So I know the famous mathematician Terence Tao starts his work on ChatGPT these days. And I also understand the inherent risk in doing that because ChatGPT will just learn from the user. Not that I'm trying to be first here because there are literally thousands of far, far more distinguished people than me articulating this much better than me. But sometimes you need to puzzle your way through just to get to the question you want answered. What exactly is the question behind SRH?
Well I gave up and just put current understanding into ChatGPT. My practical logic is rubbish having only done a single module at university so very happy to have the machine tool of ChatGPT to do the lifting here!
What is particular interesting about this Tarski et al. have already shown that meta-predicates cause terminal paradoxes e.g. Godel (incompleteness and inconsistency theorems). Yet here we are negotiating a Self-Reference Paradox non terminally! Much more advanced TODO : how does this apply to Godel's two theorems?
This is based on the statement : "every rule has an exception"
Demonstration of Paradox that is consistent and complete
Statement of the Paradox
Statement Q:
Q: ∀x (R(x) → ∃y ¬T(x, y))
Where:
R(x)
: x is a rule.T(x, y)
: The rule x is true for y.¬T(x, y)
: y is an exception to x.
Interpretation of Q
- Meaning: Every rule x has at least one exception y such that T(x, y) is not true.
Analysis of Q
Assume Q is a Rule:
- Given:
R(Q)
is true (i.e., Q is a rule).
Apply Q to Itself:
- According to Q, since Q is a rule, there must be at least one y such that
¬T(Q, y)
is true. This means there exists at least one y where T(Q, y)
is false.
Consistency Argument:
- Finding Exceptions: If there is some y such that
¬T(Q, y)
is true, then y is an exception to Q. This is consistent with Q’s assertion that every rule (including Q itself) should have exceptions. - No Inherent Contradiction: The presence of an exception y to Q does not invalidate Q; it actually demonstrates Q’s claim. Q claims that every rule has exceptions, and finding such an exception supports Q rather than contradicting it.
Summary of the Paradox
- Self-Referential Nature: Q asserts that every rule must have an exception. If Q is a rule, it must have exceptions, which aligns with its own claim.
- Resolution of the Paradox: The paradox arises from the self-referential nature of Q. However, this self-reference does not lead to a logical inconsistency. Instead, it reinforces Q’s claim because if Q has exceptions, it demonstrates Q’s validity as a rule.
Conclusion
- Consistent and Complete: The apparent paradox is resolved by recognizing that Q’s self-referential nature does not create a logical contradiction. Instead, it aligns with Q’s own definition. The fact that Q has exceptions confirms Q’s validity, making the paradox consistent and complete.
.End
==
Of course the obvious problem here is the function T(x,y) which is an emulator. It takes the unary expression x and determines whether it evaluates to true for y.
Now we will adopt Godel numbering here to expand the power of the logic. If every possible expression is ordered then we can assign it a number in the list which acts as an alternative name. This means a number can be value and it can represent a function. They are interchangeable.
Now T is not immediately self-referential because it is a binary function. However you can Curry the function into a set of functions for each y.
Ty(x) has the same result as T(x,y) which evaluates x(y).
Now we can ask T(Ty,y) which evaluates Ty(Ty)
Ty is just a number, but it is also a function that can be applied to numbers. So it is not technical. It could for example be the function 2345 in the list say x -> x^2 applied to the value 2345 evaluation to 5499025.
But we also have self-reference. By SRH we know to expect problems.
It has been shown before that unfortunately there is no way to compute whether an expression implies self-reference. So SRH depends upon a non-computable result and so SRH cannot be computed. Like Turing's Halting we know that programs may loop forever but there is no universal way to find out. Obviously by inspection 10 GOTO 10 will evaluate forever. But this cannot be developed for all possible programs. Likewise SRH.
Let us simply follow Turing. Create a function Py(x) which evaluates to T(x,y) + 1
Now what is the value of Py(Py)?
Py(Py) = T(Py, y)
But by the definition above Py(Py) = T(Py, y) + 1
We have a simple contradiction. So T(x, y) cannot exist. This is Tarski's result also.
In general if you hand an expression the power to make statements about itself then you can easily construct a contradiction by modifying that result. This indeed would be the essence of SRH. You cannot consistently/completely apply self-reference. Or to turn that around if self-reference can be shown then you cannot be consistent/complete. And in SRH that is interpreted to make the system unbounded which leads to the broader observation that knowledge and the world is unbounded and infinite.
To try and mitigate this you could try a Russel and Whitehead typing system to stop things making statements about themselves. But typing itself is contradictory. These are powerful results that break the very objectives of logic and systems itself.
