an independent axiom is a residue — undetermined by the rest, certifiable only by exhibiting a model, resolved not by proof but by choosing among consistent completions (Euclidean / hyperbolic / spherical = flat / negative / positive curvature, the three constant-curvature homogeneous geometries)

The founding case of the residue shape, two millennia before the modern ones. Euclid's fifth is neither derivable from the other four nor refutable from them: it is a residue, undetermined by the rest, and its certificate is a model — Beltrami's disk, the sphere — never a derivation. Σ₁ from below (exhibit a geometry where it holds or fails), Π₁-blocked from above (no proof settles it): an existence-form residue on the certificate-direction axis (the EX-06 shape). The three consistent completions are the three constant-curvature homogeneous geometries, so the residue is a group-action residue in the EX-05 sense — fixed only by choosing which homogeneous space, which isometry group, you inhabit. This is the residue-of-a-group-action at its founding.

Why it is the canonical case of the reverse-engineering / reverse-mathematics distinction. For two thousand years the postulate was attacked as reverse engineering under a false designer-assumption: geometry was taken to have one intended design (the plane we see), so the fifth was assumed a theorem someone had merely failed to prove, and every attempt hunted the hidden derivation. That assumption was the error — there was no single intended design to recover, so the search could not terminate. Reverse mathematics asks the correct pair of questions instead: what is the statement independent of (which axioms, exactly), and what does each completion cost (which theorems you gain or lose by choosing it). Answered that way the postulate is not a failed theorem but a fork, and the fork is the discovery.

Calibration question: when the Synthesist says "no one has managed to derive X," check whether X is a residue being reverse-engineered under a designer-assumption. If it is, the question is not "what is the missing proof" but "what is X independent of, and what does choosing cost." The parallel postulate is where that distinction was first paid for; the continuum hypothesis, the Jarlskog sign (N-012), BB(745) (P-010), and cosmic topology are the same shape recurring.