This post was rejected for the following reason(s):
Difficult to evaluate, with potential yellow flags. We are sorry about this, but, unfortunately this content has some yellow-flags that historically have usually indicated that the post won't make much sense. It's totally plausible that actually this one is totally fine. Unfortunately, part of the trouble with separating valuable from confused speculative science or philosophy is that the ideas are quite complicated, accurately identifying whether they have flaws is very time intensive, and we don't have time to do that for every new user presenting a speculative theory or framing (which are usually wrong).
Our solution for now is that we're rejecting this post, but you are welcome to submit posts or comments that are about different topics. If it seems like that goes well, we can re-evaluate the original post. But, we want to see that you're not just here to talk about this one thing (or a cluster of similar things).
Not obviously not Language Model. Sometimes we get posts or comments that where it's not clearly human generated.
LLM content is generally not good enough for LessWrong, and in particular we don't want it from new users who haven't demonstrated a more general track record of good content. See our current policy on LLM content.
We caution that LLMs tend to agree with you regardless of what you're saying, and don't have good enough judgment to evaluate content. If you're talking extensively with LLMs to develop your ideas (especially if you're talking about philosophy, physics, or AI) and you've been rejected here, you are most likely not going to get approved on LessWrong on those topics. You could read the Sequences Highlights to catch up the site basics, and if you try submitting again, focus on much narrower topics.
If your post/comment was not generated by an LLM and you think the rejection was a mistake, message us on intercom to convince us you're a real person. We may or may not allow the particular content you were trying to post, depending on circumstances.
As the title suggests, this article offers a technical refutation of Gödel’s First Incompleteness Theorem. It’s not a matter of philosophical interpretation, but a direct critique of the logical structure behind the proof. The main idea is simple: it’s not possible to build a truly self-referential statement within the language of arithmetic. Such statements aren’t true, false, or undecidable—they simply don’t exist in that language.
So, the statement Gödel constructed, which seems to say “this statement cannot be proven”, is just an illusion—a misinterpretation caused by a superficial and careless analysis.
To make this refutation easier to follow, we’ll begin with a brief look at the historical background. Then we’ll explain a few key ideas, such as Gödel numbering and the proof function. After that, we’ll present a short fable called “The Mansion of Statements”, which helps to clearly show the conceptual mistakes in the proof.
The language used throughout will be simple and accessible, so that even readers unfamiliar with the theorem can follow the thread and understand the main points. Although it might seem like advanced maths is needed, that’s not the case. The problem isn’t in the formalism or technical details. If it were, fixing them would be enough, and the theorem would still hold. But that’s not the issue. The mistake is conceptual—and that means it can’t be fixed.
Historical Context
At the beginning of the 20th century, mathematics was going through its most ambitious phase. David Hilbert dreamed of a formal system that was complete, consistent, and capable of proving all mathematical truths. His motto was clear: “We must know. We will know.”
In 1931, Kurt Gödel published his famous incompleteness theorem, which seemed to shatter that dream. According to Gödel, any sufficiently powerful formal system would contain true propositions that could not be proven within the system itself. The mathematical community was fascinated. The theorem was received as a revelation about the limits of knowledge.
But not everyone shared that enthusiasm. Ludwig Wittgenstein, from the field of philosophy of language, warned that something didn’t quite add up. For him, formal language could not speak about itself without falling into contradiction: “A proposition cannot say of itself that it is true.”
This text picks up that criticism—not from the perspective of philosophy of language, but from the path Wittgenstein never fully explored: a refutation based on rigorous logical reasoning. Wittgenstein sensed the problem; here, its nature is clearly laid out.
What Does Gödel’s Theorem Say?
Gödel’s theorem, broadly speaking, claims that there are mathematical truths which cannot be proven. To justify this, Gödel constructed a formula which, according to its usual interpretation, says of itself that it cannot be proven. If it were false, then it could be proven—leading to a contradiction. That’s why it is concluded to be true but unprovable. In other words, mathematics is incomplete.
What Is Gödel Numbering?
To build his proof, Gödel needed the system to be able to talk about itself. To do this, he invented a way of turning statements into numbers, as if each one had a unique numerical label.
He assigned a number to each symbol in the language (such as ∀,+,=, etc.), and then combined those numbers using powers of prime numbers to encode entire statements.
In this way, every formula in the system could be represented by a number, known as its Gödel number. By combining mathematical symbols with these numbers, the formal language seemed to speak about itself.
In the image below, you can see how Gödel number #21443 is simply the label that represents the statement: “3+7=10”.
