Re: Sensible and NonsenSQL Aspects of the NoSQL Hoopla

On Wed, 25 Sep 2013 15:05:06 -0700 (PDT) vldm10 <vldm10_at_yahoo.com> wrote:
> Dana utorak, 24. rujna 2013. 01:57:40 UTC+2, korisnik James K. Lowden
> napisao je:
> > However important Frege may be, predicate logic was only something
> > Codd used. I'm unware of any claim he made to extending it.
...
> > Lots of people work in a field for a lifetime without advancing it
> > one iota. For all intents and purposes, Codd invented his. Don't
> > you think that's worthy of recognition?
>
> My bad English sometimes is not appropriate, especially when I'm
> writing in a hurry. Actually I wanted to say the following: Codd just
> applied Frege's theory to database theory.
>
> The other thing I wanted to say is that Codd should quote Frege's
> work, because here come the whole theories as Propositional Logic,
> FOL, Semantics, etc. which are completely changed a spirit of
> mathematics. Frege's mathematics is applied to databases, meaning
> this is very complex application of mathematics. I mean, application
> of Frege's math are much more complex than applications of classic
> math.
>
> In my opinion, people who want to have a good understanding of
> relational databases and who want to study it, they have to know
> where the source of these important theories is. That source is the
> theories which were built by G. Frege. Otherwise, if someone does not
> read G. Frege's theory, I do not believe that this person will
> understand well RM. Every person has the right to the original source.
>
> In this thread I am talking exclusively about the data model. Of
> course, Codd is credited for the raising the theory of databases on
> the scientific level. It is clear that Codd did a lot of other things
> that have significant scientific value, for example FDs and Relation
> Algebra. Note that Codd, is perhaps the only person who was able to
> create the new theory of database and implement it in practice as
> totally dominant data model.

The above is very clear; thank you for taking the time.

Because of your fondness for Frege, and not knowing much about the subject, I spent an hour tonight reading the first hit Google returned for "first order logic history":

        http://www.mcps.umn.edu/philosophy/11_4moore.pdf

From that article, I can only conclude that one can credit many contributors to FOL. Whether or not Frege gets the credit he deserves is a question over which reasonable people can disagree.

I was particularly struck by Hilbert's disagreement with Frege over the truth of axioms,

        "... if the arbitrarily given axioms do not contradict each other with all their consequences, then they are true and the things defined by the axioms exist. For me this is the criterion of truth and existence."

because it reflects how databases are designed in practice!

Just as Hilbert justifies his axioms on their noncontradictory consequences, so are database designs iteratively modified until inferences drawn from them are deemed an accurate reflection of the reality they're intended to model. In effect, the database designer determines his axioms -- postulates about reality -- by examining their consequences.

Of course, everyone reading c.d.t. has seen databases that weren't designed that way, or even well, or even "designed" at all other than as a dog's breakfast. But if you're trying to do a good job putting your knowledge into practice, Hilbert is wind at your back.

--jkl Received on Thu Sep 26 2013 - 03:35:24 CEST