Re: A good book

From: J M Davitt <jdavitt_at_aeneas.net>
Date: Sat, 08 Jul 2006 11:20:17 GMT
Message-ID: <RxMrg.25305$Eh1.5368@tornado.ohiordc.rr.com>

Marshall wrote:
> J M Davitt wrote:
>

>>>>In brief:
>>>>
>>>>Every relation has an associated predicate.
>>>>Every element of every table is a proposition.
>>>>Every relational algebra expression is an expression of
>>>>logical inference, deriving new propositions from existing
>>>>ones.
>>
>>Predicate: good start.  "[e]lement of every table:" wrong turn here;
>>tuples, not elements, and relation variables, not tables.  And not
>>merely propositions, but *all* those that has been quantified as true.
>>And I think deduction is a better description of relational
>>expressions, isn't it?
>>
>>(Marshall, was that you?  I'm surprised...)

>
> Well, "element" or "member" are both acceptable terms for
> the things that make up a set. In the case of sets that are
> relations, the elements are tuples.
>
> I do regret the accidental use of "table" rather than sticking
> with "relation" as I started with. (D'oh!)
>
> I tend to think of "true proposition" as redundant, but
> there doesn't seem to be a term that would describe
> something that is structurally a proposition but is false.
> So maybe it's reasonable to emphasize it that way.
>
> Is there a distinction between deduction and inference?
> I'm not clear.

Deduction and induction are the processes for deriving a specific fact from a collection of facts and vise versa; Inference covers both of them, I think. Plus, in the world of databases, we're seeing the developing notion of "inferential services" so I tend to avoid inference.

> Marshall
Received on Sat Jul 08 2006 - 06:20:17 CDT