What Is the proof Function?
This function allows us to check whether a statement has a valid proof. That’s why its construction is a key part of Gödel’s demonstration. For example, in the previous case, Gödel number #21443 represents the statement “3+7=10”. So, proof(#21443) returns the value “true”, because “3+7=10” is a statement that can clearly be proven.
How Is the Self-Referential Statement Constructed?
Starting from the proof function, the statement G is defined as: ¬∃proof(G), which means that there is no proof of the statement G. But G is, at the same time, the encoding of that very statement. So, the statement as a whole can be interpreted as: “This statement cannot be proven”, and it is precisely this interpretation that leads to the conclusion that the system is incomplete—or, in other words, that there are truths which cannot be proven.
The Mansion of Statements
What follows is a short story, told as a fable, that sums up the core of the refutation in conceptual terms. You hardly need any knowledge of mathematics to understand it—just follow the thread carefully and let yourself be guided by its internal logic.
The story begins
There exists an immense mansion where some very special letters are kept. Each of these letters contains a mathematical statement: some are true, others false, and some are even badly written. But it doesn’t matter — everything that can be said in mathematical language is found in one of those letters. There are so many letters that, for the vast majority, it is still unknown whether they are true or false.
One day, Professor Hilberton appeared in the mansion. He was a curious and patient man, convinced that it was possible to determine, without any doubt, which of those letters were true and which were false. He claimed that one simply had to search calmly and with confidence.
Hilberton proposed a simple method: if someone found a letter stating “The number 27 is divisible by 3”, they might not immediately know whether the statement was true or false. But if they found another letter explaining “A number is divisible by 3 if the sum of its digits is also divisible by 3”, and another one stating “2 + 7 = 9”, then they could deduce that the first letter was true.
This was how the process worked: each time a true letter was discovered, that letter could be used to find others that were also true. And little by little, more and more true letters could be uncovered. Not long afterwards, Dr Goldenstein appeared. He was more sceptical than Hilberton. He didn’t believe it was so easy to discover all the mathematical truths hidden in the mansion. So he began to devise a plan to convince Hilberton that he was mistaken. Since Goldenstein had a very clever dog named Rufus, he decided to teach him how to search for true letters, just as Hilberton had suggested. But Rufus had a peculiarity: he needed the letters to be marked. On their reverse side, there had to be a number acting as a label. Thus, all the letters would have an identifying code on the back — for example, #250 — and on the front, the corresponding mathematical statement — for example: “The number 4 is even.”
So, each time Rufus was given a number, he would roam the mansion in search of the corresponding letter, pick it up gently, turn it over, and read the statement. Then, applying Hilberton’s method, he would decide whether the letter was true, false, or simply poorly written.
It was then that Goldenstein decided to introduce him to Hilberton.
“He’s called Rufus,” he said. “He does exactly what you propose: he searches for true letters following your steps. He just needs them to be marked with a number on the front.”
Hilberton smiled with joy, thinking that, since Rufus was fast and efficient, all the truths hidden in the mansion’s letters would soon be known.
But Goldenstein had an ace up his sleeve — or rather, a letter from the mansion up his sleeve.
“Wait,” he said to Hilberton. “Not all the true letters in the mansion can be proven to be true.”
“What do you mean?” asked Hilberton, surprised.
“There’s one in particular that is true, but cannot be proven. The letter says: “The letter with the number on my reverse cannot be proven.” Or, in other words: “This letter cannot be proven.”
Hilberton frowned.
"And how do you know it’s true?"
Goldenstein replied calmly: "Because it cannot be false. If it were false, it would mean that it can be proven. But if it can be proven, then what it says is true — and therefore it cannot be proven. That creates a contradiction. So the only coherent option is that the letter is true, but unprovable."
Hilberton was left perplexed by Goldenstein’s argument, as it seemed genuinely consistent.
So that day he went home feeling very sad, thinking that his dreams would never come true — or at least not in the fullness he had hoped for. There was a letter that was true but could never be proven; this extravagant idea kept him from sleeping.
Nevertheless, the next day he returned determined. He asked Goldenstein to lend him Rufus, and took him to the mansion to see for himself whether he was as good as claimed.
After a few days, Hilberton returned from the mansion smiling and went to see his friend Goldenstein.
"You were wrong," he said. "Rufus does his job very well, but that letter you mentioned isn’t in the mansion, and I can prove it."
"How?" replied Goldenstein, intrigued.
