# n174 — the fifth critique: the junction claim retracted to its
#         true size
# (2026-07-11. The most destructive reading yet, and mostly right.)

ACCEPTED:
1. The only DYNAMICALLY CALCULATED radiative sector is the dead one
   (headstone 34, excluded by pulsars at x423). Correct, and now
   said at headline level, not ledger level.
2. "GR's radiation content derived" RETRACTED. What n166 establishes
   is a polarization-CONTENT condition on the audit-dark sector.
   The propagator, positive norm, helicity representations, source
   coupling, factor four, and quadrupole amplitude are six unbuilt
   constructions. Weinberg-Deser constrain a built structure; they
   do not build it. Footers, status, paper III, prediction 17
   all corrected; paper III downgraded to DRAFT, not submission-ready.
3. Discovery order stated plainly (in paper III itself): the
   momentum budgets were adopted after the data kill. Their claim to
   "derived" rests on (a) the conservation-law derivation and (b)
   prediction 31 - a consequence they were NOT fitted to. If 31
   fails, the repair was curve fitting; the record commits to
   saying so.
4. alpha: the word "derived" applies to the VERTEX (integer q) only;
   the coupling normalization is open and off by x45 at first pass.

THE ANSWER TO "WHAT MUST BE CORRECTED IN THE THEORY":
Six constructions, in dependency order -
  C1. the dynamical propagator of the dark pair (second-order
      dynamics on the constraint surface; positive norm or death)
  C2. the helicity-(+-2) representation proof under the emergent
      Lorentz action
  C3. the source coupling and its normalization (one G)
  C4. the gravitomagnetic factor four (LARES-2 is the judge)
  C5. the quadrupole amplitude (the double pulsar at 1.3e-4 is
      the judge - the same executioner that killed headstone 34)
  C6. only then: Weinberg-Deser, legitimately
plus the independent test that keeps the repair honest: prediction
31's momentum channel. This is the theory's actual to-do list; the
polarization count was the doorway, not the room.