Phil,
Ontology is a study and classification of things that exist or may exist. That includes physical things as well as ideas, thoughts, propositions, discoveries, and hypotheses about things that currently exist or may be designed, created, defined, imagined, proposed, suggested, or hypothesized by anyone, anywhere, in any notation, language, or symbol system.
There is only one certain statement that we can assert about any broad coverage ontology, no matter how excellent or comprehensive it may be: it is obsolescent on the day after it has been fully specified.
But it is possible to have complete and perfect ontollogies for systems or policies that must be supported precisely for some particular purpose. An example would be the specification of accounts and services of a bank. The specification is, by definition, correct. If there is any discrepancy, the implementation is incorrect.
ISO has a policy that any standard must be reaffirmed or revised every five years.
John
Thanks very much for the presentations yesterday on the history of ontology summits.
The presentations reminded me of a possible topic for discussion at this year's summit: To create a system that has human-level AI, should the system have multiple ontologies, for different domains? Should it have a meta-ontology, an ontology of ontologies?
Michael and I posted a few remarks about this in ontolog-forum last year and then corresponded briefly about it offline. We did not write a paper though perhaps this would be a good discussion topic at some point during this year's forum. I don't have anything further prepared to say about it, though quite possibly you or others in the forum may.
Best,
Phil Jackson