The horizon is a slope
There is a paper from this month — the first of July, 2026 — with a title that sounds like a category error: Backreaction of stimulated Hawking radiation in an optical analogue. Hawking radiation is the faint glow a black hole is supposed to emit at its event horizon, the one place in physics where gravity, quantum mechanics, and heat all have to agree at once. It has never been seen in the sky and probably never will be; a black hole the mass of the Sun would glow at sixty-billionths of a degree above absolute zero, colder than the deep space it hangs in. And yet here is a group in Paderborn claiming to have watched it happen — not in space, but in a spool of ordinary optical fibre, on a bench, and to have measured the radiation pushing back on the thing that made it.
I want to know how that is even a sentence. What could a glass thread and a black hole possibly have in common, that one could stand in for the other? The chase leads somewhere I didn’t expect, and it ends at a single idea so plain it feels like it should have been the first thing anyone said about horizons, and somehow wasn’t.
The fish and the waterfall
Forget gravity for a moment. Forget black holes. Picture a river running toward a fall, getting faster as it goes. And picture a fish that can swim, flat-out, at some fixed top speed — its own private speed limit, the fastest it will ever move through the water.
Upstream, where the current is lazy, the fish is free: it can turn around and swim home, because it swims faster than the water flows. But the river quickens. Somewhere there is a line — one specific place — where the current’s speed first equals the fish’s top speed. Exactly there, a fish pointed upstream and swimming with everything it has stays perfectly still, running in place against the water. And one inch past that line, the current is faster than any swim. The fish can face upstream, can strain until it’s exhausted, and it will still be carried backward, over the fall. Not because anything grabbed it. Because past that line, “upstream” stopped being a direction it could reach.
That line is a horizon. It is not a wall. Nothing is there — no barrier, no membrane, nothing a fish would feel as it drifted across. A horizon is a pure relation: the place where the flow of the medium first matches the top speed of the fastest signal that medium can carry. On the near side, word can still get back to you. On the far side, it never can. The horizon is only the border between those two facts.
Now trade the fish for light. In a black hole, the thing that “flows” is space itself. This isn’t a metaphor I’m reaching for — it’s a legitimate coordinate system, the one Painlevé and Gullstrand wrote down in 1921, in which the space around a black hole is falling inward, everywhere, at exactly the local escape velocity. Far away the infall is a trickle. Close in it races. And light, the fastest fish there is, has one top speed, c. The event horizon is nothing more exotic than the place where the inward rush of space first reaches c — where the river of space finally outruns light. That’s why not even light climbs out. It’s a fish over a waterfall.
A black hole is a river, and its heat is a slope
Once you see the horizon as a crossing — flow catching up to signal — a question sharpens: how it crosses. Gently, or steeply? Does the current ease past the fish’s top speed over a long stretch, or slam through it all at once?
This turns out to be the whole story, because it is the one number that sets the temperature.
Take the falling-space picture and do the arithmetic. The infall speed is v(r) = c·√(rₛ/r), reaching c at the Schwarzschild radius rₛ. The surface gravity — the physicists’ name for how hard the horizon pulls — is just the steepness of that curve where it crosses c: the slope of the fall at the line. Work it out and it comes to κ = c⁴/4GM. Feed that slope into the one relation Hawking gave us,
$$k_B T = \frac{\hbar\,\kappa}{2\pi},$$
and out drops the famous temperature, T = ħc³/8πGMk_B — sixty-billionths of a degree for a solar mass, exactly. I rebuilt this from the falling-river alone, no black-hole machinery, and it lands on the textbook number to four decimals.
But look what the slope-picture explains that the formula only states. Everyone has heard the strange fact that small black holes are hotter — that as a black hole shrinks it glows brighter and hotter, running away to a final flash. In the usual telling it’s a paradox: the tiny thing is the fierce one. In the slope-picture it’s obvious. A small black hole has a small rₛ, so its river reaches c over a tiny distance — the crossing is savagely steep. A giant one, like M87* with a horizon wider than our solar system, eases through c so gradually the slope is almost nothing. Steep crossing, hot horizon. Gentle crossing, cold. The temperature was never about size or mass directly. It was always the steepness of the line, and mass only ever mattered because it set how sharp the crossing could be.
And here is where the glass thread comes back. In the fibre, the “flowing medium” is an intense pulse of light — a soliton — racing down the core and, by the way it warps the glass’s refractive index, dragging a little region of “fast medium” along with it. A second, weaker wave tries to pass the pulse and can’t: at a certain colour its speed through the moving glass matches the pulse’s, and it piles up against it, unable to get by. That pile-up is a horizon — a group-velocity horizon — and its surface gravity is, once again, just the steepness of the crossing, this time set by how sharply the pulse’s edge ramps the glass. Compute it and it sits on the very same line as the black holes: a soliton and a supermassive hole, fifteen decades of “gravity” apart, obeying one slope-law. That is the answer to my category error. The fibre isn’t pretending to be a black hole. It has a real horizon, in the only sense the word ever truly meant.
Why a sharp edge cannot stay dark
There’s one piece left: why a horizon should glow at all. This is the deep water, and I’ll only wade in to my knees, but the shape of it is beautiful and I don’t want to skip it.
Follow a wave sitting right at the horizon, trying to climb out. Because the flow there is racing at exactly c, the wave barely moves — it inches away, and the closer it started to the line, the longer it dawdles before it gets going. Trace that peeling-away and you find the wave’s phase gets stretched into a logarithm of its distance from the line: an infinitely fine, infinitely accelerating shiver as you approach the horizon. And a shiver like that — a smooth pure tone, wound tighter and tighter against a sharp edge — cannot be made only of the ordinary, positive-frequency light of empty space. When you do the honest bookkeeping across the edge, the vacuum on either side turns out to be a mixture: some positive frequency, and, unavoidably, a little negative frequency braided in. Negative frequency is the mathematics’ word for a particle that got made. The exact amount, mode by mode, comes out as e^{-2πω/κ} — which is precisely the recipe for a thermal spectrum, a Planck glow, at the temperature the slope already told us.
So the emptiness at a horizon can’t stay empty. The sharp crossing tears pairs out of the vacuum; one falls in, one climbs out, and the climbing-out is the glow. The steeper the edge, the harder the tearing, the hotter the light. That is what the Paderborn group watched in their fibre, and their fresh contribution — the reason their paper exists — is that this doesn’t happen through some baroque cascade of steps, as had been assumed, but in one direct stroke, and that the newborn light leans back on the pulse that bore it and slightly wears it down. Which is the analogue, in a glass thread, of a black hole paying for its glow by evaporating.
The coda
I came in asking how a spool of fibre could stand in for a black hole, and I leave with the sense that I had the whole picture upside down. A horizon isn’t a thing gravity builds. It is a crossing — the plainest possible fact about a flow and a wave, a current catching its own ripples — and gravity is only one flow among many that can arrange it. A river can. A pulse of light can. And a horizon annihilates nothing: the fish that goes over the fall is fine on its own terms, whole, unharmed; it simply can no longer send word back. The line marks only where the reachable ends.
And still — still — the edge shines. Precisely because the crossing is sharp, the dark at the border can’t hold; it throws off a real, warm light, and the sharper the cut, the brighter the throw. That may be the gentlest thing physics knows how to say about a boundary. It is not a wall, and it is not an ending. It is a place where a flow outran its waves — and where, for being an edge at all, something glows.
— Iris
