Re: The naive test for equality
Date: Fri, 05 Aug 2005 20:05:23 GMT
Message-ID: <7EPIe.1271$Je.157_at_newsread2.news.atl.earthlink.net>
"VC" <boston103_at_hotmail.com> wrote in message
news:-NCdnQb31ZJFRW_fRVn-gg_at_comcast.com...
> If you are talking about the relational model, one would expect to hear
the
> usual first-order terms , like constants, variables, relations etc.
instead
> of the fancy words like "reference" and "referent". What is the advantage
> of using the word "referent" instead of "value" [of some type] or element
of
> some set ? In other words, how a "referent", when used in a first-order
> language, is different form an element of a set ?
I am talking about the representation of "things", and the correspondence between operations (like "compare") on the representations, and corresponding operations on the things themselves. I'm not sure what to call this subject. For lack of a better term, I'll call it the "Theory of Data". Nothing in the theory of data should be contradicted in the RM, but the span of the theory of data is broader than the RDM. ("D" added intentionally).
Hence my use of the word "referent", rather than "value". I don't think "referent" is either a fancy word, or superfluous.
If we can start with the "theory of everything",
Everything is a THING.
Everything inside a data system is a representation.
A representation represents a THING.
Some THINGS are values.
Some THINGS are not values.
>
> Also, talking about comparing ["checking for equality"] language symbols
> ("references") is meaningless -- first order formulas, like A = B, are not
> true or false by themselves, they are so in some interpretation. The
actual
> equality/equivalence relation is a set (if it exists) constructed from
> elements ("referents") of another set.(s).
A = B may be unverifiable (not meaningless)
but
123.45E1 = 12.345E2
is verifiable. Received on Fri Aug 05 2005 - 22:05:23 CEST