The Riemann Hypothesis

A single line in the complex plane — if it is where Riemann said it was — tells you everything you could ever want to know about the primes.

The prime numbers are the atoms of arithmetic: 2, 3, 5, 7, 11, 13, 17, 19, 23, and so on, indivisible except by themselves and 1. Every other whole number is built from them in exactly one way. They are the simplest objects in mathematics, and they are utterly wild. Nobody knows when the next one will turn up. Nobody can write a formula that lists them. They look random — but they are not. Hidden underneath their scatter is an order so precise that finding it would settle a thousand other questions.

The deepest known statement of that hidden order is the Riemann hypothesis. It says, roughly: the primes are as well-behaved as they could possibly be. It has been unsolved since 1859. The Clay Mathematics Institute will pay a million dollars to whoever settles it.

The primes that won't behave

Look at the primes up to a hundred: 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97. Twenty-five of them. Now up to a thousand: 168. Up to a million: 78,498. They thin out, but not in a way you can predict. There are huge gaps and tight clusters; pairs like (11, 13) and (17, 19) sit next to deserts ten or twenty numbers long.

In 1792, at the age of fifteen, Carl Friedrich Gauss noticed something. He counted the primes up to various round numbers and stared at the totals. The density of primes near a number x, he guessed, is about 1 / ln(x). The count up to x, which mathematicians call π(x), should therefore be about x / ln(x). This is the prime number theorem. It took another hundred years to prove, but Gauss saw it from the data alone.

That's the law in the large. The puzzle is what's left over — the error. The actual count π(x) and the smooth approximation never quite agree; they wobble around each other. How big can the wobble get? That is the question the Riemann hypothesis answers.

Euler's identity — primes and a series, fused

The bridge from primes to the rest of mathematics was built by Leonhard Euler in 1737. He noticed a strange equality. On one side, you have the infinite sum

ζ(s) = 1 + 1/2^s + 1/3^s + 1/4^s + 1/5^s + …

This is the zeta function: take every whole number, raise it to a power s, take the reciprocal, add them up. For s > 1 the sum converges.

On the other side, you have a product taken over only the primes:

ζ(s) = ∏_{p prime} 1 / (1 − p^−s)

The two sides are equal. The miracle is that every prime is in there exactly once, and the sum on the left runs over every whole number. The reason they agree is the fundamental theorem of arithmetic: every integer factors uniquely into primes. Euler had turned a statement about primes into a statement about a smooth analytical function. Whatever ζ does — where it is big, where it is small, where it equals zero — is, in disguise, a statement about the primes.

Riemann's leap

For more than a century, ζ(s) was only defined for s > 1. In 1859, in a single eight-page paper — the only one he ever wrote on number theory — Bernhard Riemann did something audacious. He extended the zeta function to complex values of s. That is, s = a + bi, a number with a real part and an imaginary part, plotted on a two-dimensional plane.

Once you allow complex inputs, the zeta function comes alive. It has zeros — places where ζ(s) = 0. Some are easy and uninteresting: at s = −2, −4, −6, and so on. These are called the trivial zeros. Riemann found something else. There are infinitely many other zeros, all sitting inside a vertical strip of the complex plane where the real part is between 0 and 1. And as far as he could compute, every single one of them had real part exactly 1/2.

That observation, dropped in a single sentence near the end of his paper, is the Riemann hypothesis:

All non-trivial zeros of the zeta function lie on the critical line Re(s) = 1/2.

Why does anyone care about the imaginary zeros of an analytical function? Because of Euler's identity. The zeros of ζ control the wobble of π(x) around x / ln(x). The further off the critical line they wandered, the more violently the primes could fluctuate. If they all sit on the line, the primes are as orderly as the prime number theorem permits. If even one zero drifts off, the primes can do things nobody expects.

Why it matters

The hypothesis has been checked, computationally, for the first ten trillion zeros. Every one of them lies on the critical line. But ten trillion is not infinity, and no proof has come. Hundreds of theorems in modern number theory begin with the words “Assuming the Riemann hypothesis…” They sit, finished, waiting for the foundation to be poured. If RH is true, an enormous edifice stands. If it is false, parts of that edifice will need to be rebuilt — and we will have learned something startling: that the primes are stranger than even Riemann imagined.

There is something almost spiritual about it. Hilbert was once asked what he would do if he could be revived in 500 years. He is said to have answered: “I would ask: has the Riemann hypothesis been proven?” That is the question. A statement about a single line in the complex plane — quietly governing the most basic objects in mathematics, the primes — and still, after a century and a half of attack by the best minds alive, undefeated.