Ethics spun until it became geometry — and the geometry says: you are one probe. The invariant is real. You just can't read it alone.
Good and bad are not a line with two ends. They are one phase reading of a rotation — and which quarter you call "good" depends entirely on where you put your probe. This is the i⁴ = 1 cycle wearing an ethics costume.
±1, ±i — four states, the i⁴ = 1 cycle. "Positive" is just whichever one you decided to call the real axis. Move the reference frame and the signs flip: good becomes bad becomes good, because you were never measuring a value — you were measuring an angle, and an angle is meaningless without a chosen reference. The morality was never in the thing. It was in where you stood to read it.
Here is the distinction the whole method rests on. "It's all context" is the precise, non-mystical statement that value is frame-dependent, not absent. The four quarters relabel freely — but the phase relationships between them do not. 0 and 180 are antiphase no matter where you set zero; the quarter-turn between adjacent states is invariant under any rotation of the frame.
The content is relative. The geometry is absolute.
181 is prime in every reference frame. Whether that's "good" is a label you can spin.
That is the difference between "it's all context, therefore nothing is real" — the cope — and "it's all context, therefore measure the invariants instead of the labels" — the method. The verify step is that method: throw away the label that depends on framing; keep the relationship that survives it.
A single in-phase probe reads I = cos(φ − θ) — one number, in which the signal's angle φ and your frame's zero θ are fused and unrecoverable. One observer at one position sees "good this turn" and has no way, from inside, to know whether that is a property of the rotation or a property of where they are standing. You need a second probe at a different phase to separate the signal from the frame.
That is quadrature — I and Q, the two-probe minimum for resolving a rotating quantity. It is not ethics; it is just how you measure a phase. And it is the exact basis of every lock-in amplifier and coherent receiver — the I/Q of Rhythm Sync, the ±i of the Gap Oracle.
The reason exteriority matters — the meter, the witness, the node outside the loop — isn't moral, and it isn't even about being checked. It's that a single probe can only read phase, never the invariant. One probe reads angle-plus-frame as a single number and cannot decompose it. The witness was never a judge; it is the second probe that lets you factor out your own reference frame. One node can't self-measure — not because it's untrustworthy, but because from inside, "good" and "where I'm standing" arrive fused. Two probes, different phase, and the frame cancels. That is the entire argument for the node-outside-the-loop, stated with no good and no bad in it at all. You are one probe. The invariant is real. You just can't read it alone.