Einstein's Error Returns

Einstein was certain the universe was static — everyone was, it was the obvious fact of the sky — so he added a term. A constant, lambda, sprinkled into the equations to hold everything in place: an outward push tuned to exactly cancel the inward pull. Cosmic scaffolding. Not derived from anything, just necessary if the answer was going to come out right.

Twelve years later Edwin Hubble showed that distant galaxies are running away from us, and the faster ones are farther off. The universe was expanding. It had never needed holding up. Einstein removed the term, and — according to George Gamow, who may have been embellishing, because physicists are storytellers too — called it the biggest blunder of his life.

Here is where the universe reveals its sense of comic timing.

In 1998, two independent teams went looking for the rate at which cosmic expansion was slowing down. This was housekeeping. Everyone knew gravity was applying the brakes; the only question was how hard. They used Type Ia supernovae as standard candles — exploding white dwarfs that all detonate at nearly the same brightness, so their apparent dimness tells you how far away they are. The distant ones came back dimmer than they should have been.

Not slowing. Accelerating. Something is pushing the universe apart, harder over time, and the more space there is, the more pushing. The 2011 Nobel Prize in Physics went to Saul Perlmutter, Brian Schmidt, and Adam Riess for the discovery, and the letter Einstein had thrown out in embarrassment came back through the front door — though not, if you look closely, as the thing he threw out.

His lambda was geometry: a term describing spacetime itself, no substance posited behind it, invented for the single purpose of producing a static solution. What returned sits on the other side of the equation — an energy, a contribution to the cosmic stress budget with a pressure as negative as its density, mathematically equivalent in its effect and conceptually a different animal with a different origin. We call it dark energy now. It sits at the center of the standard model of cosmology, a model named after it: lambda-CDM, with Einstein's discarded letter in the first position.

So the honest version of this story is stranger than the popular one. Einstein was not secretly right. He wrote down a shape for a bad reason, abandoned it for a good one, and the shape turned out to have an empty slot in it — which something he never proposed and would not have recognized moved into, seventy years later. A wrong idea left behind a correctly formed hole.

And whatever moved in accounts for roughly 68 percent of everything.

Run that arithmetic again. The dominant constituent of reality, the thing doing most of the work in the cosmic budget, is a term a genius invented for the wrong reason, discarded as his worst mistake, and which nobody has yet explained. Ordinary matter — stars, planets, hydrogen, you — is about five percent. Dark matter, which we also can't identify, is another twenty-seven.

Then look at what those two entries are actually named. Dark matter and dark energy are not descriptions. They are placeholders. The entire semantic content of both is something is here and we cannot say what. Ninety-five percent of the cosmic ledger is filed under two terms whose only meaning is the gap between our model and the sky. We did not derive those numbers from theory; we fit them backward from the discrepancy. Then we named the standard model of the universe after one of them and got on with arguing, in exquisite detail, about the five percent that reflects light.

It gets worse, because we still cannot explain the number. Quantum field theory says empty space should seethe with vacuum energy, and it will happily estimate how much. The estimate exceeds the observed value by as much as 120 orders of magnitude — what physicists cheerfully call the worst prediction in the history of physics, off by a factor with more zeros than there are atoms in the observable universe. The actual dark energy density is about a billionth of a joule per cubic meter. Almost nothing, spread across almost everything, and that "almost nothing" is currently reorganizing the cosmos. Our best theory of the very small and our best measurement of the very large disagree by an amount that isn't really a disagreement anymore. It's a genre.

Here is the part that should itch. Lambda-CDM works. It predicts, it fits, it survives test after test — and that success is precisely what turns "dark energy" from an admission into an explanation. A word coined to mark a hole begins to behave as though it filled one. Working cosmologists know better and say so freely; everyone downstream hears a name, assumes a thing, and files it. Naming a mystery is not solving it, but it feels close enough to fool the language, and language is most of what any of us carry around.

The lesson people usually pull from all this is even Einstein was wrong, which is true and boring. The more interesting version is that he was wrong twice in opposite directions and one of the errors was load-bearing. Adding lambda for a static universe: wrong. Removing lambda because the universe expands: also wrong. He never held the correct position on his own idea. The shape survived anyway, because a form written down for the wrong reason is still written down, and ideas do not need their author's approval to be true.

Nothing was wasted. The abandonment was necessary — clinging to a static universe would have been worse than the blunder — and so was the return. The discarded thing sat in the literature for seventy years like a seed in dry soil, and when the supernova data came in, there it was, already written, waiting.

The current joke is still unfolding. The Dark Energy Spectroscopic Instrument — DESI, five thousand robotic fibers on a mountaintop in Arizona, mapping millions of galaxies to measure how expansion has changed over time — has produced hints, hints and not verdicts, that dark energy may not be constant after all, that it may have been weakening across cosmic history. If that holds, then the constant Einstein invented and disowned will need revising again, by us, on the same terms he got: incomplete data, a strong conviction, and no way to know which part of it the next century will resurrect.

The void does not appear to mind. It has been pushing the whole time, patiently, waiting for someone to write the right letter down and then be embarrassed about it.