Fermat's Last Troll

The Fires of History Newsletter, No. 9

Sometime around 1637, a French lawyer named Pierre de Fermat was reading his copy of Diophantus’s Arithmetica, a third-century Greek text on number theory, and he scribbled a note in the margin. The note, written in Latin, translates roughly as follows:

“It is impossible to separate a cube into two cubes, or a fourth power into two fourth powers, or in general, any power higher than the second, into two like powers. I have discovered a truly marvelous proof of this, which this margin is too narrow to contain.”

Then he closed the book, went about his life, and died in 1665 without ever writing the proof down anywhere else.

His son found the note and published it in 1670. And with that publication, Fermat lit a fuse that would burn for 358 years.

The Claim

The claim, now known as Fermat’s Last Theorem, is deceptively simple. It states that no three positive integers a, b, and c satisfy the equation a^n + b^n = c^n for any integer value of n greater than 2.

For n = 2, you get the Pythagorean theorem, and solutions are everywhere: 3, 4, 5; 5, 12, 13; and so on. But Fermat said: bump that exponent up by even one, and solutions vanish entirely. Forever. For all numbers.

And he could prove it.

But the margin was too small.

Write it on another piece of paper, Pierre. Write it in a letter. Write it on the wall. The excuse is so transparently inadequate that it functions as a taunt: I have the answer and you do not, and the reason you do not have it is that I could not be bothered to find a larger piece of paper.

The Structure of a Perfect Troll

Consider the elements.

The claim is extraordinary. Fermat was not asserting something modest. He was asserting a universal negative – that something is impossible across an infinite domain. Proving a positive is hard. Proving a universal negative is orders of magnitude harder. To claim you have done it and then decline to show your work is not modesty. It is provocation.

The excuse is absurd. “This margin is too narrow to contain it” is the seventeenth-century equivalent of “I would explain but you wouldn’t understand.” It is calibrated to infuriate. A serious mathematician who had achieved what Fermat claimed would have moved heaven and earth to publish the proof. The deliberate casualness of the margin note – the offhandedness of I have done the impossible but cannot be bothered to document it – is what makes it a troll rather than a claim.

The delay is infinite. Fermat died without producing the proof. The note was published posthumously. The audience, every mathematician who read it for the next three and a half centuries, was left provoked, uncertain, and unable to resolve the provocation. Did he have a proof? Was he bluffing? Was he wrong? The only way to answer the question was to actually prove the theorem. That turned out to be one of the hardest problems in the history of mathematics.

The troll cannot be resolved by ignoring it. This is the critical feature. If Fermat’s claim was wrong – if solutions existed for some exponent n > 2 – then finding a counterexample would have killed the troll instantly. But nobody ever found one. For 358 years, every attempt to find a counterexample failed. The absence of a counterexample did not prove Fermat right (absence of evidence is not evidence of absence), but it kept the troll alive. You could not dismiss it. You could not resolve it. You could only keep working.

The Obsession

And they worked. For three and a half centuries, Fermat’s margin note consumed some of the best mathematical minds in Europe and, later, the world.

Euler proved the case for n = 3 in the 1770s. Sophie Germain, one of the great mathematicians of the early nineteenth century, who had to submit her early work under a male pseudonym because the Ecole Polytechnique did not admit women, proved a broad class of cases in 1823. Ernst Kummer, in the 1850s, developed entirely new mathematical tools (ideal numbers) specifically to attack Fermat’s theorem and proved it for all “regular” primes. Each generation of mathematicians chipped away at the problem, proving special cases, developing new techniques, pushing the boundary further – without ever reaching a general proof.

The theorem became more than a problem. It became a grail. The Wolfskehl Prize, established in 1908, offered 100,000 marks for a proof. Paul Wolfskehl, according to legend, had been about to commit suicide when he became absorbed in a mathematical paper related to Fermat’s theorem; the distraction saved his life, and he established the prize in gratitude. Whether the suicide story is true or apocryphal, it captures something real about the theorem’s hold on mathematical imagination: it was a problem so compelling that it could, quite literally, give you a reason to live.

Fermat’s margin note generated entire branches of mathematics. Algebraic number theory exists, in substantial part, because mathematicians needed new tools to attack the theorem. The theory of elliptic curves, modular forms, and Galois representations – the mathematical machinery that would ultimately prove the theorem – was developed over two centuries by mathematicians working on problems that Fermat’s three-sentence note had made urgent.

