Re: Using the RM for ADTs

Brian Selzer wrote:

>It's not a question of philosophy. Identity and identification are
>completely different concepts. In particular, identity involves all
>properties while identification involves just those properties that serve to
>distinguish one object from another. There can be many sets of properties
>that identify something at any given time, but only something that has all
>the same properties at all times can be considered identical.

It's only completely different if you postulate an absolute object universe in which every object absolutely exists. You have not demonstrate the need for this postulate. Just considering identity and identifiability to be the same thing, namely, a property of terms used in a given a conceptual model, makes a lot more sense to me. The difference you refer to just the difference between identity in two conceptual models to me. Except that according to you, one is The Absolute Object Model, which will always remain elusive (we'll never agree on all the details).

Of course our conceptual models do have a correspondence with reality, but I don't think it's best expressed in terms of Absolute Real Objects.

>> The closest I get to ontology is entity-relationship modeling.
>
>What does ontology to do with identity and identification? Or taking things
>out of context?

E/R expresses identity ('identifiability' if you prefer) explicitly.

>[...] I have read numerous academic papers that at least in part support the
>positions I take, and I have in the past cited some of them here, so I don't
>consider myself out of step.

I think you're correct. E.g. the Tractatus Logico-Philosophicus (which I haven't read, but I browsed it for this discussion) seems to agree with you both in viewpoint and in terminology. It particular it uses the term 'object' in the same way, as far as I can see.

-- 
Reinier