Now while T(x,y) cannot exist, it does however demonstrate in the first part that a Paradox may not be Inconsistent or Incomplete. This in fact disproves the SRH. That not all instances of self-reference or even Paradox lead to contradiction.
That itself is Paradoxical as surely the definition of a Paradox is contradictory result. But by incorporating paradox into the expression we can allow for it.
==
Unedited ChatGPT conversations leading up to the conclusion. Interesting to see how ChatGPT works.
Definitions
- Predicate for Rules: Let R(x) represent "x is a rule."
- Predicate for Exceptions: Let E(x,y) represent "y is an exception to rule x."
Formal Statement
To express "every rule has an exception," you can use the following logical formulation:
∀x(R(x)→∃yE(x,y))
Negating the Statement
To negate this statement, follow these steps:
Original Statement:
∀x(R(x)→∃yE(x,y))
Apply Negation to the Entire Statement:
¬∀x(R(x)→∃yE(x,y))
Use the Negation of Universal Quantifier:
The negation of a universal quantifier ∀ becomes an existential quantifier ∃:
∃x¬(R(x)→∃yE(x,y))
Negate the Implication:
To negate the implication R(x)→∃yE(x,y), use the fact that:
¬(A→B)≡A∧¬B
Here, A is R(x) and B is ∃yE(x,y). So:
¬(R(x)→∃yE(x,y))≡R(x)∧¬(∃yE(x,y))
Negate the Existential Quantifier:
The negation of ∃yE(x,y) is:
¬(∃yE(x,y))≡∀y¬E(x,y)
Therefore:
¬(R(x)→∃yE(x,y))≡R(x)∧∀y¬E(x,y)
Combine Everything:
Putting it all together, the negation of the original statement is:
∃x(R(x)∧∀y¬E(x,y))
Summary
The negation of ∀x(R(x)→∃yE(x,y)) is:
∃x(R(x)∧∀y¬E(x,y))
This means that there exists at least one x such that R(x) is true and, for all y, E(x,y) is false.
Now let us use the symbol Q to refer to ∀x(R(x)→∃yE(x,y))
Q:∀x(R(x)→∃yE(x,y))
Finding the Negation of Q
To find the negation of Q, follow these steps:
Statement Q:
Q:∀x(R(x)→∃yE(x,y))
Negate the Statement Q:
To negate Q, apply negation to the entire statement:
¬Q=¬∀x(R(x)→∃yE(x,y))
Apply Negation to the Universal Quantifier:
The negation of a universal quantifier ∀ becomes an existential quantifier ∃:
¬Q=∃x¬(R(x)→∃yE(x,y))
Negate the Implication:
To negate the implication R(x)→∃yE(x,y), recall that:
¬(A→B)≡A∧¬B
Here, A is R(x) and B is ∃yE(x,y). So:
¬(R(x)→∃yE(x,y))≡R(x)∧¬(∃yE(x,y))
Negate the Existential Quantifier:
The negation of ∃yE(x,y) is:
¬(∃yE(x,y))≡∀y¬E(x,y)
Therefore:
¬(R(x)→∃yE(x,y))≡R(x)∧∀y¬E(x,y)
Combine Everything:
Putting it all together:
¬Q=∃x(R(x)∧∀y¬E(x,y))
Summary
The negation of the statement Q, where Q is:
Q:∀x(R(x)→∃yE(x,y))
is:
¬Q=∃x(R(x)∧∀y¬E(x,y))
This means there exists at least one x such that x is a rule and for every y, y is not an exception to x.
Can you confirm that R(Q) is true, given that Q is a rule.
Certainly! Let's analyze the situation using the following definitions:
Statement Q:
Q:∀x(R(x)→∃yE(x,y))
This statement Q asserts that for every x, if x is a rule (i.e., R(x) is true), then there exists at least one y such that y is an exception to x (i.e., E(x,y) is true).
Assuming Q is a Rule:
Suppose we consider Q itself as a rule. In other words, we treat Q as a specific instance of x for which R(Q) (i.e., Q being a rule) is true.
Let’s substitute Q into the statement and check the validity:
- If Q is a rule, then R(Q) is true.
Given that Q is a rule, the statement Q says:
Q:∀x(R(x)→∃yE(x,y))
Specifically for x=Q:
R(Q)→∃yE(Q,y)
If R(Q) (i.e., Q is a rule) is true, then:
∃yE(Q,y)
This means that there must exist at least one y such that E(Q,y) is true. In other words, there should be some y for which y is an exception to the rule Q.
Conclusion
- If Q is indeed a rule, R(Q) is true by definition.
- For Q to satisfy its own condition, ∃yE(Q,y) must also be true. This means there should exist some y such that E(Q,y) holds true.