A scribble in a margin produced more mathematical progress than most research programs produce in decades.

The Proof

In 1993, Andrew Wiles announced a proof of Fermat’s Last Theorem at a lecture in Cambridge. He had worked on it in near-total secrecy for seven years, telling almost no one, retreating to his attic to work. The announcement made headlines worldwide – one of the vanishingly rare occasions when a mathematical result was front-page news.

There was a gap in the proof. For over a year, Wiles struggled to fix it. In September 1994, with the help of Richard Taylor, he found the fix. The final proof was published in 1995: 129 pages long, relying on the modularity theorem for semistable elliptic curves – a connection between two branches of mathematics (elliptic curves and modular forms) that did not exist as a conjecture until 1955 and was not remotely within the technical reach of seventeenth-century mathematics.

Wiles used tools that would not be invented for centuries after Fermat’s death: Galois representations, Hecke algebras, deformation theory. The proof is a monument of twentieth-century mathematics, and it is absolutely, categorically beyond anything Fermat could have produced.

Did Fermat Have a Proof?

Almost certainly not.

The mathematical consensus is overwhelming: whatever Fermat thought he had, it was not a valid proof of the general case. He may have had a proof for specific exponents – n = 3 and n = 4 have relatively elementary proofs that were within his reach. He may have had a flawed argument that he mistook for a complete one. This happens to mathematicians all the time; the history of Fermat’s theorem is littered with failed proofs by brilliant people who thought they had it.

Or – and this is the possibility that makes mathematicians twitch – he may have known exactly what he was doing when he wrote that note.

He may have been trolling.

There is no way to know. Fermat is dead. The margin note is all we have. And the note is perfectly ambiguous – it admits both the sincere reading (he believed he had a proof and was too lazy to write it down) and the provocative reading (he knew the claim was stronger than his evidence and wrote the note to see what would happen).

The result, either way, is the same. A single sentence in a margin generated 358 years of mathematical effort, produced entire new branches of number theory, motivated the development of algebraic geometry, and culminated in one of the great intellectual achievements of the twentieth century.

If that is not a productive provocation, the term has no meaning.

The Lineage

Fermat is the first in a lineage of mathematical trolls – people who made claims so outrageous that the establishment’s attempt to refute them produced more progress than the claims themselves.

After Fermat came Georg Cantor, who proved in 1874 that there are different sizes of infinity. The statement sounds like nonsense. It is not. Cantor demonstrated, with rigorous proof, that the real numbers are a bigger infinity than the natural numbers. Leopold Kronecker, one of the most powerful mathematicians of the era, responded by calling Cantor a “scientific charlatan” and a “corruptor of youth” – the same charge Athens leveled at Socrates. Kronecker could not prove Cantor wrong. He could not. So he tried to destroy Cantor’s career instead. Cantor suffered mental breakdowns and died in poverty. His mathematics is now the foundation of set theory, topology, and real analysis. Kronecker is remembered primarily as the man who was wrong about Cantor.

After Cantor came Kurt Godel, who proved in 1931 that any consistent mathematical system powerful enough to express basic arithmetic contains true statements that cannot be proven within the system. He used mathematics to troll mathematics. He turned the tools of formal logic against formal logic itself. The proof is self-referential: Godel constructed a statement that says, in effect, “This statement cannot be proven.” If the system proves it, the system is inconsistent. If the system does not prove it, the statement is true and the system is incomplete. The statement is a mathematical liar’s paradox, except it is not a paradox. It is a theorem. It is true.

And after Godel came Alan Turing, who proved that computation cannot compute its own limits, and then asked whether a machine that can fool you into thinking it is human should be called intelligent. The Turing test, as I argued in last week’s essay, is a formalized trolling problem.

Fermat, Cantor, Godel, Turing. Margin note, diagonal argument, incompleteness theorem, imitation game. Each one a provocation. Each one dismissed before it was accepted. Each one now foundational.

The tradition of trolling – of making a claim so outrageous that the world is forced to respond, and in responding, discovers something it did not know – is not incidental to how mathematics progresses. It may be the primary mechanism.

This essay draws from The Fires of History, coming fall 2026.

Next week: The compliance industry – PCI, cyber insurance, AI safety boards – is the most expensive theater production in history. And the audience is starting to notice the props are made of cardboard.

Source URLs