On Jun 15, 2014, at 12:22 PM, John F Sowa <sowa@xxxxxxxxxxx> wrote: (01)
> Mark and Dick,
>
> MHL
>> I WISH that the SBVR designers had chosen to define it in terms
>> of a system with a sound semantic basis, such as CL/IKL.
>
> The failure to start with a sound semantic basis has plagued computer
> systems for years. I blame both the programmers and the logicians.
> Their excuses are stated in many different ways, but the reason is
> always the same: they don't understand each other's point of view.
>
> MHL
>> Ed and I and the Date-Time Vocabulary (DTV) team chose to define
>> key aspects of DTV in CL + IKL, and in UML + OCL, partly to ensure
>> a firm semantic foundation for our work.
>
> That's good. But there have been many good examples like that over
> the years, and the next group of designers never learn the lesson.
>
> Ted Codd started with logic as a foundation for RDBs. The SQL
> WHERE-clause can be used to represent FOL, but in a way that I used
> to call the worst notation for logic ever invented. Datalog is an
> excellent query and constraint language for any DB (relational or
> graph based). But there are many turf battles over doing anything
> to simplify and standardize DBs.
>
> For the Semantic Web, Tim B-L proposed a foundation called SWeLL
> (Semantic Web Logic Language) in 2000. Pat Hayes and Guha developed
> a logic called SWEL for specifying RDF -- with a semantics that was
> almost identical to CL. But the OWL gang wanted their own model
> theory, and most of the voters in the W3C were clueless about any
> issue that was related to logic.
>
> Fundamental problem: Most programmers don't understand how simple logic
> really is, and most logicians don't know how to explain logic to anyone
> who doesn't already speak their language. For a simpler intro, see
>
> http://www.jfsowa.com/talks/egintro.pdf
>
> MHL
>> What I meant is that the "that" operator (in whatever semantic system /
>> language) is necessary for adequately capturing important aspects of real
>> business rules and vocabularies.
>
> I agree. Note slide 4 of egintro.pdf, which shows Peirce's version. (02)
Which, as usual with Peirce, is simple and elegant :-) (03)
But its not at all obvious what a proposition *is*. For sure, it can't be
simply a sentence or even an equivalence class of sentences. It can't be
anything that is defined purely in terms of syntax, because propositions,
unlike sentences, are typically *about* something. Consider for example, the
proposition that I, Pat Hayes, am an American citizen. You can write a sentence (04)
(American PHayes) (05)
but what ensures that the name "PHayes" refers to me? A: nothing ensures it,
and that name might refer to anything as far as the semantics of the formal
sentence is concerned. The nearest thing to a proposition here has to be (an
equivalence class of) sentence(s) PLUS a mapping from some of the names in
those sentences to actual entities, ie a partial interpretation mapping. This
essentially *semantic* component is essential to the very nature of
propositions: no construction based purely on operations on syntax can do the
required job. See the diagrams in
http://www.ihmc.us/users/phayes/IKL/SPEC/SPEC.html for one way to do this
formally. (06)
>
> RHM
>> John Sowa's earlier remarks about IKL and "that" identified the
>> fundamental issue underlying this entire thread. The details relating
>> to graphs and triples are not important. What's important is the
>> treatment of propositions and sentences (using John's terms). I will
>> simply use the term proposition. (07)
Please, I beg you, on my knees, DONT get sentences confused with propositions.
It took us (literally) years to get this straight in IKL, and confusing these
has made (again, literally) years of work on context logics worthless, because
they claimed to be about propositions but were in fact about sentences.
Propositions and sentence are not the same KIND of thing. A sentence can
*express* a proposition, but it is not that proposition and it does not denote
it. (So for example, in this thread it is sometimes said that RDF reification
names a triple and also names a proposition, as though these were synonyms: but
a triple is a sentence, not a proposition.) (08)
>
> I'm glad that we can agree on something. But the same proposition
> can be expressed by different sentences. Note slides 8 to 11 of
> egintro.pdf. When you're trying to relate multiple notations, it's
> essential to distinguish the readable mark (sentence) from the
> meaning of the mark (proposition). (09)
True, but... (010)
>
> In the following note, I define a proposition as an equivalence class
> of sentences (011)
...that's not enough by itself. You also need to explain how a proposition can
be about someTHING. See above. (012)
Pat (013)
> that can be mapped from one to the other by a meaning-
> preserving translation (MPT): http://www.jfsowa.com/logic/proposit.htm
>
> 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
>
> (014)
------------------------------------------------------------
IHMC (850)434 8903 home
40 South Alcaniz St. (850)202 4416 office
Pensacola (850)202 4440 fax
FL 32502 (850)291 0667 mobile (preferred)
phayes@xxxxxxx http://www.ihmc.us/users/phayes (015)
_________________________________________________________________
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 (016)
|