Discussion with John Sowa and others

Paul,

I recall that you made various points about abstraction,

but I'm not sure what exactly you're referring to.

> I wonder if you ever see the point about abstraction

> that I make often, and which you never acknowledge.

Could you please resend a copy of a note that covers

that point.

John

***********************

John,

I will develop something specific to communicate to you (in your language, as well as I can) what I mean by the "formation of natural categories" and the "human induction of signs and the meaning that these signs come to have".

In my mind, there are currently three steps involved in creating web ontology.

1) observation by a mind or minds of natural ontology's expression as signs and symbols

2) the formation of natural category in a human mind (as i discuss in various chapters at

http://www.bcngroup.org/area3/pprueitt/book.htm

3) the use, BY KNOWLEDGE ENGINEERS, of specified standards (such as RDF, OWL, XML, UML, SOA-IM, etc) to construct data structure

I wish to create an OASIS standard (over the next year) called "Community Centric Service Orientation” which will build on the recently approved BCM (business centric methodology) standard

www.businesscentricmethodology.com

(which focuses on developing choice points for human decision making)

The purpose of this standard will be to shift the origin of "induction of information structure" from the small and often not interdisciplinary community of knowledge engineers to normal everyday people.

Your call for standardizing common logic (which is simple plain logic that everyone can understand) is in line with this purpose.

I see the following facts

1) there are limitations to the capability of OWL, for which most of the SemWeb community is in denial over.

As you said:

RDF is a terrible base for supporting logic,

OWL demonstrates how bad RDF is, and there was much better

technology available instead of the bunch of crap that was

dumped on us by the W3C.

which is similar to what I have said, and others have said.

2) groups like BioPAX is using OWL Full and extending in ways that are not really "standard" OWL so as to attempt a data integration of many massive bioinformatics databases (many in RDBMSs) . This work and the work, in my opinion, at DERI in Germany and others (www.wsmo.org) are very close (6 months) to using WSDL (web service description language) to create the "best that the TBL layer cake will produce>

The paper you cite at:

http://www.secondschool.net/beads/ontologyMapping/44.htm

"A Revised Architecture for Semantic Web Reasoning" makes points similar to yours and to mine (and others) regarding the limitations of OWL.

However, the proposed Community Centric Service Methodology would fence off the limitations using a strategy that Nigam Shah has recently been discussing on the BioPAX e-forum, and which is very similar to my BCNGroup Roadmap...

This suggestion is (in Nigram's words)

"I just sent an email to Alan about a more "richer" representation involving slots for a variety of "states" and "contexts" that we are using in the HyBrow system www.hybrow.org . I am currently trying to translate this to OWL so that HyBrow can be made a semantic web application. (I cc'ed my email to Alan to the BioPAX group, but will forward it to this group of people as well)

I strongly believe that a Semantic Web version of a HyBrow like system (which allows users to build "hypotheses" about a biological system and then tells them how well is it consistent with existing data items, what are the items that contradict and what are the ones that support the hypothesis) will be a powerful demonstration of the Semantic Web. "

Issues relative to the SemWeb is that a pragmatic axis to human awareness (of information as processed by the mind-brain system) is left out, for reasons that a touched on my your and other's principled criticism of the TBL layer cake.

Pragmatics relates, in my mind and in the writings of others, to what exists in real time...

Semantics would then be “defined” as the existing archypal structure to natural categories being expressed by the "system". In cell signal pathway context the system is a valid producer of natural category, and the cell signalling is a system induction of signs and symbols.