Re: In an RDBMS, what does "Data" mean?

Anthony W. Youngman wrote:
> If we can't set up an experiment (even a Gedanken thought experiment),
> then relational theory is not provable, therefor it is not scientific,
> therefor it is irrelevant to the real world, therefor why the hell are
> we using it :-)

Newtonian mechanics is more like a particular instance of a database in the relational model, rather than the model itself.

The relational model is really just an implementation of first-order predicate logic that is suitable for computers.

Logic is more like a "meta-theory": it's kind of how we reason *about how we reason*, so it's a bit self-referential.

For a particular database we can test it experimentally: we add data, query it and check that the answers correspond with reality.

For first-order predicate logic itself, it's almost axiomatic that it corresponds to reality, because we are saying this is how we argue logically by definition. Godel proved that first order logic is "complete" in some sense (see here for example: http://www.sm.luth.se/~torkel/eget/godel/completeness.html), though the whole area of Godel is guaranteed to cause confusion and misunderstanding, and will possibly explode your brain.

>> (Incidentally, can we agree that while consistency is not sufficient to
>> prove the correctness of a data model, it is necessary?)
>>

But in order to prove the model is accurate externally we'd have to use logic. So we've got a chicken and egg situation here. What logic is external to logic itself?

Paul. Received on Thu May 20 2004 - 11:18:19 CEST