For many years in the 1980s and 1990s, there was some agreement that a semantic model had three components:

· Declarations

· Procedures

· Constraints

One of the definitive sources for this 3-part structure was the 1982 data modeling book by Tsichritzis and Lochovsky, but other data modeling sources of the times used the same distinctions. There was also Sowa’s notion of “conceptual relativity” which hinted that some kinds of knowledge could be represented in more than one of these ways (like CurrentAge which could be either a declarative or a procedural representation), so the choices were in many cases relative.

My question concerns the efficacy of this three part structure with regard to ontologies. Is there a source defining ontologies that uses this same set of distinctions? Is the structure a viable way to compartmentalize an ontology’s description (subject to conceptual relativity arguments of course)?

William E. McCarthy

Michigan State University

517-432-2913

mccarthy@xxxxxxxxxxx

http://www.msu.edu/user/mccarth4