PeteOlcott wrote: ↑Fri Apr 05, 2019 2:55 am
You are conflating some things there. I will try to boil it down much simpler.
If we define truth as derived on the basis of some set of expressions of language
in the same sort of way that theorems are derived from axioms using rules-of-inference
and the same way as true conclusions are derived from valid deduction from true premises
Then it could not be possible for an expression of language G to assert that
it is not provable in F and have this assertion be proven in F, thus true in F.
Well this response only makes me more confused, not less.
I begun a thread that is unlikely being read to address the historical considerations involving "incompleteness". Originally, rational numbers were thought sufficient to define all numbers. But once the Pythagoras Theorem was proven, say using Euclid's system of reasoning, the hypotenuse was deemed, "unexpressible" by a rational expression.
This just implied that the system of real numbers must be expanded for being understood as "incomplete" without doing so. A newer system of reasoning can be added to the prior logic of Euclid and any theories of Number it assumed in extension of its postulates (common notions, that is). The question was whether we could have a universal system of reasoning that links all mathematics under one universal system?
We can PRETEND that there IS such a universal logic system, say 'G', and ask if we can show it to lead to a contradiction. However, this assumes some
apriori assumptions about the system we permit to be useful to judge this. For this simple question, we already would be assuming at least two factors in this very challenge: that we can "pretend", and that "contradiction" is something we think is not permitted in our judging logic system.
We MAY try to be more inclusive by permitting a system that judges without bias to any apriori conditions. But such a system is itself "trivial", literally meaning, 3-values for semantic meaning. In other words, it is already understood that there can be universal machine (system logic) that covers all possible logic systems under one umbrella. But it would require an acceptance of "truth" to be relative and 'fuzzy' with respect to totality. This would also remove the meaning of seeking for some universal logic system because all systems would be considered logical no matter what in the most inclusive way.
So we have to pick a judging logic that is based upon the limited definitions of a 'finite' world: those that allow us to 'pretend' (a rule about assumptions) and have a system devoid of contradictions that lack closure. If we permit a system that lacks closure (completeness), we cannot have any reason to pose the question, "is there such a universal logic" or we'd be expanding the meaning of "logic" to include irrational systems of reasoning. [Note how I'm bringing back the word, "irrational", in context to the historical root of the word as well?]
So the first step is to find the simplest systems of reasoning that is closed with respect to a finite domain. We then want to pretend that this system is itself a qualifying 'judge' about other systems. Certainly if there IS a universal system based on its own minimal limits, then there has to be some truth about it being subject to that universal system itself.
Then we use this 'judging' logic, knowing it is complete, and then PRETEND that there is a universal system. IF this very system you are using is itself only a subset of a complete universal system, then we assume it complete as 'proven' by the meaning of "completeness" and then try to see if we can extend this calculating system to ask possible questions about some larger domain OUTSIDE of it's defined one.
When we PRETEND that there is a goal we can reach outside of the minimal rules of logic, if we find a contradiction, we must treat the contradiction as proof that what we pretended is either NOT true OR there is some newer system of reasoning that requires extending the minimum laws about things like contradiction or pretenses.
Using this system, with the assumption that contradiction is not allowed, leaves us into a contradiction of the judging system. This means that either the 'judging system' is flawed or at least, "incomplete".
So PRETEND that you are correct, ...that any Incompleteness Theorems are false. You have some system 'judging' this to be true on what minimal logical assumptions? If you permit contradiction in your system, you are always certain to be 'proved' correct because contradiction allows for more than two truth values, 'true', versus 'false'. Your ARE asserting the theorems 'false' with exclusive distinction are you not? If not, you would at best be saying they are both 'true' and 'false'. If you don't permit something to be both, then you default to a LIMITED system of logic as already understood by those making those theorems. That is, they already assumed non-contradiction as a primary limitation. And this further means that they, like you, require treating "truth" about these systems as BINARY, not tri-vial.
Using your system, you'd require to show that your own reasoning is itself binary in truth values or you are defaulting to assume your system to allow for contradiction and you'd be hypocritical to use it to demonstrate that an external rationale is at fault for assuming non-contradiction systems complete or not. This is something you do not seem to care to do as you rely on an assumption of external machines to do this for you (like the Lambda calculus or Prolog, etc).
The point is that the "theorems" used to show incompleteness are supposed to end in their own contradiction. This way you can ask if that 'judging system' itself using its minimal rule of non-contradiction is sufficient as a PARENT logic to all systems it can build upon it. (?) So either "G" is contradictory or not. If proven contradictory when we PRETEND it is the most sufficient machine to exhaustively cover all domains, then it means either that system is wrong or that it is correct but means there will always be some domain outside of it to which some other MORE complete logic is constructed from.
The theorems stand true precisely for your own apparent intent to be non-contradictory and yet be unable to do so without permitting all systems that include at least a third truth value.
This is where I interpret contradiction itself as a foundational reality to everything. It IS a 'trivial' system that implies everything possible in some greatest domain, I call "Totality". It rationally permits "contradiction" to be a
force that constantly tries to BE consistent when it can only do so at the cost of becoming 'inconsistent' for trying. Thus a 'universal logic' can exist, but then reduces the meaning of such a 'logic' as inclusive of those things we think of as 'non-logical'. So we have to stick with some fixed limitation of the term 'logical' by some stricter subset of all possibilities or we lose meaning of 'logical' for nothing to be 'illogical'.