Le 24-juin-05, à 20:27, [EMAIL PROTECTED] a écrit :

<x-tad-bigger>Bruno, I have to be honest and say that I'm just starting to get into this stuff out of a passing interesting and that I probably don't have time and priority to study the math that would be sufficient to make a significant contribution in my view.

</x-tad-bigger>

To be sure I was not asking a contribution! But you did point on something interesting.

<x-tad-bigger>For instance, I just learned about Church's lamba calculus last night.

</x-tad-bigger>

It is probably better than learning about Church's lambda calculus tomorrow.

<x-tad-bigger>So I probably went in over my head in citing Lowenheim-Skolem. But is not my statement correct with regard to Lowenheim-Skolem and cardinalities? If so, then perhaps the iffy part is the application to this topic (so perhaps I committed the 1004 fallacy here). Nevertheless, regarding the application, on the surface it just seems that to make any conclusions about whether there is a non-zero probability of something being true or happening, you need to know the cardinalities of the sets you are working with.

</x-tad-bigger>

Actually, not really. You need a measurable space. It is a set with a sigma algebra of subsets. Cantor found uncountable sets (high cardinality) with measure zero. It is very tricky, especially with comp. But with modal logic I have been able to isolate the measure one logic, without investing to much in measure theory.

<x-tad-bigger>I will be gone on a marriage retreat this weekend, so I'll be back on Monday.

</x-tad-bigger>

A marriage retreat! This is what I should suggest to my friend Jack! Thanks, ;-)

Bruno

http://iridia.ulb.ac.be/~marchal/