Summary: If Q is a rule, then R(Q) is true. According to the statement Q, this implies that there must be at least one y such that E(Q,y) is true. So yes, R(Q) is true if Q is a rule, provided that Q correctly adheres to its own definition and there exists a y such that E(Q,y) is true.
Let us take a y such that E(Q,y) is true. This means that y contradicts Q.
Let's explore the idea of y contradicting Q when E(Q,y) is true.
Understanding E(Q,y)
Given:
- Q is the statement ∀x(R(x)→∃yE(x,y)).
- E(Q,y) means that y is an exception to the rule Q.
In the context of Q, this translates to:
E(Q,y) means that there exists some y such that E(Q,y) holds true.
Interpreting the Statement Q
The statement Q says:
Q:∀x(R(x)→∃yE(x,y))
For x=Q:
R(Q)→∃yE(Q,y)
- If R(Q) is true: This means Q is a rule.
- Then: There must exist at least one y such that E(Q,y) is true.
Contradiction Analysis
To say that y contradicts Q, we need to understand what it means for E(Q,y) to be true:
- If E(Q,y) is true, then y is an exception to the rule Q.
- For Q itself: This implies E(Q,y) indicates that y is a case where QsËŠ condition does not hold.
Specific Case of Q
If E(Q,y) is true, then y is an exception to Q. In other words:
There exists some y such that E(Q,y) holds true.
To be an exception to Q, y must be such that y does not adhere to the condition specified by Q.
Since Q is:
Q:∀x(R(x)→∃yE(x,y))
- For Q itself: The rule Q states that if something is a rule, then there must be an exception.
- E(Q,y) means y is an exception to Q. This implies that y is an instance where Q does not hold.
Contradiction:
- The statement Q asserts that every rule has an exception. If E(Q,y) is true, it means y is an exception to Q, which, in turn, is a rule asserting that every rule has an exception.
- If E(Q,y) is true, y shows that Q itself has at least one exception, which fits the assertion of Q.
Conclusion
- When E(Q,y) is true, y being an exception to Q aligns with Q's own assertion that every rule (including Q) has an exception.
- There is no inherent contradiction here because E(Q,y) upholds the statement Q by showing that Q has an exception, which is consistent with Q's own claim.
In summary, y being an exception to Q does not contradict Q; instead, it supports Q by demonstrating that Q has an exception, as Q asserts every rule must have at least one exception.
But this means there is no y that contradicts Q which is then a contradiction.
You are correct in recognizing that if Q is a rule, then having an exception y that contradicts Q leads to a paradoxical situation. Let’s delve deeper into this to clarify why it results in a contradiction.
The Paradox of Q
Given:
Q:∀x(R(x)→∃yE(x,y))
Analysis
Assuming Q is a Rule:
- R(Q) is true because we are assuming Q is a rule.
Implication of Q:
- Since R(Q) is true, the statement Q demands that there exists some y such that E(Q,y) is true. This means y is an exception to Q.
Meaning of E(Q,y):
- If E(Q,y) is true, then y is an exception to the rule Q. This means y should be a case where Q does not hold.
Evaluating the Contradiction
To see why this leads to a contradiction:
The Paradox
Assumption:
- Assume Q is a rule, and by definition of Q, it must have an exception.
Contradiction:
- If E(Q,y) is true, then y is an exception to Q. This means y shows a case where Q does not apply. But this would also imply that Q itself is inconsistent or self-contradictory, because it would be an instance of a rule that supposedly cannot be properly applied without leading to contradictions.
Summary of the Paradox
Self-Referential Nature: Q is a self-referential rule that states it must have exceptions. When assuming Q is a rule and finding exceptions, we face a paradox because any exception to Q (i.e., E(Q,y)) implies that Q itself cannot be universally applied without contradiction.
Logical Contradiction: This self-referential situation creates a paradox similar to the liar paradox, where the rule’s own assertion about having exceptions leads to a logical inconsistency.
In essence, if Q is a rule, it requires exceptions by its own definition, but the existence of an exception also suggests that Q is failing in some self-referential way. This circular reasoning and self-reference result in a paradox, demonstrating the inherent contradiction in treating Q as a rule with exceptions.
Can we try and rephrase this. Can we replace the predicate E(x,y), with ¬T(x,y) where T(x,y) states that the rule x is true for y and so ¬T(x,y) means that y is an exception to x.
Certainly! Let's rephrase the original statement using the predicate T(x,y), where T(x,y) indicates that the rule x is true for y, and ¬T(x,y) indicates that y is an exception to the rule x.
Rephrasing with T(x,y)
Original Statement:
Q:∀x(R(x)→∃yE(x,y))
Here, E(x,y) indicates that y is an exception to the rule x.
