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

Anthony W. Youngman wrote:
> So if you use Newtonian Mechanics to prove where Mercury was 400 years
> ago, your proof is more accurate than Tycho Brahe's observations - which
> place it somewhere else?

The proof will still be 100% accurate.
Newtonian Dynamics assumes certain axioms, which we now know to be slightly wrong. The first-order logic is still perfectly accurate; it's just your starting assumptions have changed.

> You are making exactly the mistake that made me start this thread - you
> are assuming that the DBMS *defines* reality, rather than carrying out
> experiments to show that the DBMS accurately *describes* reality.
>
> What you should have said is "IF the dbms is an accurate model of real
> life then ...". Which is basically what I said - if the dbms and real
> life disagree then the dbms model must be wrong. You seem to be saying
> that it's reality that's wrong ...

I'm just talking about the system of logic that enables us to talk about our database (our "theory" if you like). Whether our theory has axioms that correspond to the real world, or whether our interpretation (or "model") of our theory is accurate, is a totally different question.

> The problem I have is that the mathematicians seem to have taken C&D's
> idea of "data" and built this wonderful theory on top of it.
> Unfortunately, what they have not done is to define "data" in real-world
> terms (rather than mathematical), and as such there is no way we can go
> from a "proof within the model" to a formal description of the reality
> that that proof represents. So you can come up with all the proofs you
> like within the dbms, but you cannot show that the equivalent real-life
> scenario is true because you cannot describe that scenario accurately.

What I'm saying isn't really relying on DBMSs at all, it's just pure logic. A DBMS is just an example of a system that uses it. We have several layers:

1. First-order logic itself (our meta-language)
2. Our theory (all the relations and tuples in the database, our axioms)
3. Our model (how we interpret our theory in the real world)

All I'm saying is that we know that part 1 is guaranteed to be complete and consistent. Parts 2 & 3 can be totally wrong, which is when your database will give answers that diverge from reality.

> So by definition the theory is unscientific because you cannot show that
> the dbms proof is true (or false) in real life.

Given that your axioms and your interpretation are correct, then I think you can show the DBMS proof is true in real life (for the reasons given above and in previous posts).

I know that the language used by logicians can seem very inpenetrable but I think it does actually make sense; it's not just a conspiracy of people talking gibberish and pretending to understand each other.

I don't know how much you've read about logic but it is very mathematical and well worth the steep learning curve. Wikipedia is a good place to start. Be warned though: logicians to have a tendency to go insane in later life; it is a serious brainfuck if you think about it too much!

Paul. Received on Thu May 27 2004 - 12:16:51 CEST