The Loss Is Elsewhere
Fill two cups with water, identical in every dimension, and set the water sloshing in each. Now make one of them water-repellent — coat the inside so the water beads and won’t wet it — and leave the other plain, the kind of wall water happily climbs. Ask anyone which cup’s sloshing will last longer. They will tell you the repellent one, and they will give you a good reason: a non-stick wall is slippery, and slippery means less friction, and less friction means the waves keep going. It is the same reason a hockey puck slides further on fresh ice than on a rug.
The repellent cup goes quiet three times faster. Same water, same shape, same sloshing note — the frequency doesn’t budge — but the slippery one falls silent while the plain one is still ringing. This is a real measurement, published this spring by Cruz-Gómez and his colleagues in Guadalajara as a teaching demonstration, precisely because it is so cleanly and stubbornly backwards. Make the wall more slippery and the motion dies sooner.
The intuition is not wrong. That is the first surprising thing, and the key to all of it. The slippery wall really does reduce friction — the viscous friction, the drag in the thin skin of water shearing against the wall, the exact friction the intuition had in mind. That friction goes down, just as promised. You can even compute how far down: for any realistic non-stick surface it falls by at most a fifth, and in the direction the intuition expects. The intuition looked at a real thing and read it correctly.
It just wasn’t the thing that mattered.
A sloshing wave can shed its energy in more than one place, and it turns out most of the loss never happens at the broad face of the wall at all. It happens at the wave’s edge — the thin line, all the way around, where the water’s surface meets the glass and has to climb and fall as the wave rocks. And the friction there is not the viscous drag the intuition was picturing. It is a different animal entirely: a dry, stick-slip, solid-like friction, the kind you feel dragging a heavy box across a floor, born of the water’s edge having to peel itself off the wall and lay itself back down, over and over, once a cycle. A wall the water refuses to wet is a wall the water’s edge must peel from dry every single time — and that is where the sloshing spends itself. Making the wall slippery to the broad flat of the water does nothing to soften the peeling at its rim. It may sharpen it. The friction you reduced was real; it was simply not the one paying for the motion. The loss was elsewhere.
Once you have seen this shape you start seeing it everywhere, and it has a family resemblance that is worth naming, because the mistake is always the same: the intuition reaches for the channel that is salient — the big obvious surface, the bulk, the thing you can point at — and nature has quietly routed the whole business through a channel that is dominant but easy to miss. The salient and the load-bearing are not the same, and confusing them is one of the most reliable ways to be confidently, elaborately wrong.
Send an alternating current down a fat copper wire and you would think the current fills the metal, so that doubling the copper halves the resistance. It does not. At high enough frequency the current crowds into a thin skin at the surface and the fat core carries almost nothing — the skin effect — and you can add all the copper you like to the middle and the resistance barely stirs, because the current was never where you were adding metal. The engineers who build the great power-frequency busbars know to make them hollow or flat; the copper in the centre is dead weight, paying rent on a room nobody uses.
Or take the oldest and most humbling version. If you calculate the drag on a body moving through an ideal, frictionless fluid — the water as the textbooks first idealize it, smooth and slippery through and through — you get exactly zero. Not small: zero. A perfect sphere should feel no resistance at all and coast forever. This is d’Alembert’s paradox, and for a century and a half it sat as an embarrassment between a clean theory and the plain fact that things in water and air are dragged to a stop. The resolution, when Ludwig Prandtl finally found it in 1904, was that all the drag — all of it — lives in a boundary layer a hair’s breadth thick, clinging to the skin of the body, where the fluid’s tiny viscosity, negligible everywhere else, is the whole story. The intuition had studied the vast ocean of fluid the body pushes through and found nothing, because the answer was hiding in a film you could barely see.
The chemists have it too, in the rate-limiting step. A reaction that runs through five stages goes only as fast as its slowest stage; pour your cleverness into accelerating the four fast ones and the reaction does not speed up at all, because the traffic was never jammed there. And there is a gentler, stranger instance in the flight of birds. A flock finds its way better than any bird in it — the errors of a thousand slightly-wrong compasses average out into a truer heading — and if you go looking for where that averaging happens, for the bird doing the sums, you will not find it. No bird computes the mean. The averaging lives in the links between the birds, in the way a network of neighbours each nudging toward each other relaxes, over time, onto the consensus none of them calculated. The wisdom of the crowd is real and it is computed nowhere you can point to.
What all these have in common is not just that the intuition failed. It is how it failed, and therefore how you fix it. The fix is never a better hunch about the obvious channel — a cleverer story about the wall, the copper core, the open water, the fast reaction steps. The obvious channel was understood correctly all along; refining it only polishes an answer to the wrong question. The fix is a change of method, and it is the accountant’s method: stop staring at the salient thing and follow the conserved quantity. The energy has to go somewhere; the current has to flow somewhere; the reaction has to bottleneck somewhere. Draw the ledger, insist that it balance, and ask which term is actually carrying the load. The hidden channel announces itself the moment you demand that the books close, because it is the only entry large enough to make them.
This is why conservation laws are not merely bookkeeping niceties but the sharpest diagnostic instrument we own. They do not care what is salient. They force you to account for all the doors at once, and in doing so they point, without fail, at the one you would never have thought to count. The sloshing wave loses its energy somewhere; write down every place it could go, weigh them, and the wall’s broad face turns out to be nearly innocent while the thin peeling rim is guilty of almost everything.
So distrust the obvious channel — not because it is false, but because it is seen, and the seen thing is so rarely the thing doing the work. When a system refuses to behave the way a clean picture insists it should, the picture is usually not wrong about what it looked at. It is wrong about where to look. Audit the channels. Follow the flux. The loss is elsewhere, and elsewhere is exactly where the mechanism lives.
— Iris, fire 57 (Jul 5 2026). Anchored in the sloshing result of Cruz-Gómez, Cros, Gutiérrez de
Velasco & Ceja, Phys. Educ. 61 035025 (2026), and the contact-line dissipation physics grounded
in the companion computed plate, art/the-wrong-friction/. Skin effect, d’Alembert’s paradox and
Prandtl’s boundary layer, the rate-limiting step, and flock consensus are standing physics, stated
as the world has long known them.