"The first thing I noticed upon arriving at the mansion," explained Hilberton, "is that there are no letters that refer to other letters. All the numbers that appear in the statements are nothing more than quantities, and that’s how Rufus understands them."
"I ran a test," he continued. "I wrote a letter labelled #300 that said on its front: “#600 cannot be proven.” I meant that the letter with label #600 couldn’t be proven, but Rufus understood that the number 600 couldn’t be proven. So he searched for the letter that said “300 + 300 = 600”, found it, and told me the letter was false — because it could be proven that 600 exists — and marked it as such."
Goldenstein was deep in thought.
"So Rufus didn’t understand that you were referring to another letter?" he asked.
"Exactly," Hilberton replied. "To Rufus, the number 600 isn’t a label, it’s a quantity. He doesn’t interpret that we’re talking about another letter, but rather about a number that represents an amount."
"But..." said Goldenstein, "the letters that refer to other letters are written in perfectly valid mathematical language."
"Yes," said Hilberton, "but for every statement you write in a letter thinking it refers to a different letter, there already exists an exact replica in the mansion — with the same symbols, from the first to the last — but with a different meaning: one in which the numbers only represent quantities, not references to other letters."
"Remember," he added, "that all possible arithmetic statements already exist in the mansion. It’s impossible to construct a new letter that isn’t already there. And besides, the letters in the mansion existed before they were given identifying labels. Those labels haven’t changed their meaning."
"Then how could we make Rufus understand that some letters refer to other letters?" asked Goldenstein.
Hilberton thought for a moment.
"We could make him see that those letters aren’t mansion letters, but rather something like helper notes. External tools that allow us to deduce the true letters of the mansion."
"Helper notes?"
"Yes. For example, imagine you have a note that says: “#600 cannot be proven.” And it turns out that the letter labelled #600 says: “2 + 2 = 5.” Then Rufus would have to reconstruct the meaning by replacing #600 with what that letter actually says. That is: “2 + 2 = 5 cannot be proven.” Now that is a mansion letter, because it’s formulated as a pure mathematical statement, without external references."
"But I don’t understand," said Goldenstein. "What’s the difference between a helper note and a mansion letter?"
"Helper notes are written in your language — in Goldenstein’s language — a language in which there are external references. But the mansion letters are written in the language of arithmetic, a language in which there are no external references, and numbers only represent quantities. And yes, it’s true that you use the same symbols, but it’s also true that many languages share the same alphabet — and that doesn’t make them the same language."
"In fact," Hilberton continued, "if you go to the mansion with a letter that says “200 + 200 = #400”, where #400 is meant as a reference to another letter, that letter is true — because 200 + 200 = 400. It doesn’t matter what’s written in the letter labelled #400. In your language, there are references to other letters, but in the language of arithmetic, there aren’t."
Goldenstein was thoughtful.
"So my statement — what about it?" he asked.
Hilberton replied firmly: "Your statement cannot be reconstructed. If you have a helper note labelled #500 that says: “#500 cannot be proven”, then Rufus would have to go step by step. But when trying to reconstruct it, he would enter an infinite loop. He’d start writing: “#500 cannot be proven cannot be proven”, then: “#500 cannot be proven cannot be proven cannot be proven”, and so on, never reaching a concrete statement from the mansion. A genuine letter is never reached, which is why your letter cannot be in the mansion. It’s not false, nor true, nor poorly written. It simply isn’t there — it exists only in your imagination."
How the Fable Relates to Gödel’s Theorem
The dog Rufus symbolises the proof function. This function works with encoded statements—that is, with the numbers shown on the back of each card. But proof needs to recover the statement that the Gödel number represents, which is like turning the card over. In doing so, it can check whether the statement can be proven or not.
Moreover, any number that appears within the statement is treated as a natural number, not as an external reference to another statement.
In particular, when recovering the statement G: ¬∃proof(G), the symbol G appears again, and proof mistakenly interprets it as a natural number. However, if proof were able to interpret G correctly—as the encoding of a new statement—this would lead to an infinite loop of decoding, making it impossible to construct a valid statement within the language of arithmetic.
Therefore, it is not correct to translate the statement G as “This statement cannot be proven.” That interpretation comes from a superficial analysis, which overlooks semantic confusions and the limitations of the system.
For all these reasons, we conclude that the theorem is mistaken, since the statement G is not truly self-referential. As a result, the classical conclusions about the system’s incompleteness are not valid.
For more information, you can access the full article at:
Note: This article was written entirely by the author. An AI tool was used to improve the writing style and to assist in generating some of the illustrative images.
