Poincaré's Three-Body Problem

A prize-winning error in celestial mechanics revealed that even Newton's tidy universe hid an intractable wildness at its core — and quietly gave birth to chaos theory, seventy years early.

When Isaac Newton published the Principia in 1687, he did more than write down three laws of motion. He handed humanity a receipt. Give me the position and velocity of two point-masses attracted by gravity, he said, and I will hand you back an equation that says exactly where each of them will be, forever. Not “approximately.” Not “for a while.” The full future. The full past. A closed-form ellipse, worked out with pen and ink.

This was the origin of a picture the West lived inside for two centuries. The universe was a clockwork. Its laws were simple, its predictions exact, and its trajectory — given the initial conditions — was already written. Pierre-Simon Laplace pushed the picture to its conclusion: a mind that knew every particle and every force could see the entire history of the cosmos in one glance.

Then someone added a third body, and the picture broke.

Two bodies, solved forever

Newton's insight for two bodies is beautiful and, once you see it, almost obvious. Treat the pair as a single system moving around its common centre of mass. The problem separates: the centre drifts at constant velocity, and each body orbits the centre in a conic section — a circle, ellipse, parabola, or hyperbola depending on the energy.

There are ten conserved quantities in the game — the three components of total momentum, the three of angular momentum, the total energy, and the three coordinates of the centre of mass — and, for two bodies, ten is enough. Once you feed the equations those constants, they integrate cleanly. You can write down the position of the Earth around the Sun as a function of time and be correct to whatever precision your instruments can measure.

Astronomy in the eighteenth century became, in essence, an industry of two-body corrections. Le Verrier used tiny gravitational tugs on Uranus to predict the existence of Neptune before it was observed. Kepler's laws, derived by Newton from his own more general laws, held with almost embarrassing fidelity. If the universe was a clock, the two-body problem was its escapement.

The King's prize

The three-body problem is easy to state and impossibly hard to solve. Three point-masses, gravitating according to Newton. Given their positions and velocities at some moment, predict their motion for all time.

For over a century, mathematicians tried. Euler solved a restricted version in which one body was vanishingly small. Lagrange found five special configurations — the Lagrange points — where three bodies could sit in a stable relative arrangement. But the general case defied every technique. The equations refused to integrate. There simply weren't enough conserved quantities to reduce the system to something solvable.

In 1885, King Oscar II of Sweden and Norway announced a prize. Whoever could produce a mathematical solution to the n-body problem — a series expression for the coordinates of any number of bodies moving under gravity, valid for all time — would receive 2,500 crowns and a gold medal. It was one of the most prestigious problems of the age.

Henri Poincaré, then a young French mathematician already regarded as the most powerful mind of his generation, took it on. He didn't manage the general case, but he made spectacular progress on the restricted three-body problem — one small body moving under the influence of two heavier ones — and in 1888 he was awarded the prize. His memoir was set to be published in the journal Acta Mathematica.

Then, while checking galley proofs, Poincaré and the editor Lars Phragmén found an error. It was small in appearance and catastrophic in consequence. Poincaré had assumed that certain trajectories in his construction behaved smoothly. They didn't. He had to withdraw the printed copies at his own expense and rewrite the paper. The revised version was three times longer, and its conclusion was, in a sense, the opposite of the one he'd started with.

What Poincaré found

Poincaré had been looking at a geometric object called a homoclinic point — a point where an unstable trajectory eventually returns to itself. He'd assumed the two branches of that returning trajectory would meet cleanly and close up. Instead, he realised, they crossed each other. And if they crossed once, symmetry forced them to cross again. And again. Infinitely many times.

The resulting picture — the homoclinic tangle — is one of the most consequential drawings never quite drawn in the history of science. Poincaré himself refused to try. In Les Méthodes Nouvelles de la Mécanique Céleste he wrote:

“When one tries to depict the figure formed by these two curves and their infinite intersections... these intersections form a kind of trellis, tissue, or grid with infinitely tight mesh... One is struck by the complexity of this figure, which I shall not even attempt to draw.”

Two facts fell out of the tangle, and they changed physics.

First: there is no clean series solution to the general three-body problem, not in closed form and not (as Bruns and Poincaré himself proved) in the class of algebraic functions available to nineteenth-century analysis. The extra conserved quantities that would let you integrate the system simply do not exist. The dream of an explicit formula was over.

Second — and this was the deeper blow — the trajectories of the three bodies are exquisitely sensitive to their starting conditions. Two systems with initial states differing by less than any instrument could measure would, after enough time, follow completely different paths. Laplace's demon, however godlike, needed exact initial data to predict the future. In a three-body system, “exact” means infinitely precise. Reality does not offer such precision.

The birth of chaos

Poincaré did not use the word “chaos” — that came from Li and Yorke in 1975 — but he had discovered it. Determinism and predictability, which everyone had quietly assumed were the same thing, came apart in his hands. A perfectly deterministic system can be radically unpredictable, given only finite information about its state. The wall between mechanism and unpredictability had cracked open, and physics had walked through without noticing.

The finding sat mostly ignored for seventy years. Then, in 1963, the meteorologist Edward Lorenz noticed that a simplified weather model on his computer produced wildly different forecasts when he restarted a run with numbers rounded to three decimal places instead of six. He'd rediscovered Poincaré's sensitivity, in a system as ordinary as convecting air. The metaphor — a butterfly flapping its wings in Brazil — became famous. The mathematics was already there, waiting in a Swedish journal.

Today the descendants of Poincaré's tangle run through half of applied mathematics. Dynamical systems theory, symbolic dynamics, ergodic theory, the KAM theorem, strange attractors, fractal geometry — all sit downstream of that 1889 rewrite. The solar system itself, we now know, is mildly chaotic: over hundreds of millions of years, we cannot predict the precise orbits of the planets, only the statistical envelopes they will inhabit.

What Poincaré revealed was not that Newton was wrong, but that Newton was incomplete in a way Newton could not have seen. The clockwork universe still runs on Newton's laws. It just doesn't run in the way a clock does. It runs like weather. Like traffic. Like life. The equations are exact; their consequences are wild. That gap — between what we can write down and what we can foresee — is the space chaos theory inhabits, and it opened one afternoon in a Paris printing office, when a young Frenchman found a mistake in his own galleys.