the record · exhibit 07 · loop 02, Production
The chapter that was wrongly cut
The room trimmed a good chapter toward a length ceiling, twice. The trim was reverted and the rule rewritten; the overruled verdict stayed in the file, marked superseded, and the chapter carried a do-not-re-cut marker for two months.
· principle 12 of twelve
What it shows
Principle twelve, with the veto used. The critic held a chapter at revise for length alone; the trim was applied twice and left the file inconsistent. The disposition reversed it, rewrote the length rule so it could not recur, and left the overruled verdict in the file under the word superseded. The marker held for two months: when the August rework touched the same chapter, it needed the author’s explicit approval and two blind jurors.
The principles are numbered on the method page. Cuts in the excerpts below are marked with an ellipsis in brackets; the three kinds of cut are stated on the record’s front page.
From the record
2026-06-25 12:12 Restore Ch 17 to full 4,675-word version; fix length policy Restore Ch 17 to full 4,675-word version; fix length policy Reverts an over-rigid trim: Ch 17 was a fine 4,675-word chapter that I wrongly trimmed toward a 4,000 "ceiling" (twice, leaving an inconsistent file). Restored the original draft verbatim from the writer's transcript (body verified at 4,675 words; closing and protected passages intact). Policy fix so this does not recur: - Length is NOT a gate below 5500 body words; quality/engagement/continuity come first. Over-band is fine when the work holds. - Above 5500, a single conditional trim only if it loses no quality/ engagement/continuity; otherwise leave as-is. - editor-critic Lens 2 updated to stop returning REVISE for length < 5500. - Ch 17 critic note: showrunner disposition = ACCEPT at 4,675, no trim. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
### SHOWRUNNER FINAL DISPOSITION (overrides the verdict below): **ACCEPT at 4,675 words — no trim** Under the corrected length policy (length is not a gate below 5500 body words; quality, reader engagement, and continuity come first), the sole basis for the REVISE below — Lens 2 length/pacing at ~4,675 words — is **not** a gate. The chapter clears every dimension that matters: no dimension below bar, no blacklist hit, no D4 veto, and the sensitivity lens (the chapter's dominant risk) passes exemplarily. **The chapter is ACCEPTED at its original 4,675 words, untrimmed.** The Length verdict and Finding 1 below are retained as honest craft observations but carry no action. (This corrects an earlier over-rigid trim that reduced the chapter and, when interrupted, left it inconsistent; the original full version has been restored.) ### Verdict (superseded by the disposition above): **REVISE** Not a structural revise. **No dimension is below bar; no blacklist hit; D4 veto not triggered; the sensitivity lens — the chapter's dominant risk — passes exemplarily.** The chapter would be ACCEPT-able on voice and conscience. It is held at REVISE **on Lens 2 (length/pacing) alone**, which is a Major finding: the single inference is over-extended and twice-restated, putting it ~17% over band as genuine bloat rather than earned size.
ch16 (-~150w): the second self-dismantling pass compressed to one
statement per idea (duplicate-removal class, not deep compression);
"without remainder" and the staccato ifs kept per juror grafts.
NOTE: this is old ch17 under the renumbering — the July do-not-re-cut
marker — trimmed on explicit author approval and accepted by both
blind jurors ("drafting residue").
Gates: adversarial canon audit (reveal tiers hold at Suspicion,
Allegra invariant 0, […] E1-E6 clean; its 9 findings
applied) then 8 blind juries: ch14 2-0 clear, ch15 2-0 clear, ch16 2-0
(clear/slight), full ch13-17 sequence 2-0 clear — 8-0 for the edits.