A Gaussian sky's Fourier phases must be uniformly random, all information confined to amplitudes, yet phase-statistics analyses from Chiang-Coles through the Planck era keep finding organized phases at the largest scales: correlations among low multipoles, coupling aligned with the anomaly directions, cross-scale coherence. The familiar anomalies are this finding in projection, since aligned multipoles and modulated power are what speaking phases look like.
Gaussian randomness is the generating mechanism's signature: inflation's independent mode-stretching guarantees mute phases, with no parameter to loosen the guarantee. Every detected correlation must therefore be contamination or chance, a defense the model must maintain across every estimator simultaneously, forever.
SCT's large-scale field was not drawn from a Gaussian ensemble; it was deposited by events with geometry. A cascade of collisions lays down structure whose largest scales encode the deposition's shape, the axis, the gradient, the finite domain (P48, P44, P55), and a field with shape is precisely a field with organized phases: phase correlations are the native mathematical language of deposited rather than drawn structure. The cascade's budget applies here as everywhere, many stages drive the small-scale field toward Gaussian muteness (the central-limit machinery that keeps f_NL near 10⁻²), while the few largest stages, statistically unaveraged, leave their phases talking at exactly the low multipoles where the analyses keep hearing them (P36, P37). Weak, large-scale-confined, axis-aligned phase organization is the registered shape of the signal.
The unification is the point: the alignments, the modulation, the parity preference, and the phase correlations are one fact about one deposited field, expressed in four statistical dialects. The standard model pays four separate chance-bills; the cascade pays one geometry. Foundations in Paper 3, From Chaos To Concordance Spectra (the cascade-statistics machinery), with the imprint geometry in Paper 8, From Chaos To Constructive Relativity.
No new premises are spent: the same deposition that organizes the anomaly family is what phases measure most directly.
The registered kill: polarization-era phase statistics (LiteBIRD, CMB-S4) finding the large-angle phases fully random, no recurrence of the correlations, no axis-alignment with the temperature family, would certify the temperature-era detections as contamination and erase the deposited-field signature. Phase organization recurring but misaligned with the anomaly axes would break the one-field reading just as cleanly.