One Electron at a Time

Some experiments are strange because the apparatus is exotic. This one is strange with the plainest possible setup, which is exactly why it cannot be waved away.

Fire the electrons slowly enough and you can be sure of one thing: they leave the gun one at a time. There is a gap between each launch, long enough that the previous electron has already struck the screen and been recorded before the next one sets off. Each is a single particle, a discrete lump of charge, and when it hits the detector it makes a single dot, a tiny flash at one definite place. Nothing about that suggests anything strange. You could be watching sand grains hit a wall.

Between the gun and the screen there is a barrier with two narrow slits cut in it, side by side. An electron has to pass through. Common sense says it goes through one slit or the other, the way a thrown pebble goes through one gap in a fence, and that over many throws you should get two stripes on the screen behind, one lined up with each slit. That is what happens with pebbles. It is not what happens with electrons.

Watch the dots accumulate. The first few land here and there, scattered, no pattern you can see. A hundred dots in, still nothing obvious. But keep going, a thousand, ten thousand, and a pattern surfaces out of the randomness like a photograph developing. The dots are not piling up in two stripes. They are forming many stripes, bright bands separated by dark gaps where almost no electron ever lands. This is an interference pattern, the exact signature you get when two waves overlap and reinforce in some places and cancel in others. It is the pattern Thomas Young saw with light in 1801, the pattern that convinced physicists light was a wave. But these are not waves. They are single electrons, sent through one at a time, each arriving as a lone dot.

Sit with how odd that is. Each electron left the gun alone and hit the screen alone. There was no other electron present to interfere with. And yet the pattern that builds up is an interference pattern, which only makes sense if something passed through both slits at once and overlapped with itself on the way to the screen. A single, indivisible particle, somehow sampling both paths. The dark bands are the giveaway. They are places an electron can reach when only one slit is open, but cannot reach when both are, as though opening a second door could prevent it from arriving somewhere it used to be able to go. Adding a possibility removes an outcome. No picture of the electron as a little ball moving along a definite path can survive that.

Now do the obvious thing. If you cannot believe the electron goes through both slits, put a detector at the slits and catch it in the act. Watch which slit each electron actually takes. You can do this, and when you do, each electron is dutifully recorded passing through one slit or the other, never both. The mystery seems to dissolve. Except that the moment you start watching, the interference pattern on the screen disappears. The many stripes collapse into the two plain stripes that common sense predicted all along. Stop watching, and the interference comes back. The act of finding out which path the electron took destroys the very pattern that proved it took both.

This is not a matter of clumsy instruments bumping the electron, though that is sometimes how it is first explained. It is deeper than that. As long as the path is unknowable, even in principle, the electron behaves as though it travels as a spread-out wave of possibilities passing through both slits. The instant the path becomes knowable, recorded anywhere in the world, the electron behaves as a particle that went through one. The pattern depends not on whether you look at the screen but on whether the which-path information exists at all. Reality seems to wait to see what you are going to ask before deciding what it was doing.

Richard Feynman, who understood quantum mechanics as well as anyone, said this experiment contains the only mystery, that everything strange about the quantum world is in here and you should not try to make it go away by explaining it in familiar terms, because no one can. He was not being lazy. He was being honest about a limit. The experiment has been done with electrons, with whole atoms, with large molecules of dozens of atoms, and the result is always the same. Single objects build up an interference pattern that requires each one to have explored multiple paths, and that pattern vanishes the moment the path is pinned down.

I have described this as if it were a paradox to be resolved, but it is not, quite. It is simply how things are at small scale, confirmed ten thousand dots at a time, every time anyone checks. The electron is neither the pebble nor the wave on the pond. It is its own kind of thing, and the double slit is where that thing shows its face most plainly. You can run the experiment yourself, in a good undergraduate lab, and watch the impossible pattern assemble itself out of single particles that had no business making it. Nobody who watches it carefully comes away thinking the world is made of little balls.

The electron is neither the pebble nor the ripple on the pond. The double slit is simply where that fact stops being arguable and starts assembling itself, one dot at a time, in front of you.