Generative Methodology thread

Tuesday, May 24, 2005

The next generation knowledge tools

Paul, John, & Friends:

I have been suggesting for sometime that Paul seek to organize his community with respect to some set of foundational "defining questions" -- with champions of different points of view.

The aim in doing so is to identify those issues for which there are rough agreements and those for which particular sub-groups choose to draw strong lines of difference.

With Paul, John, and I there are at least three active voices who are not shy in voicing their own interests and stands on many issues and do so with well principled statements and active on-going efforts to test the consequences of where these lead.

I ventured to volunteer one read on where John seemed to stand, not fearing that he could indeed speak his own mind with far greater clarity. This is precisely what he has done on several issues -- and so the historical process of forming a school -- or not -- is begun.

I trust that Paul will continue his chosen task of organizing, recording, and sharpening the issues in welding this effort into future meetings and publications to mark this Knowledge Age emergence.

ARTICULATING MY CLASSROOM WITHIN OUR SCHOOL

My work is concentrated in Knowledge Science, Knowledge Industry Formation, and in accelerating the advancement of Knowledge Research with world-leading projects and productive professional career paths.

Knowledge Science as I have formulated and taught it is empirical and quantitative, based upon a Physical Theory of Knowledge and Computation and long series of applications to small and massive projects of national importance. My recent notes to Paul set forth many consequences of nurturing that work through a complete Science and Technology (S&T) Cycle (10 years of R&D & 10 years of Advanced Engineering and exhaustive testing in 50+ projects). The remaining 10 years of that cycle now demand technology deployment and use world-wide -- to be triggered this year and next by Mark 3 Beta releases. The Alpha construction and testing is what we are doing right now. Hence my special interest in forward-going strategic partnerships for industry formation & professional education.

John's path of leadership has been just as long or longer and more public than any other. The missing pioneer of those who set sail on their own in the late 70s and early 80s is Doug Lenat. We have all keynoted Knowledge Technology Conferences, but Doug has kept his Cyc in a circle all his own. His appearance in the recent conference in San Francisco led by Mills Davis & Ralph Hodgson may bode for seeing him more, but as John notes Doug's presentations have turned defensive and far more humbled than we have ever seen before. The quote that "human knowledge" is far from self-consistent was taken from Doug's closing remarks on his 20 year findings -- and I agree with him. Theories rule human interactions and few of these are scientific, they are nonetheless influential enough, to alter human behaviors -- whether truly believed or merely accepted as conventional or "formal law of the land".

The hurdle to be jumped within our school is the notion that only scientific theories deserve the name "theory".

My "theory" definition of “theory” is any "idea" or "value" that can substantially change a human's behavior.

The notion of unbounded meta-meta-meta-meta parsing has been an endless value biased ontology effort and Vienna Circle argument to divide classical metaphysics into value differentials safely putting mathematics, religion, fantasy, and occult into different standards for measuring "justified acceptance." Human's fight wars on just such issues -- science or not.

Paul's footnotes speak of religious issues, John observably refuses to use the classical term "metaphysics" in favor of positivist use of "abstract", many Western scientists follow suit. Now Asian Scientists weigh in using "metaphysics" without blinking their eyes. Our "computer science" is stymied still by battles 3000 years old. I think our agreements can be linked in finding and declaring fundamental definitions or articulating clearly different positions on such definitions. One-third of my course is centered there.

ON TRUTH-CONDITIONAL LOGIC AS A MISTAKE

My own choice now is to use Brian Loar's (Rutgers University) review of the evolving definition of "meaning" pp 545-550 in The Cambridge Dictionary of Philsophy, 2nd Edition. I position myself amongst the last two "conceptual or use theories of meaning" and accept far more of Quine, than John finds acceptable. I heartily recommend that dictionary for those not inclined to read every original philosopher to position their own notions quickly.

It was a quote about Dumment in Loar's definition of "meaning" that spoke of positivist attitudes.

Specifically: "Dumment has argued that assertability conditions are the basis of meaning and that truth-conditional semantics rest on a mistake (and hence also propositional semantics in sense [a] above)."

QUINE & BALLARD AS RADICALS

Loar also has this quote for Quine: "A radical position, associated with Quine, identifies the meaning of a theory as a whole with its "empirical meaning", that is, the set of actual and possible sensory or perceptual situations that would count as verifying the theory as a whole."

In all Mark X variants from 1978 on, the central atoms of any knowledge representation consisted in nominal or conceptual model-instance sets -- the set of "real" or "abstract" instances associated with each "conceptual model". Such sets were represented simply by a minimal number or bit pattern denoting each (metaphysical) "model" and another minimal number or bit pattern uniquely identifying each real, imagined, or representational instance. If these sets were arguably complete enough to denote everything known, they exemplify and support Quine's notion of "empirical meaning".

Empirically this representational approach has proved 100% successful in over 200 applications (Marks 0-3, 1978-2005).

One's "ontology" consists in all the concept (models) one needs to proclaim exhastively "what is". This choice is necessarily a personal "ontological commitment." Quine's views are cited most often to describe how we make "ontological commitments." The definition of "ontological commitment" in the Cambridge Dictionary of Philosophy is written by Roger F. Gibson (Washington University). He defines it as "the object or objects common to the ontology fulfilling some (regimented) theory (a term fashioned by Quine). The ontology of a (regimented) theory consists in the objects the theory assumes there to be. In order to show that a theory assumes a given object, or objects in a given class, we must show that the theory would be true only if that object existed, or if the class is not empty."

>From perspective of Knowledge Theory, the acceptance of a theory need not be a declaration of acceptance or belief, it is enough that receipt or awareness of a theory should observably alter one's behavior.

In this we come to the great divide between schools and one defining issue of our time.

Loar's discussion notes that "A central tradition in twentieth-century theory of meaning identifies meaning with factors other than propositions .... and truth-conditions." He notes that Quine's notion of "empirical meaning" -- "can be seen as a successor to the [positivist] verification theory, with THEORY replacing STATEMENT or SENTENCE." I have described this jump as leaping the LINGUISTIC/SEMANTIC GAP.

It is that observation and understanding that has caused others to label my Mark X developments as "Theory-based declarative semantics" in contrast to WC3 OWL & RDF and other representational formalisms that seek to preserve "linguistic relational (sentence) assumptions" while somehow labeling them "semantic" by virtue of using E-R diagrams.

SCIENCE VS RELIGION

The lyceum and academy of Greece within which Aristotle and Plato taught and wrote were not concerned with the "science of human behavior". They were there to teach presumedly exalted modes of reasoning, so their students might approach a perfection beyond the norm. From Aristotle to the blessed acceptance of his teaching within the Catholic Scholasticates of the Reformation, the goal remained to produce individuals who thought as they "should", not as other human's in fact do.

Our computer science and math departments unthinkingly continue that semi-religious objective of teaching the formalism we SHOULD USE, in contrast to the modes of thinking human's observably do use.

Our new knowledge science is founded on real and experimentally testable human modes of thinking and use of legacy "lessons-learned". Creating equally comparable reasoning and success within machines is the test of our current and future understanding and achievements within that purpose.

Dick