
> > constructions, in particular the use of nested expressions and
> > quantifier scoping; to abandon the idea that syntax is the same as,
> > or best coded using, reification; and to make a committment to
> > precision in specifying logical meanings (which might include being
> > precise about what one is NOT saying, by the way: precision doesnt
> > entail being exhaustive.)
>
>http://lists.w3.org/Archives/Public/www-rdf-logic/2001Apr/0000.html
>where Peter F. Patel-Schneider said:
>
> > 1/ There are many places in the RDF and RDFS documents where vitally
> > important statements about the semantics of RDF or RDFS are presented with
> > no formal backup, and in a totally inadequate fashion.
>
>snip
>
> > 2/ The triple model does not provide sufficient representational power to
> > transfer interesting semantic content.
>
>http://lists.w3.org/Archives/Public/www-rdf-logic/2001Apr/0024.html
>where Peter F. Patel-Schneider said:
>
> > > 1) How could RDF be changed/augmented/better documented to make it a 
> firmer
> > > base on which to build DAML+OIL?
> >
> > 1/ Provide a firm semantic foundation.
> >
> > 2/ Remove the parts that cannot be given a firm semantic foundation.
> >
> > 3/ Remove the insistence that everything can be represented as
> >    RDF-interpreted triples.
> >
> > 4/ Include facilities for datatypes.
> >
> > 5/ Remove the non-monotonic portions of RDFS (e.g., domain) and the
> >    non-compositional portions of RDFS where possible (e.g., range and
> >    acyclicity).
>
>http://lists.w3.org/Archives/Public/www-rdf-logic/2001Apr/0037.html
>where Peter F. Patel-Schneider described "A Personal View of Changes to
>Make in RDF and RDF Schema to support good representational practices",
>including:
>
> > 1/ Stick to a single purpose
>[involving removing collections and reification]
>
> > 2/ Don't grab all content
>
> > 3/ Make RDF Schema monotonic
>
> > 4/ Semantics
>[provide a good semantic foundation]
>
>Jonathan Borden's overall summary was also pretty fair:
>
> > To broadly summarize: strip out all the stuff from RDF that no one 
> agrees on
> > (such as reification). Start with a simple but rock solid foundation, and
> > add concepts in only as absolutely necessary.
>
>Much of the discussion on the "What do the ontologists want?" thread was
>effectively additional illustration, motivation, and discussion of these
>points (and other related points, such as the need to clarify URIs and
>resource semantics).
>
>Obviously, some additional digestion is necessary, but this ought to be
>a start.  It's also worth noting that in some cases a *choice* is
>necessary, i.e., "either do X or do Y, but don't try to do both" (with,
>however, additional comments about the actual value of RDF if it becomes
>purely X or purely Y).
>
>--Frank
>
>
>--
>Frank Manola                   The MITRE Corporation
>202 Burlington Road, MS A345   Bedford, MA 01730-1420
>mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-8752

------------------------------------------------------------
Graham Klyne                    Baltimore Technologies
Strategic Research              Content Security Group
<Graham.Klyne@Baltimore.com>    <http://www.mimesweeper.com>
                                 <http://www.baltimore.com>
------------------------------------------------------------

Received on Tuesday, 22 May 2001 10:24:29 UTC