Sujet : Re: Minimal Logics in the 2020's: A Meteoric Rise --- eternal september failure
De : polcott333 (at) *nospam* gmail.com (olcott)
Groupes : sci.logic comp.theoryDate : 09. Jul 2024, 02:00:37
Autres entêtes
Organisation : A noiseless patient Spider
Message-ID : <v6huj5$12ktu$2@dont-email.me>
References : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
User-Agent : Mozilla Thunderbird
On 7/7/2024 10:09 PM, olcott wrote:
On 7/7/2024 10:02 PM, olcott wrote:
>
Formal logic is a subset of this.
Not-a-logic-sentence(PA,g) ≡ (~True(PA,g) ∧ ~True(PA,~g))
There are no truth preserving operations in PA to g or to ~g
>
https://liarparadox.org/Tarski_275_276.pdf
Within my analytical framework this Tarski sentence is merely
self-contradictory
(3) x ∉ Provable if and only if x ∈ True. // (1) and (2) combined
There are no truth preserving operations in Tarski's
theory to x if and only if There are truth preserving
operations in Tarski's theory to x
There cannot possibly be an infinite proof that proves
that there is no finite proof of Tarski x in Tarski's theory
The infinite proof of the Goldbach conjecture
(if it is true) continues to find more true
cases than it had before, thus makes progress
towards its never ending goal (if its true).
The cycles in the following two cases never make any progress
towards any goal they are merely stuck in infinite loops.
The Prolog unify_with_occurs_check test means that
LP is stuck in an infinite loop that makes no progress
towards resolution. I invented Minimal Type Theory to
see this, then I noticed that Prolog does the same thing.
?- LP = not(true(LP)).
LP = not(true(LP)).
?- unify_with_occurs_check(LP, not(true(LP))).
false.
LP := ~(L ⊢ LP)
00 ~ 01
01 ⊢ 01, 00
02 L
The cycle in the direct graph of LP is
an infinite loop that make no progress
towards the goal of evaluating LP as
true or false.
-- Copyright 2024 Olcott "Talent hits a target no one else can hit; Geniushits a target no one else can see." Arthur Schopenhauer