Re: Surrogate primary key plus unique constraint vs. natural primary key: data integrity?

On 13 mar, 00:42, "James K. Lowden" <jklow..._at_speakeasy.net> wrote:
> On Tue, 12 Mar 2013 11:24:19 -0700 (PDT)
>
> Cimode <cim..._at_hotmail.com> wrote:
> > > It is a consequence of defining the enterprise of interest.  A
> > > value assigned outside the enterprise of interest that is a key
> > > within it is "natural".
>
> > An *enterprise of interest* does not say anything about the fact that
> > a surrogate key may at some future point in time be considered a
> > natural key.  But the point is that subjectivity can not be taken
> > from the equation in any scheme involving establishing a unique
> > identifier.
>
> I suppose you're on solid ground there, in a way.  The determination of
> what is an entity, of what will be modelled in the database, is surely
> subjective.  We might even say "establishing a unique identifier" is,
> at best, heuristic.  It's unique until it's not.  I remember when
> currency mapped 1:1 to country.
That is not what I said.

The point I am trying to express is that the process of making a natural key/primary key is inherently subjective even though a unique identifier has to respond to precise criteria. I am not talking about entities in general.

> Roy's point is that the surrogate/natural nomenclature derives from
> (subjective) perspective.  Do you disagree with that, or just think
> the matter too trifling to discuss?
I have nothing to agree or disagree with.

The precise point of my comment is that any unique identifier is *fundamentally* artificial prior to a designer's effort which may never occur. It only becomes natural once a designer's effort at formalizing a unique identifier imposes a reasonable (but subjective) choice of attribute subset to constitute a candidate key, then a natural key. As far I am concerned, the process of *natural key* generation in itself is way more interesting than the *surrogate vs natural key* debate, which is more an implementation issue than a theoretical issue in RM.

But that is just me. Received on Wed Mar 13 2013 - 02:37:38 CET