Eight small moons sit inside Saturn's rings. Every one of them is a bearing between two walls running opposite ways. Three carry a shape found nowhere else in the solar system — and not one of the three has ever had its spin measured.
Generating premise: a body in a ring is not orbiting alone — it is wedged between two flows moving opposite ways past it. Nearer material orbits faster and overtakes; further material orbits slower and falls behind. That is a bearing between two belts.
Kepler makes the inner wall overtake and the outer wall fall back. Both push the same way round.
Take the orbit anticlockwise. Material at smaller radius is faster, so in the moon's own frame it streams forward along its inner flank. Material at larger radius is slower, so it streams backward along the outer flank. A force in +ŷ applied at −x̂, and a force in −ŷ applied at +x̂, give the same torque: −ẑ.
So the shear does not cancel. It adds. Every ring-embedded body sits in a standing couple that wants to turn it backwards relative to its own orbit.
Not a metaphor. Two opposed velocity fields in permanent contact with one body, at every instant of its life.
A tidal effect. Tides act along the Saturn line; this couple acts about the vertical axis and has the opposite sign.
Generating premise: the shear drives the body backwards and the tide drives it forwards, so its spin state is a contest. For the Moon the contest was settled long ago — one face, always. For a ring moon it has never been measured at all.
A body that rolled would show every face. A locked body shows one. Locking is the roll cancelled — exactly one counter-turn per orbit.
Tidal locking drives prograde spin at one turn per orbit. The Kepler shear couple drives retrograde. They are not competing for magnitude; they are competing for sign. Whichever wins is written in the body's rotation, and rotation is measurable.
For Pan, Daphnis and Atlas it has never been measured. Their synchronous rotation is assumed, on the reasonable ground that locking timescales are short for small close bodies. Assumed is not measured, and a ledger fed an assumption returns it.
Feed T_spin ≡ T_orb into any exchange ledger and it returns unity because unity went in. That is arithmetic, not evidence.
A spin sense. Not a period, a sign. Prograde settles for the tide; retrograde settles for the seat.
Generating premise: Pan and Atlas carry equatorial skirts, and the skirt lies in the ring plane — which is also the plane perpendicular to the shear axis. Orientation alone cannot tell accretion from rolling. Profile can.
Atlas is 1.91 times wider across its waist than pole to pole. Nothing else in the solar system is shaped like that.
But the obvious metric is a trap, and it caught me while building this page. R_eq/R_pol does not measure a ridge — it conflates a skirt with a sausage. Two bodies outrank Atlas on it and neither has a skirt: Aegaeon at 3.50 and Prometheus at 2.28, against Atlas's 2.21. Both are simply elongated along the orbit.
The metric that separates them is R_mid/R_pol. A prolate body is long in one axis and round in the other two (a≫b≈c). A skirted body is round in two and flat in the third (a≈b≫c). Score the middle axis against the pole and the two morphologies come apart.
Atlas and Pan still stand clear of the pack. The two deepest in dense ring material are the two with skirts.
Daphnis drops into the crowd at 1.28. Its skirt is marginal, and any three-body story built on it was built on the wrong axis ratio.
Generating premise: an inclined moon imprints the same vertical phase on both gap edges; an encounter imprints opposite ones. Both offsets and both relative velocities reverse between the inner and outer edge, so the circulation sense goes as their product — and flips.
The wavelength is not free. Kepler shear alone gives λ = 3π·Δa — about 190 km at 20 km offset, against roughly 200 km measured.
The heights are honest too: they come off a ruler. At the low solar elevation of the 2017 ring-grazing passes, Daphnis and the crests throw long shadows, and shadow length times the tangent of the sun angle is the relief. About a kilometre. No model in the chain.
The inclination is a different animal. It is fitted — quite possibly fitted from the very waves it is then invoked to explain. That is the circularity this investigation must not inherit, and the parity test sidesteps it completely: it asks only whether the two edges agree or oppose, which no fitted parameter can absorb.
Wavelength against offset. Vertical relief from shadow length. Both model-free, both already in the Cassini frames.
The 0.0036° inclination. Until its provenance is traced it may not enter the chain — only an INFORMATIONAL-RIVAL column.
Generating premise: depth does not explain the skirts. It would be convenient if it did — koppa is the framework's own currency. It does not, and the phase that shows this is a pass, not a failure.
Saturn's c-boundary is Ϟ = 0.4221 m. Every moon's depth is z = Ϟ/a — and across the eight it slides smoothly by a quarter, with no break anywhere.
Pan sits at 3.16 × 10⁻⁹, Aegaeon at 2.52 × 10⁻⁹, and the six between them fall on a smooth line. The skirts, meanwhile, stop dead somewhere near 138,000 km. A smooth quantity cannot draw a sharp boundary.
So depth is excluded as the ridge criterion, in the framework's own currency, before anyone gets attached to it. That is what a pre-registered null is for — it spends the attractive explanation early, in public, at no cost.
Nothing. A null declared in advance is information; a null discovered after a claim is a retraction.
Contact with dense ring material, which is not a depth and not in the chain — so it enters as rival information only.
Generating premise: the whole question reduces to the sense of one rotation. Not a period — a sign. And it has never been taken for any of the three bodies where it matters.
Cassini is gone. But the frames are not, and neither are the propeller moonlets, the next mission's targets, or the archive nobody has read this way.
Two predictions, stated before any data is touched: under tidal lock, the body turns prograde once per orbit and keeps one face to Saturn. Under a shear-won seat, it turns retrograde, and the face presented to Saturn cycles. Those are opposite signs on a directly imageable quantity.
What this page is: a specification. Nothing here is claimed. Act Five is a declared null, Act Three is degenerate with accretion until Act Four resolves, and Act Four has not been run. The value is that every one of those states is written down before the numbers arrive.
SPEC. Not a result, not a claim, not evidence. Six phases with numeric gates and six falsifiers, published before execution.
If both gap edges carry a common vertical phase, the seat mechanism loses its only imaging test and this page says so.