Re: So let me get this right: (Was: NFNF vs 1NF ...)

"Jan Hidders" <jan.hidders_at_REMOVETHIS.pandora.be> wrote in message news:65qPd.9956$ve3.633922_at_phobos.telenet-ops.be...
> paul c wrote:
>>

Hello Jan,

>> i'm sure your motives are pure. that's not my question. i'd just like
>> to know: "what is the theoretical problem with RVA's?"
>
> There are no theoretical problems with RVA's. Usually they are typed,

I'm not sure what you mean by typed. Are you referring to domains and a declaration of them? Or is this a new activity in typing RVA variables by associating pre-declared domains and higher order relation variables in some manner?

> which prevent's Russel's paradox, but even if you don't like that, then
> you can prevent it by restricting yourself to non-recursive values, and
> even if that is too strict for you you can use non-well-founded sets and
> still not have any problems with paradoxes.

Given the gap between the internal and external predicate in a system, I don't know how one could guarantee this, even theoretically. It would seem reasonable to allow a set of redundant recursive domains within a nested structure where the FOL predicate does not result in a paradox. Can a model discern potential paradoxes without understanding the semantics of the set of predicates?

>
> The real question is if there are *practical* problems with RVA's. That
> question is much more important and also much harder to answer.
>
I agree!

> -- Jan Hidders
Received on Sat Feb 12 2005 - 23:02:29 CET