Re: the distinction between data and intelligence

From: erk <>
Date: 10 Jun 2005 04:37:41 -0700
Message-ID: <>

mountain man wrote:
> In Codd's 1979 paper, "Extending the database
> relational model to capture more meaning", Codd
> opens his paper with the following phrased aim:
> "The intent is to capture (in a more or less formal way)
> more of the meaning of the data so that database design
> can become more systematic and the database system
> itself can behave more intelligently".
> Codd was aware of evolution, and was very much
> part of that process.

That paper is a collection of things, some of which may have been prompted by other researchers (the bibliography is extensive). It's not really an evolution, at least in my opinion, as the contents are:

1. a description of a (possible) DBMS catalog
2. some formalization of patterns of DBMS usage
3. an interesting discussion of nulls and 3VL (although I disagree with
his thing justification of the concept, his presentation is much more solid than most)

Your proposal, as I understand it, primarily concerns the following: 1. process/program "objects" as DBMS entities - in other words, how they can be managed in the system catalog 2. change management - this also would have a large impact on the system catalog to allow the DBMS to manage versions and schema evolution

> It is quite conceivable that, by their ability to also store
> processes and programs (such as stored procedures)
> current DBMS software is not just a repository for
> data, but for what may reasonably termed "intelligence"
> in respect of the organisation which has invested time
> and resources in incrementing automation.

The term "intelligence" is so overloaded as to be useless in this discussion, in my opinion. The above are certainly assets.

> Do you agree that the distinction between data and
> intelligence may be drawn informally as simply as I
> have done above (by the union with processes)?

One standard view of a database is storing facts - the raw material on which computations operate. But a relational language as applied to program objects is still manipulating data, albeit a higher order. The DBMS presumably doesn't parse the programs - it doesn't access the structure of programs (e.g. their abstract syntax trees). So a program text is essentially a new domain for the system catalog.

> Otherwise, how would you define the distinction
> between data and intelligence (for the purposes
> of database theory only budding philosophers!)?
> Or, is such a distinction of little consequence?

Your topic is useful; using the word "intelligence" might not be worth the confusion.

  • Eric
Received on Fri Jun 10 2005 - 13:37:41 CEST

Original text of this message