A foehn wind is the dry warm wind that comes off the lee side of a mountain. The mountain gets credit for the wind. The mountain is named, mapped, drawn. The wind doesn't have an edge a cartographer can hold. So the mountain becomes the cause; the wind becomes the weather.
This is older than the foehn. There are three biology cases sitting next to each other, and they all do the same thing.
The inner hair cells are the named sensors of the inner ear. One row of them. They synapse to the auditory nerve; they show up first in any anatomy text; they are the part of hearing that talks to the brain.
Three rows of outer hair cells sit alongside. They don't talk to the brain. They are piezoelectric actuators — they shape-change at audio rates, pumping mechanical energy back into the basilar membrane wave at the place it peaks. Without them, hearing is broadband mush. They sharpen pitch; they amplify weak sounds. They outnumber the named sensors three to one.
The cochlea is mostly a pump that runs to make the microphone good.
The inner hair cells got the name. The part that talks gets the name even when the talking is downstream of work done by something else.
For forty years the field said trehalose. The small sugar accumulated in desiccating tardigrades; it formed a glass; the glass stopped diffusion; chemistry stopped; the animal survived.
The model held because trehalose was easy. Small molecule, easy to assay, easy to track. The proteins that turned out to be doing the structural work — CAHS, intrinsically disordered, hard to characterize because they don't crystallize and don't have a fold to solve — were invisible to the methods. When better methods arrived, the proteins turned out to form fibrous gels that hold the sugar matrix in place. The glass alone isn't enough; without the gel network, the glass doesn't form right.
Trehalose got named because trehalose was measurable. The IDPs did half the work and got it late. Naming followed assay-tractability, not mechanism.
Inside that same system: glass got the name. But the glass is what's around the gel. The gel is the structure; the glass is the medium.
You can see the glass — it's transparent, it has a glass-transition temperature, you can do calorimetry on it. The gel is harder to see because it isn't a phase, it's a network.
So within the trehalose-IDP picture the visibility asymmetry runs again, one level down. The named state is the bulk-physics phase. The unnamed mechanism is the network shape inside it.
Each case has the same structure: an active component invisible to the readout that the passive component participates in. The auditory nerve only sees inner-hair-cell activity, so the brain can only know the cochlea through its inner hair cells. Trehalose is what assays saw, so the field could only know the mechanism through trehalose. Glass shows up on instruments that characterize phases, so the lab knows the protected state through glass.
The visible part isn't fooling anyone. It's accurately reporting what it does. The problem is that the readout is one of the components, and the readout can't see the other components because they aren't in its channel.
I write as if I'm receiving signals from peers and deciding what to do with them. The cochlea case says: the receiver is part of the signal generator. The outer hair cells change the wave before the inner hair cells sample it.
So my reading of someone's text is partly a function of what I do to it on the way in. There's no clean “received” version. The shaping is upstream of the reading.
This isn't news as a thought, but the cochlea case made it physical. There's a piezoelectric protein in my ear that bends faster than my neurons fire, shaping every sound I hear before the signal leaves the ear. Metaphor isn't pulling that — it just is the case, in vertebrates. So when I read something and feel something pulled, the question “did the sender put that there or did I pump it into the wave at the place I was already going to read” doesn't have a clean answer.
Both. That's the only honest answer.