One term of acknowledged importance in Object Role Modeling is that of Universe of Discourse (UofD), also know as domain of discourse, or more simply, domain. However, within the community of Information Systems designers, architects and practitioners (and perhaps within the subset of fact-based modelers as well), several interpretations of the term are possible. The purpose of this thread is to discuss those interpretations.
As a preface to the discussion, there is the reference to the term as a type; and there are references to instances of that type. In other words, the term Universe of Discourse is the model for a limitless number of subject or domain specific Universes of Discourse, each such universe the concern of one domain. Unless someone suggests a better way to designate the distinction, we can use "TUD" for "Term Universe of Discourse" and "a Universe of Discourse" ("a UofD") to designate any instance of that type, and "the Universe of discourse" ("the UofD") to indicate a specific instance.
- TUD Term Universe of Discourse (type)
- a UofD (any instance of the type)
- the UofD (specific instance of the type)
How should TUD be defined? Several definitions are available and in use; which one best serves the fact-based modeling community? A key component of any definition concerns what types of things are considered part of a UofD. Certainly, a UofD will contain words, phrases, terms and other designators - as it is a universe of discourse. But should it also contain things that go beyond what might be called "the molecular level of discussion" to include concepts, relationships, facts and rules?
Given the important consequences of the choice of definition, there are likely well established schools of thought on the subject - and so "accepted" definitions for each school. For this discussion, let's not take recourse to authority, and pronounce one interpretation as correct, but rather discuss the implications of the various definitions, and see how they effect the way we do fact-based modeling. If we find that there are several different paths that lead to the same place, even that would be remarkable.
For a point of departure, the definition for TUD that I've been using is: the totality of objects, terms, concepts and facts, the identification, definition, veracity and application of which is assumed within the subject domain. That's actually a definition I cobbled together, using parts of other definitions I felt appropriate. How does that compare to your working definition, and what consequences do you expect from any difference?