Thursday, March 16, 2006
The Second School of Semantic Science
ontologyMapping Glass Bead Games
Communication from Paul S Prueitt
Colleagues
Comments from Azamat on
ontology à [30]
Over the past two months I have been working on OASIS specifications, more recently on the Business Centric Methodology. The BCM “mind map” shows that cultural transformation is at least 25% of the purpose of BCM. As depicted by the mind map, the four “quarters” of the methodology are:
1) Cultural Transformation
2) Emphasis Enterprise Agility
3) Layered Focus Methodology
4) Business Centric Methodology Objectives
Quote on page 30 of the OASIS BCM specification:
“Experience indicates that today’s organizations are too complex to be
modeled and easily understood with lines and boxes in a CASE tool. Current modeling techniques are adequate
for showing subclassing, path options, sets of codelists, or object-role
variances; but they fall short in tracing the thread of user choices. This is where the BCM differs significantly
from the current methodologies as it directly embraces and provides support for
choice.”
The work that my group is doing is in support of the Choice Points that are a non-removable part of the methodology.
In our work we extend from the BCM using a theory that was derived from a review of natural science literatures, and which is summarized as being of the “Second School”. This phrase “second school” allows us to make a contrast to a viewpoint often expressed in artificial intelligence and in knowledge engineering. Since this viewpoint is widely reflected in our current culture, particularly in commercial circles and in the media, we allow the use of the phrase “second school” to reflect a need for cultural transformation of the type anticipated by the business centric methodology specification from OASIS (2004).
We also use the phrase “Human-centric Information Production”, or “HIP”, to indicate that the viewpoint of these methodologies and technologies are NOT from the viewpoint of the First School. The First School asserts strong assumptions about the purpose of information science and even the purpose of life. In some cases, the second school agrees that there are some utility derived from what we consider to be a reductionist mind set. However, the second school asserts that almost in the entirety that the first school is incorrect in foundational assertions about the purpose of information, and the purpose of life.
The Choice Points defined with the OASIS BCM specification has no counterpart in the W3C standards. In the W3C standards, the focus is computer scientist or knowledge engineer centric. In all honesty the W3C should create a standard called KECM (Knowledge Engineer Centric Methodology) where the methodology is all about assuming power over all assertions into the hands of a professional elite. This is an unfortunate state of affairs for everyone, actually; since those individuals who have made a life time commitment to the profession are driven continuously into the professional centric viewpoint.
We have learned that one should not critic a powerful profession, such as the IT and KE professions, without being able to offer clarity and methodology that addresses the complete spectrum of issues.
We feel that the OASIS specifications have been moving in this direction over the past few years. These specifications include the OASIS SOA-IM, SOA-CS, FERA, and BCM. Additional work in line with the second school is the TopQuadrant work on “OWL Modeling of Service-Oriented Architecture Reliability, Privacy and Security. We also applaud the W3C work as seen in the discussion by Hao He at the Thompson Corporation
http://www.xml.com/lpt/a/ws/2003/09/30/soa.html
At core to part of this W3C contribution are the specification on URI/URL based representation (a key part of the W3C paradigm)
http://www.w3.org/2001/tag/webarch/#retrieve-representation
To open up the issues differentiating the first and second school approaches, we would like to ask the reader to review the short presentation by Hao He (link above). This is 6 pages long.
Up to page 3 there are only two uses of the term “semantic”, and in both cases we find a point of view difference which is essential to understanding how the first school orients the “truth” about the nature of web services. In both cases, the term “semantic” can be replaced with the phrase “information structure” without loss of coherence. What is different is the implicit information (assertion) that is made by the mainstream and bulk of the W3C efforts. There is an assertion that meaning is captured by the information structure, and additionally that this meaning is complete in some sense. However, almost all of the initial point of view by Hao He is to suggest that a loose coupling between service information structures is necessary in order to adapt to “the duties that IT is supposed to perform”. The OASIS BCM explicitly defines a “pragmatic” dimension to service interactions, which must be mediated by human choice points. There is no W3C standard that becomes human centric in this fashion, and thus the primary difference between the first and the second school is illustrated.
The first school discipline is a very powerful memetic, in that no matter how one starts out one turns around and the effort becomes first school. Hao He’s presentation is an further illustration of this phenomenon. The presentation stats out as if form the second school perspective, then then after the “misuse” of the term “semantic” (seen from the second school point of view) Hao He states (page 3)
“The more restricted a message is, the easier it is to understand the
message, although it <the ease of understanding> comes at the expense of reduced
extensibility. “
(Italics are mine of clarification.) The first thing to mention here is that there is a strong assertion that is not in general a true assertion. For humans, messages are NOT understood or understood based on simplicity, but rather based on a communicative act where the listener’s knowledge of the world is active.
“I am alive” is NOT a simple message. “This is red” is (may be) a far simpler message, but is still a message that might be considered quite complex (in the Rosen sense of complex), if one is talking about a context in which a physical structure seen from satellite is about to ignite and cause a forest fire. The “simplicity of a statement” is depending on purpose of the transaction. Communicative acts between humans are a different type of “transaction” than are the exchange of information structure and/or data between computers.
The strong W3C polemic comes “fully into view” with the definition of “extensibility”. This makes the switch fully from any suggestion of being human (or business) centric to being technology centric.
The W3C definition of extensibility is at:
http://www.w3.org/TR/webarch/#extensibility
“Requirements change over time. Successful
technologies are adopted and adapted by new users. Designers can facilitate the
transition process by making careful choices about extensibility during the
design of a language or protocol specification.
In making these choices, the designers must
weigh the trade-offs between extensibility, simplicity, and variability. A
language without extensibility mechanisms may be simpler and less variable,
improving initial interoperability. However, it's likely that changes to that
language will be more difficult, possibly more complex and more variable, than
if the initial design had provided such mechanisms. This may decrease interoperability
over the long term. “
“Who” are the designers?
The OASIS Business Centric Methodology does not depend on this notion of extensibility. The second school (of semantic science) would state clearly that this notion “for extensibility” is a false notion since the asserted purpose (interoperability in a real world) is not meet.
What is meet is an ideal conceptualization that is the core of the W3C viewpoint, and which is repeatedly asserted as if true.
The actual requirement, according to the second school viewpoint, is for a methodology that offers to the business person (or human stakeholder) an ability to choose between a specific set of information structures (called blueprints), each of which have been developed through a methodology where interoperability is contextualized based on variations of purpose.
BCM Choice Points anticipate that humans will talk about purpose and create a contract based on blueprints. These blueprints are composed through a process that assures interoperability between data structures. This assurance depends strongly on an alternative to description logics. This alternative uses formative category theory as well as description logics.
Formative category theory is not yet expressed as a technical specification, but we are working to define this specification.
Using the extension of BCM based on Formative Category Theory (FCT), there is no need for anyone to assert that meaning is “fully” encapsulated by the information structure. Assertions of this type come under the control of the human domain experts (business experts).