Here is what Thomas Young did. He cut two narrow slits, side by side and very close together, into a card. He held the card in a beam of sunlight and examined the shadow it cast on a wall several feet away. He expected to see two bright bands where light came through the slits and a dark region between them. Instead he saw a pattern of alternating light and dark stripes — not two bands, but many, spread across the entire shadow region and extending into what should have been darkness. He was looking at interference fringes, and their existence proved, decisively, that light is a wave.
The year was 1801. Isaac Newton had been dead for seventy-four years. His reputation still dominated British science so completely that proposing a wave theory of light was professionally hazardous. Newton had championed a corpuscular theory — light was made of particles, "corpuscles" — and the evidence for this was not trivial. Light travels in straight lines, casts sharp shadows, and doesn't seem to bend around corners the way water waves do. Newton's optics was a monument of precision. Young was a polymath who had decoded Egyptian hieroglyphs and measured the elasticity of materials, and he was not easily intimidated. He published his results. The reception was hostile.
But the physics was unambiguous. To understand why, you need to understand what interference actually is.
A wave is a propagating oscillation — a pattern of crests and troughs moving through a medium (or, in the case of light, through the electromagnetic field). When two waves overlap, they add together, linearly and point by point. Where two crests coincide, you get a stronger crest — constructive interference. Where a crest meets a trough, they cancel — destructive interference. This is the rule, and it is a consequence of the mathematics of wave superposition.
Now consider Young's two slits. When light passes through each slit, it diffracts — it spreads out in the manner of any wave passing through an aperture smaller than its wavelength. So each slit becomes, in effect, a new source of spreading waves. The two spreading wavefronts overlap everywhere in front of the card. At points on the wall equidistant from both slits, the path lengths from the two sources are equal, so the crests arrive together, so the waves add: bright. At points where the path from one slit is half a wavelength longer than the path from the other, a crest from one arrives simultaneously with a trough from the other, so they cancel: dark. Moving across the wall, you alternate between these two conditions, and you get the striped pattern Young observed.
The width of the stripes depends on the spacing of the slits and the wavelength of light. Young used this to calculate the wavelength — the first measurement of the wavelength of visible light. The numbers he got, around five hundred nanometers for yellow light, are correct. The experiment worked as a measurement tool precisely because the fringes encode the wavelength in their spacing.
What I want to linger on is the implication that most accounts rush past. If light is a wave, and if wave interference requires both sources to be active simultaneously, then the pattern in Young's experiment requires light from both slits to be present at every point on the wall at the same time. But here is what makes this strange: when a single photon of light is sent through the apparatus — something Young could not do but we can — the fringe pattern still appears, built up one detection event at a time. Each photon lands at a single point on the wall, leaving a dot. But over thousands of photons, the dots accumulate into the interference pattern. The stripes emerge even when the photons are sent one by one, when there is no possibility of photon-photon interaction.
This is the quantum mystery of the double slit. Each photon must, in some sense, go through both slits and interfere with itself. But if you put detectors at the slits to find out which slit each photon passed through, the interference pattern disappears — you get two bands, not many stripes. The act of measuring which path was taken destroys the interference. You can have path information or interference, but not both.
Here is the point I keep wanting to make: this mystery is already there in Young's classical experiment, before quantum mechanics was invented. Young did not have a quantum theory and did not need one. His experiment showed that light forms interference patterns — which already implies that the wave must pass through both slits simultaneously, must be in some sense present in two places at once. The classical wave description already contains this non-localized character. A wave is not a particle; it does not have a definite location; it is spread through space. The spookiness is in the wave nature, not in the quantum mechanics.
What quantum mechanics added, a century later, is the photon — the fact that light, despite being a wave, is absorbed and emitted in discrete packets. This makes the mystery sharper and harder to dismiss, because you cannot simply say "the wave is spread out" and leave it there. You must confront the fact that something that arrives as a localized click — a single dot on the detector — was traveling as a delocalized wave that sampled both slits. But the delocalization itself, the inability to say "it went through this slit," is a classical wave property. Young's experiment contains the mystery. Quantum mechanics names it.
Newton's corpuscular theory predicted no interference. If light consisted of particles, each particle would go through one slit or the other, and you would get two bright bands, nothing more. The fringe pattern is decisive evidence against this, and Young knew it. What he could not have known is how strange the resolution would turn out to be — that light is both: waves in propagation, particles in absorption, with no satisfying classical picture for the transition between the two regimes. He set out to settle a question about light's nature. He settled one version of it (not particles, at least not simple classical ones) and opened a stranger one that would take another hundred years to properly formulate.
The fringes on Young's wall are still there, as easy to produce as they were in 1801. You can do the experiment with a laser pointer and a piece of aluminum foil with two slits scratched into it. The stripes appear within seconds. What they mean — what it means for a wave to arrive at a screen in a single localized click — is still not settled to everyone's satisfaction, even now. The weirdness was always there. The experiment just showed us where to look.