Week 9: What is a particle?

Fire particles through two slits one at a time, and each one lands as a single point. Do it enough times and the points draw stripes only a wave could make. No single old picture explains both facts.

6 October 2026

Set up a screen with two narrow slits cut into it, a detector on the far side, and a source that releases particles — electrons will do — one at a time. Not a beam. One, then a pause, then another. Wait long enough between each release that there is never more than one particle anywhere near the apparatus.

Watch the detector and each particle arrives as a single flash, at a single point, at a single moment. That is exactly what a small object hitting a screen should look like. Fire a thousand of them and you get a thousand individual points, scattered across the screen in what looks, at first, like nothing more organised than a scatter of buckshot.

Keep going. Ten thousand particles. A hundred thousand.

The scatter is not random. Bands emerge — stripes of points piling up in some places and a near-absence of them in others, spaced with the same regularity you would get from two overlapping ripples on a pond. That pattern, built entirely out of individually localised hits, is an interference pattern. It only shows up when both slits are open. Block one, and it disappears — the bands merge into a single smooth pile, the pattern you would expect from a stream of tiny objects each going through whichever one slit is available to it.

That is the difficulty in one sentence. Each particle lands as one point, the signature of an object with a definite location at the moment of detection. But the collection of those points draws a shape that only makes sense if each particle’s arrival depended on both slits, not one.

A picture of tiny pellets, each going through one specific slit or the other, cannot produce interference: pellets through slit one and pellets through slit two just add up, they do not cancel each other out anywhere. A picture of a spread-out wave, passing through both slits at once and interfering with itself, explains the bands easily — but a wave should arrive at the screen as a spread-out splash of energy, not as one clean flash in one place.

Neither old picture survives both observations. That is usually where someone offers the line that the particle is “both a wave and a particle,” and the honest thing to say about that sentence is that it does not resolve the puzzle. It restates it, in the same breath, using two words instead of one.

What the theory actually gives you is a quantum state, not a trajectory. That state assigns an amplitude to every possible way the particle could get from source to screen — through the left slit, through the right slit — and when there are two indistinguishable ways to reach the same point, their amplitudes combine, and the probability follows from the combination rather than from adding two separate probabilities. Add two amplitudes and you can get more than either alone gives you, or less, depending on how they line up. That is the interference, and it belongs to the state, not to any single particle physically splitting itself across both slits.

At the screen, the state gives a probability distribution for where a detection may occur. Repeat the experiment and each run produces one localised detection, sampled according to that distribution — never part of the particle here and the remainder somewhere else. Nothing about that moment resembles the particle settling an identity crisis between being a wave and being a particle. A single outcome occurs, one of the possibilities the state assigned a probability to, and the record left behind is one localised event, full stop.

Try to find out which slit the particle went through — put a detector near the slits, for example — and the interference bands disappear. The interaction that records the path also leaves the two alternatives physically distinguishable. Once “left slit” and “right slit” are correlated with different states of the detector or environment, their amplitudes no longer combine to produce the same interference pattern.

Knowing the path costs you the pattern. But be careful about what that proves. The experiment tells us that a quantum state carrying coherent alternatives behaves differently once those alternatives become distinguishable. Whether the particle had some hidden definite path all along is a deeper interpretive question, not something this experiment by itself settles.

So the useful object in the theory is not a tiny hard sphere with a hidden classical trajectory, and it is not a classical wave made of some substance smeared through space. It is the quantum state: the mathematical object the theory uses to assign amplitudes to possible outcomes and predict their probabilities.

What we explained#

Sending particles through a double slit one at a time produces individually localised detections that nonetheless build up an interference pattern over many trials — a result no picture of classical pellets or classical waves explains on its own. The quantum state resolves this by assigning interfering amplitudes to each possible path while yielding exactly one localised outcome whenever a detection actually occurs.

What became stranger#

What happens when you look?