The Unwanted Solution

In January 1916, two months after Einstein published the field equations of general relativity, Karl Schwarzschild solved them exactly for the case of a spherically symmetric mass. He was serving at the Russian front, calculating between artillery bombardments, and would be dead of pemphigus within four months. His solution described the geometry of spacetime around a point mass. It also contained a singularity: at a radius of 2GM/c squared, the metric coefficients diverged. Below this radius, the roles of space and time reversed. Nothing, including light, could escape.

Einstein did not want this. He acknowledged Schwarzschild's mathematics but spent the next two decades arguing that nature would not permit matter to collapse to such densities. In 1939, he published a paper in the Annals of Mathematics that purported to prove it: "On a Stationary System with Spherical Symmetry Consisting of Many Gravitating Masses." His conclusion was that particles in a gravitational cluster would reach a limiting density before reaching the Schwarzschild radius. The argument was physically reasonable. It was also wrong.

In 1939, the same year, Oppenheimer and Snyder published "On Continued Gravitational Contraction," showing that a sufficiently massive star, once it exhausted its nuclear fuel, had no mechanism to halt collapse. The mathematics was straightforward application of Einstein's own equations. In 1965, Roger Penrose proved the singularity theorem: under general conditions, gravitational collapse to a singularity was not a special case but inevitable. The first strong observational candidate, Cygnus X-1, was identified by 1971.

The field equations did not care what Einstein believed about nature's taste. They described a geometry. The geometry permitted a singularity. The singularity existed.

In 1928, Paul Dirac wrote down the equation that unified quantum mechanics with special relativity for the electron. The equation was elegant: first-order in both space and time, Lorentz invariant, and it predicted the electron's spin and magnetic moment without additional assumptions. It also predicted something Dirac did not ask for: solutions with negative energy.

A free particle governed by the Dirac equation could occupy states of arbitrarily negative energy. This was not a minor artifact. If negative-energy states existed, every electron would radiate photons and cascade downward forever, releasing infinite energy. The equation seemed to predict the impossibility of stable matter.

Dirac's response was not to discard the equation but to interpret the unwanted solutions away. In 1930, he proposed the "Dirac sea": all negative-energy states are already filled by an infinite number of unseen electrons. The Pauli exclusion principle prevents any positive-energy electron from falling into them. A hole in this sea — a missing negative-energy electron — would behave like a particle with positive energy and positive charge. Dirac initially suggested the hole might be the proton. Hermann Weyl showed it had to have the electron's mass. Dirac conceded, reluctantly, that his equation predicted a new particle: a positive electron.

In August 1932, Carl Anderson, working with a cloud chamber and a strong magnetic field at Caltech, photographed a track that curved the wrong way. The particle had the electron's mass and opposite charge. He called it the positron. The equation had been right. The unwanted solution was not an artifact. It was antimatter.

In 1861, James Clerk Maxwell was assembling the equations of electricity and magnetism into a unified framework. Ampere's law described how electric currents generate magnetic fields. But Maxwell noticed an inconsistency: for time-varying electric fields, Ampere's law violated conservation of charge. The divergence of the curl of B should be zero, but with only conduction current as a source, it was not.

Maxwell added a term. He called it the displacement current: a contribution to the magnetic field proportional to the time rate of change of the electric field. No experiment had detected a displacement current. No physical mechanism was proposed for how a changing electric field in empty space could act as a current. Maxwell added it because without it, the equations were inconsistent.

With the displacement current included, Maxwell's equations became symmetric between electric and magnetic fields. They also became a wave equation. The equations predicted self-propagating transverse electromagnetic disturbances traveling at a speed determined by two laboratory constants: the permittivity and permeability of free space. Maxwell computed the speed. It was the speed of light.

Maxwell published "A Dynamical Theory of the Electromagnetic Field" in 1865. He died in 1879. In 1887, Heinrich Hertz generated and detected electromagnetic waves in his laboratory in Karlsruhe, confirming what a mathematical consistency fix had predicted twenty-six years earlier. The equations had discovered a physical phenomenon that their author added as a bookkeeping correction.

In 1949, Kurt Gödel presented a solution to Einstein's field equations describing a universe filled with rotating dust. The solution was exact. It was also, by the standards of physical intuition, absurd: it contained closed timelike curves. A particle — or a person — following one of these curves would return to its own past. General relativity, the theory that described gravity as the curvature of spacetime, permitted time travel.

Gödel was not a physicist making an approximation. He was a logician reading the formalism literally. He and Einstein were colleagues at the Institute for Advanced Study. He presented the solution as a birthday gift for Einstein's seventieth, and published it as "An Example of a New Type of Cosmological Solutions of Einstein's Field Equations of Gravitation." The paper was nine pages. The mathematics was correct.

Einstein's response, published the same year in a volume of replies, was careful. He accepted the mathematics. He expressed discomfort with the physical implications. He suggested that solutions permitting closed timelike curves might be excluded by boundary conditions or by the requirement that the universe be expanding rather than rotating. The observable universe does not appear to rotate. But this was a constraint on our universe, not on general relativity. The equations did not distinguish between universes Einstein found acceptable and those he did not.

Subsequent work found more solutions with closed timelike curves. The Kerr metric, describing a rotating black hole, contains them in its interior. The van Stockum dust cylinder, the Tipler cylinder, the Gott cosmic string solution — all are exact solutions of general relativity that permit causal loops. None have been observed. None have been ruled out on mathematical grounds. The formalism permits them. The physicists ignore them. The disagreement between the theory and its interpreters has not been resolved in seventy-seven years.

In each of these cases, the formalism was more general than the intuition that produced it. Einstein supplied the axioms of general relativity; the axioms supplied black holes and time travel. Dirac wrote an equation for the electron; the equation wrote antimatter. Maxwell fixed an inconsistency; the fix predicted light.

The physicist's rejection of an unwanted solution is not the same kind of statement as the solution itself. The solution is a theorem: it follows necessarily from stated premises. The rejection is a judgment about which theorems correspond to reality. Schwarzschild's singularity is as certain as Schwarzschild's geometry. Einstein's objection was not that the mathematics was wrong but that nature would not permit the mathematics to apply. He was right about the mathematics and wrong about nature.

The formalism does not argue. It does not insist. It simply contains what it contains. The unwanted solution sits in the equation with the same logical standing as the wanted ones, waiting for someone to look where the author refused to.