Phase 4: paper-review + humanizer

[neuromechanist]

Summary

Phase 4 of epic #46. Internal manuscript:paper-review pass + manuscript:humanizer polish.

Outputs:

• manuscript/narrative-review/reviews/internal-review.md (full peer-review-style critique with 0 critical, 4 major, 10 minor concerns)
• manuscript/narrative-review/reviews/humanizer-log.md (humanizer pass record; 11 of 29 patterns clean at baseline due to project style guide; 3 patterns triggered minor edits)

Manuscript edits:

• M1: Figure 3 caption "twelve" -> "thirteen" no-coverage cells (factual correction)
• M2: Section 2 restructured from theme catalogue into perspective-driven structure
• M3: Section 7.4 (narrative-position objection) merged into a single open-questions/limitations subsection now numbered 7.4
• M4: Abstract adds 100 Hz sampling-rate caveat as a one-clause disclosure
• m1: Abstract opening tightened from 125 to 133 words then back to budget; final 133
• m2: "stop-motion or stills" corrected to "natural movie clips and static images" (Dorr 2010 actual categories)
• m4: "this Review" decapitalised in body (Cell Press style)
• m5: Section 5b opens with a thesis sentence matching 5a's structure
• m9: Section 6 adds explicit latency windows (80-300 ms occipital; 200-500 ms frontal) matching Figure 4
• m10: "intracranial cortical recording (ECoG)" -> "electrocorticography (ECoG)" first-use definition
• m8: Klin autism findings moved from Section 4 to Section 7.4 limitations
• Humanizer: 3 small rewrites for patterns 3, 11, 25 (-ing endings, synonym cycling, anthropomorphic framing)

Closes #50. Part of epic #46.

Final word counts (sections 1-7 main text)

Section	Words	Delta from Phase 3
1. Introduction	320	-6
2. Scaffold	394	+60
3. Psychophysics	370	-4
4. Action	420	-56 (autism moved to S7)
5. Language	556	-14
6. Emotion	529	+25 (latency windows added)
7. Synthesis	656	+32 (absorbed 7.4 + autism)
Total main text	3245	-10

Main text well inside TiCS Forum Review ~4000-word ceiling. Section 7 still over its individual budget but the absorption of narrative-position into open-questions is structurally sound.

Test plan

• markdownlint, typos pass
• All Highlights bullets <= 80 chars
• No em-dashes, no emojis, no AI-vocabulary tells
• All 0 critical, 4 major, and 10 minor findings from internal review either applied or explicitly deferred to Phase 5
• F1-F5 critical findings from prior self-review still satisfied
• Final Cell Press numbered reference conversion (Phase 5)
• Final formatting + figure font-size rescale (Phase 5)