Prev: A proof that Geometry is clean but Algebra is tainted by not having finite-number definition #619 Correcting Math
Next: Wouldnt It be Cool if These WereEquivalent?
From: Frederick Williams on 1 Jul 2010 11:42 Charlie-Boo wrote: > > ... I was > going to figure out the exact relationship between that approach and > mine. Go on then. -- I can't go on, I'll go on.
From: Frederick Williams on 1 Jul 2010 12:00 Charlie-Boo wrote: > get. (This is something the authors of ZFC axioms' representations do > not realize when they choose FOL as the language of choice.) That'd be Skolem. > And we can generate an extremely wide array of possibilities of other > very interesting questions regarding formal systems: Is P equivalent > to |-|-P? |-P and ~|-~|-P? See L\"ob 'Solution of a Problem of Leon Henkin' JSL, 1955; or, if you read German, Hilbert & Bernays. Oh, and Jeroslow, 'Redundancies...' JSL, 1973. -- I can't go on, I'll go on.
From: Frederick Williams on 3 Jul 2010 05:30 Charlie-Boo wrote: > > Well, let's see. I call this system ABC because I represent wffs > using letters where > > A = |- > B = ~ > C = (all X) > > Notice that P is actually free. The empty string [] represents P. So > we have e.g. > > A = |-P > B = ~P > C = (allX)P(X) > AA = |- |- P > AB = |- ~P > AC = |- (allX)P(X) > BA = ~|-P > BB = ~~P > BC = ~(allX)P(X) > CA = (aA)|-P(A) > CB = (aA)~P(A) > CC = (allX)(allY)P(X,Y) > etc. Is this alphabet soup of interest to anyone other than you? -- I can't go on, I'll go on.
From: Frederick Williams on 3 Jul 2010 14:08
Charlie-Boo wrote: > > On Jul 3, 5:30 am, Frederick Williams <frederick.willia...(a)tesco.net> > wrote: > > Charlie-Boo wrote: > > > > > Well, let's see. I call this system ABC because I represent wffs > > > using letters where > > > > > A = |- > > > B = ~ > > > C = (all X) > > > > > Notice that P is actually free. The empty string [] represents P. So > > > we have e.g. > > > > > A = |-P > > > B = ~P > > > C = (allX)P(X) > > > AA = |- |- P > > > AB = |- ~P > > > AC = |- (allX)P(X) > > > BA = ~|-P > > > BB = ~~P > > > BC = ~(allX)P(X) > > > CA = (aA)|-P(A) > > > CB = (aA)~P(A) > > > CC = (allX)(allY)P(X,Y) > > > etc. > > > > Is this alphabet soup of interest to anyone other than you? > > Are you saying that you don't understand it? Do you know what I'm > doing? I'm listing the first few wffs. Their representation is any > string of alphabet {A,B,C} so it's real easy to list wffs. Then the > idea is to see how each would be represented using the provability > predicate, to compare the two approaches. > > Does that help? Does your ABC system have any theorems? If so, do they have any proofs? -- I can't go on, I'll go on. |