For seventy-two seconds in the summer of 1977, something came in loud on the quietest frequency in the sky.
It rose. It peaked. It faded out. And the curve it traced on the way through was the exact curve the sky makes when a telescope that cannot move is carried past a fixed point by the turning of the Earth.
Then it was gone. It has never come back.
Nobody was watching when it arrived. The machine wrote it down on paper, and the paper sat in a stack for days.
The stack of paper
Picture a man working through a pile of fan-fold computer printout by hand.
Jerry Ehman was a volunteer on the search program at the Ohio State University Radio Observatory. The telescope ran unattended. Its output was a continuous roll of characters, and somebody had to read the roll. That somebody, a few days after 15 August 1977, was Ehman.
Most of the page is the ordinary hiss of the universe. Blanks, ones, twos. The sky breathing.
Then one column read 6EQUJ5.
He drew a circle around it in red ballpoint and wrote a single word in the margin beside it.
Wow!
That is the whole naming ceremony. A volunteer, a red pen, and an exclamation mark. Nearly fifty years later it is still what the thing is called.
How to read six characters
The code is simple once you know it.
Each character stands for one measurement of signal strength above the background noise. The system counted 0 through 9, then kept going with letters. A was 10, B was 11, and so on up.
So 6EQUJ5 reads as 6, 14, 26, 30, 19, 5.
The samples came roughly twelve seconds apart. Six of them span about seventy-two seconds, which is close to how long a fixed point in the sky needed to drift across that telescope's beam.
Now look at the shape of those numbers. Up, up, up, peak, down, down. Not a spike. Not a burst. A smooth rise and a smooth fall, matching the antenna's own sensitivity as the target slid through the center of its view.
And the peak was thirty times the background.
The quietest frequency there is
The receiver was listening near 1420 megahertz.
That number is not arbitrary. Neutral hydrogen, the most common substance in the universe, radiates there. Every radio astronomer on Earth knows it, and the argument since Giuseppe Cocconi and Philip Morrison published their 1959 note in Nature is that anybody else with a radio telescope would know it too. It is the one channel two strangers might both think to check. International agreement keeps it clear of terrestrial broadcasting.
The receiver split its coverage into fifty narrow channels, each about 10 kilohertz wide. The signal turned up in one of them. Accounts of the exact figure differ, and the value usually quoted is 1420.4556 megahertz, though other published discussions give 1420.356.
What matters is the narrowness. Natural radio sources smear across bandwidths. Stars, gas clouds, hot dust, lightning, all of it spreads. A signal confined to a single 10 kilohertz slice is doing something that most of nature does not do.
So set it out at full strength. For seventy-two seconds, from the direction of Sagittarius, the instrument recorded a narrowband transmission on the hydrogen line, thirty times louder than the sky around it, rising and falling in the precise signature of an object beyond the atmosphere.
If you had been asked in advance to sketch what a beacon would look like on that printout, you would have sketched 6EQUJ5.
The telescope that could not turn
Big Ear was a strange machine even by the standards of radio astronomy.
Designed by John Kraus and built on land at the Perkins Observatory near Delaware, Ohio, it had no dish you could swing around. It had a flat tilted reflector at one end, a curved reflector at the other, and an aluminum ground plane between them covering an area comparable to three football fields.
It could not point. It could only wait.
You set the tilt, and the rotation of the planet dragged the sky through your field of view. The observatory turned its attention to a long-running search for artificial signals in 1973 and kept at it for years. The recording computer was an IBM 1130, with less memory than a modern phone uses to open a menu.
That is why the encounter exists only as ink on continuous paper, found late. There was no alarm, no console, no operator. Just a machine keeping a ledger in the dark.
The case against
It never happened again. That is the blunt version, and it has to be said plainly.
Ehman went back to the same patch on following nights. Nothing. Robert Gray spent years hunting for a repeat, using the META array at Harvard in the late 1980s, the Very Large Array in the mid-1990s, and a 26-meter dish in Hobart, Tasmania at the end of the decade with the astronomer Simon Ellingsen. Nothing came back.
Then there is the horn problem, which is the strongest objection anyone has raised.
Big Ear fed two receiving horns, offset from each other. A genuine source out in the sky should have crossed both, a few minutes apart, and registered twice. It registered once. Nobody has ever produced a clean account of why.
The same gap leaves the position uncertain. Because it is not known which horn caught it, there are two candidate coordinates rather than one, a few minutes apart in right ascension, in a region near the Chi Sagittarii group at a declination of roughly minus 27 degrees.
Explanations have kept arriving. In 2017 Antonio Paris proposed that hydrogen released by passing comets could account for it, naming 266P/Christensen among the candidates. Most astronomers who responded rejected the idea, on the grounds that comets are not known to emit strongly at that frequency, and that a comet would have had to cross both horns as well.
More recently, a team led by Abel Mendez at the Planetary Habitability Laboratory in Puerto Rico worked through archived Arecibo observations and suggested a cloud of hydrogen briefly brightened by a passing transient, such as a flare from a magnetar. That work was circulated in 2024 and is still being argued over.
The part that will not go away
Every proposal so far explains a signal. None of them has produced this one.
Start with the shape, because the shape is the hard part. The rise and fall matched the beam pattern of the antenna. A ground transmitter, a passing aircraft, a stray harmonic from a piece of equipment in the building does not trace that curve. It comes and goes on its own schedule, not on the schedule of the Earth's rotation. Whatever made those six characters behaved like something standing still while the planet turned underneath it.
Add the narrowness. Add the frequency, on a band the world's governments agreed to keep clear.
And add the failures. Comets, clouds, debris, interference. Each candidate has been named. Not one has ever been shown to make that mark again, on that band, in that shape.
Ehman himself has never claimed it was a message. He has spent decades pointing at the missing repeat, which is the honest thing to point at. The signal that will not resolve is a poor foundation for a conclusion in either direction.
What is left is six characters
Big Ear did not survive. It was demolished in 1998, the ground cleared to extend a golf course. The instrument that made the only record of the event no longer exists to be pointed anywhere.
So the case rests on a page.
Either something ordinary crossed that beam once, in a form nobody has yet reproduced, and the paper preserves an accident.
Or, for seventy-two seconds on a summer night in Ohio, a machine nobody was watching wrote down something that had come a very long way.
The printout has not said which.