Replace E(x,y) with ¬T(x,y):
E(x,y)≡¬T(x,y)
Therefore:
Q:∀x(R(x)→∃y¬T(x,y))
This means:
- For every x, if x is a rule, then there exists at least one y such that T(x,y) is not true (i.e., y is an exception to x).
Negation of the Rephrased Statement
To find the negation of this rephrased statement:
Original Rephrased Statement:
Q:∀x(R(x)→∃y¬T(x,y))
Negate the Statement:
¬Q=¬∀x(R(x)→∃y¬T(x,y))
Apply Negation to the Universal Quantifier:
¬∀x(R(x)→∃y¬T(x,y))≡∃x¬(R(x)→∃y¬T(x,y))
Negate the Implication:
Recall that:
¬(A→B)≡A∧¬B
Here, A is R(x) and B is ∃y¬T(x,y). So:
¬(R(x)→∃y¬T(x,y))≡R(x)∧¬(∃y¬T(x,y))
Negate the Existential Quantifier:
The negation of ∃y¬T(x,y) is:
¬(∃y¬T(x,y))≡∀y¬¬T(x,y)≡∀yT(x,y)
Combine Everything:
So:
¬(R(x)→∃y¬T(x,y))≡R(x)∧∀yT(x,y)
Thus:
¬Q=∃x(R(x)∧∀yT(x,y))
Summary
Original Statement:
Q:∀x(R(x)→∃y¬T(x,y))
This means: For every rule x, there exists some y for which T(x,y) is not true (i.e., y is an exception to x).
Negation of the Rephrased Statement:
¬Q=∃x(R(x)∧∀yT(x,y))
This means: There exists at least one x such that x is a rule and for all y, T(x,y) is true (i.e., x applies to every y, so there are no exceptions).
Understanding the Paradox
If Q is a rule and E(x,y) is replaced with ¬T(x,y):
- Q asserts that every rule has an exception.
- If Q itself is a rule, then Q must have exceptions.
- However, if you find an exception y to Q, it means T(Q,y) is not true for that y, which contradicts the rule Q itself if Q claims every rule should have an exception.
Thus, the paradox arises because if Q claims that every rule has an exception and you assume Q itself is a rule, then finding an exception to Q paradoxically validates Q, but also points out that Q might not universally apply, creating a self-referential contradiction.
But can you not argue that Q does universally apply because in the one instance that it does not apply namely to itself, it is actually true.
Yes, you are touching on an interesting point related to self-reference and consistency in logical systems. This is reminiscent of the discussions in formal logic about self-referential statements and paradoxes.
Rephrasing and Self-Application of Q
Let's explore the idea that Q could still be universally applicable, even if it seems to contradict itself in the specific instance of its self-application.
Rephrased Statement:
Q:∀x(R(x)→∃y¬T(x,y))
This means: For every rule x, there exists at least one y such that T(x,y) is not true (i.e., y is an exception to x).
Self-Application of Q:
When applying Q to itself:
- Assume Q is a rule. Therefore, R(Q) is true.
- According to Q, there must be an exception y such that E(Q,y) is true, which translates to ¬T(Q,y) being true. In other words, there should be some y for which T(Q,y) is not true.
Argument for Consistency
Self-Consistency Argument:
Assumption:
Requirement:
- Q asserts that every rule must have at least one exception.
Exception to Q:
- If Q is a rule, then according to Q's own assertion, there must be at least one y such that T(Q,y) is not true. This aligns with Q’s definition as a rule.
Self-Consistency:
- The fact that Q has exceptions does not invalidate Q; rather, it confirms Q's assertion. In this way, Q can be seen as universally applicable because, even in the case where Q applies to itself, the presence of exceptions (as required by Q) demonstrates that Q correctly fulfills its own criterion.
Addressing the Paradox
Application to Itself:
- Q states that every rule has exceptions. If Q itself is a rule, then it must have exceptions.
- Finding exceptions to Q validates Q's claim rather than contradicting it.
Validation of Q:
- Q’s self-reference implies that the presence of exceptions to Q is consistent with Q's definition. This is because Q itself asserts that every rule (including itself) has exceptions.
Conclusion:
- The apparent contradiction is resolved by recognizing that Q's self-referential nature does not invalidate it. Instead, Q’s requirement that it has exceptions reinforces its own claim and adheres to its definition.
In summary, Q can be argued to universally apply even when considering its own exceptions. The presence of exceptions to Q supports the consistency of Q rather than contradicting it. This reflects a form of self-consistency where the self-referential rule Q aligns with its assertion by demonstrating that it indeed has exceptions.