by Bill Cheswick

He pedaled quickly away from his surprisingly cheap Victorian house with strong, idealistic post-doc legs. The car that would spread his waist and speed his trip was still several years in the future.

Besides, he liked to pedal: it gave him time to think about the project that had come to him a week before: a little, unofficial experiment.

Subatomic reactions don't seem to care which way time is flowing: they are perfectly reversible. The positron, the antimatter twin of the electron, acts just like an electron going back through time. What if it actually does go backward through time? If it walks and quacks like a duck, maybe it is a duck.

Is a positron just an electron in a little time machine? Time machines aren't very popular research subjects. If you write a paper about making a real time machine, the press gets wind and bothers you with stupid questions. In fact, the physics community has adopted code words when writing about such things.

He wasn't thinking about the popular idea of a time machine. He couldn't build something to go back and kill Hitler in 1922, or listen to Beethoven conduct one of his symphonies, or chat with Leonardo about the design of a steam engine.

But maybe an electron could reach a few femtoseconds into the past and kill it's own "grandfather", the reaction that created it. Nothing spectacular: just a little lab experiment.

He came up with a reaction that might work, assuming that a positron is a backward traveling electron. It was a complicated reaction, quite improbable, and all of it a cheat, a rare dance of virtual particles that never happens long enough to really count, hiding behind Heisenberg's kludge. It needed the power of a beam-line to make it real.

As he cycled along the Illinois highway he smiled at the arrangement. A small accelerator, a laser, and some other stuff borrowed from various labs while the accelerator was down. Most of the physicists were on vacation, or visiting Switzerland, or putting their work on a strong theoretical footing while the beam was retuned and adjusted. It would take a couple more weeks. The beam would be on, but too wide, or badly calibrated---useless for the usual experiments. No one would begrudge him the beam time, and his regular research was on hold. This was vacation.

He turned onto the long country road, under the power lines that marched into the distance. The locals didn't really trust this place, vaguely mistrusting assurances that it couldn't go up in a nuclear explosion or some inexplicable technological mishap. They do stuff with atoms there!

Back in the early days of the Manhattan Project, a panel of scientists met to try to figure out whether something really bad would happen if we set off an atom bomb. What would high-energy particles do to the atmosphere? Would a nuclear explosion set off a chain reaction of the ocean's hydrogen, converting the whole globe to a vast hydrogen bomb?

The experts decided that nature had already run the experiments. If the ocean was susceptible to this, cosmic rays would already have set it off. Nature has already tried everything possible. In fact, there is a rule in physics and cosmology (and they are the same subject) that if something is possible, nature has already done it. Black holes and neutron stars were deemed possible, and later discovered. There are even natural masers in cosmic gas clouds.

As he cycled past the herd of bison (can it be dangerous if they have herds of bison there?) he wondered: has nature tried my experiment? There was plenty of energy in the first tenth of a second after the big bang, but the mix was all wrong. No positrons, either.

Later, the conditions are a little better, but the energy is missing. Black holes emit streams of positrons (attract hordes of electrons going back in time?!), but the energy is rare: a few monster cosmic rays, and the arrangement is still wrong. Besides, even when the conditions are right the chance is one in a billion that an electron will kill its grandfather.

He propped his bike near a convenient entrance. There is no need to lock it in these parts. As he hustled up the steps of the glassy modern building, he worked out the probabilities. His estimates were wilder than any grade school student's guess: this many galaxies, so much antimatter, the universe weighs about this much.

Even with generous guesses, it looked extremely improbable that his experiment had ever happened in the natural universe. (No speculation about alien physicists allowed.) Such a grand thought is not unprecedented: for a few tens of nanoseconds, the heart of an exploding atom bomb is thought to be the hottest place in the entire universe. The light of the flash itself has the density of lead in the bomb's interior!

He strode down to the borrowed beam line. These things fire a hair's width beam of intensely energetic particles at a target. The target is surrounded by a massive and expensive detector that tastes the fragments of the impacts.

He had no detector. In fact, he had no idea how to tell if his tiny time machine would work. But this was vacation time. He'd make the motions, and let her rip for a week or so. If he was right, there were better than even odds that at least one event would occur.

The beam was switched on later that day. Two days later, at 3:41 AM Central Standard Time, for the first time ever, an electron traveled back in time and suppressed its own creation, the first time paradox.

The universe crashed with a General Protection Fault.