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

From: Marshall Spight <mspight_at_dnai.com>
Date: Sat, 22 May 2004 03:25:55 GMT
Message-ID: <7Lzrc.20286$zw.10403_at_attbi_s01>

"Anthony W. Youngman" <wol_at_thewolery.demon.co.uk> wrote in message news:uMz8IRDsXHqAFw+B_at_thewolery.demon.co.uk...

>

> But both are attempts to apply a mathematical model to a real world
> problem. Viewed from a dispassionate oversight, both are instances of
> the SAME problem, and the same techniques can be applied to solving
> them. Namely "how well does my mathematical model work in the real
> world?".

It's clear you love this analogy, but it doesn't work.

What we put in the database is data, not the real world. Neither do we attempt to say anything about the real world with our databases. Consider a payroll database. Does it contain one single fact about the natural world? It does not. It has names, social security numbers, addresses, salaries, phone numbers, etc. These are all 100% human constructs; none of them are found anywhere in the real world; they are exclusively in our heads.

I suppose you will counter with some NASA database or something. But what will it have in it? Let's say it's full of the positions of rocks. But how do we record those positions? With a GPS machine that tells us lattitude and longitude. Note that we don't have *actual* rocks in the database; we only have data for the lat/lon pairs. You could comb all over the surface of Mars or Earth and never find a lattitude line.

The internal predicate is in the database; the external predicate is in our heads. Humans convert from one to the other; machines can't. It's imperative that that the humans be able to tell the difference.

Marshall Received on Sat May 22 2004 - 05:25:55 CEST