Oracle FAQ Your Portal to the Oracle Knowledge Grid

Home -> Community -> Usenet -> comp.databases.theory -> Fraud Number Null:

Fraud Number Null:

From: Bob Badour <>
Date: Wed, 21 Jun 2006 03:12:42 GMT
Message-ID: <KO2mg.234$>

: "I state that BB is wrong in saying RAM SQL tables representations are multidimensional..."

BB replies: The SQL table is multidimensional--or more correctly has a number of dimensions equal to its degree--no matter the medium in which one represents it: in a linear memory, written in chalk on a two dimensional blackboard, suspended in a cube of glass. The dimensions of the medium have absolutely no effect on the dimensions of the logical entity.

Which brings us to the all-important question: How many SQL tables can fit on the head of a pin? Or is that in a pinhead? I forget.

BB states: "Non sequitur does not require a false premise. Sound logical derivations require valid logic and true axioms."

  replies: "Don't express yourself about mathematics when you don't understand a thing about it...You have redefined in 2 lines the meaning of Axiom as it was defined from Antiques times till now... "

BB rebutts: I used premise and axiom synonymously as a statement presumed true. I fail to see how I redefined the meaning of either by doing so.

I contrasted valid logical derivations or conclusions with sound logical derivations or conclusions. The primary context of the terms valid and sound are in logical arguments or propositional logic.

Thus, consider the following logical argument:

  1. If Bob is credible and wrong, he admits he is wrong.
  2. I have proved Bob wrong.
  3. Bob has remained silent.

Therefore, Bob is not credible.

The above argument is a valid logical argument; however, it is not a sound logical argument. If Bob has remained silent, he has not admitted he is wrong. By modus tollens, therefore, the statement that Bob is credible and wrong is false. But we know Bob is wrong therefore the only way the statement "Bob is credible and wrong" can be false is if the statement "Bob is credible" is false. All of that is perfectly valid.

However, the argument is not sound. The argument relies on three premises or axioms. As it happens, all three of the axioms are false.

Premise #1 is false. For instance, if Bob has twit-filtered some self-aggrandizing ignorant or similar crank, Bob sees nothing at all to respond to. Thus, his silence in this case merely reflects the worth Bob ascribes to the person ignored ie. at best no worth at all.

Premise #2 is false. While I am sure the crank genuinely believes he has demonstrated that Bob is wrong, only the crank himself (or perhaps another crank) imagines such a thing.

Premise #3 was true until now.

One could forgive the poor reader at this point for wondering: "What the hell does any of this have to do with databases or their theory?!?"

However, it has everything to do with them.

We use dbmses so that we can draw correct inferences from the data we own. To draw correct inferences, the data must be correct and the manipulation of the data must be correct.

It doesn't matter how valid the manipulations are if the data are wrong. It doesn't matter how accurate the data are if the manipulations are invalid.

This point pops up again and again. For instance, many newbies, cranks and self-aggrandizing ignorants focus on structure to the exclusion of integrity and manipulation. This is a mistake that one can easily forgive newbies for making.

Defects like null and duplication affect the validity of results.

Integrity can only prevent unreasonably wrong data from getting into a database. Whether the data are right depends on whether they accurately reflect what we intend them to reflect from the real world.

See point #3 at

etc. Received on Tue Jun 20 2006 - 22:12:42 CDT

Original text of this message