Executive Summary — Hebrews (English → Cantonese)

The bottom line: Hebrews is translatable into Hong Kong Cantonese using the established Protestant church register, but it carries the highest cultural-collision density of any curriculum built on this Language Package so far — because its central argument (a unique, once-for-all, unrepeatable priestly sacrifice) runs directly against Hong Kong’s living, recurring ritual-offering culture. Phase 1 is complete; every artifact required for Phase 2 enforcement is in place.

---

Why it matters

Hebrews 9:11–28, the curriculum’s core passage, makes the single most doctrinally load-bearing claim in the whole book — Christ’s sacrifice is 一次(而)永遠(有效) (once for all) — directly opposite the logic of daily/seasonal incense offerings, Ching Ming/Chung Yeung ancestor rites, and temple vow-and-fulfillment cycles that are part of ordinary religious life for much of the target audience. Get this term wrong anywhere in the book, and the central argument collapses into “the most recent in a series” rather than “the only one that was ever needed.” Nearly every other doctrine in the curriculum (High Priest, Mediator, Mercy Seat, Blood, Altar) sits in the same collision zone.

Key findings

• 32 doctrines were catalogued in assets/doctrine_risk_registry.json: 12 Critical, 14 High, 4 Medium, 2 Low. 26 of 32 (81%) require mandatory human theologian review — a notably higher proportion than typical for this pipeline, reflecting Hebrews’ concentration of sacrificial/priestly vocabulary.
• 105 terms were catalogued in assets/translation_memory.json (32 inherited exactly from the Romans baseline + 73 new to Hebrews): 22 Critical, 34 High — 56 terms (53%) requiring theologian-level enforcement before any segment is approved.
• The single highest-frequency collision in the book is 大祭司/祭司 (High Priest/Priest) vs. living Hong Kong ritual specialists (道士, 法師) — it recurs in nearly every chapter from 2 through 10.
• Hebrews introduces a new class of doctrinal-bias risk absent from the Romans package: Seventh-day Adventist “investigative judgment” readings of Hebrews 8–9, and Jehovah’s Witness (New World Translation) deity-of-Christ softening at Hebrews 1:3/1:8. Both are now explicitly screened in analysis/12_ai_translation_requirements.md.
• Every chapter of Hebrews (1–13) has been walked through in analysis/07_semantic_analysis.md and analysis/11_doctrine_analysis.md; no chapter was silently omitted, per the PRD’s full-book-coverage mandate.

Risks

• Once-for-all softening (Critical): 一次(而)永遠(有效) drifting toward “one of a series” is the single greatest risk in the curriculum — it would silently convert the book’s central claim into something Hong Kong’s temple-going culture already assumes is normal (repeatable offerings).
• Priest/ritual-specialist conflation (Critical): 大祭司/祭司 read as interchangeable with 道士/法師/和尚 would import a fee-based, repeatable-ritual framework the book explicitly argues against.
• Apostasy flattening (Critical): 背棄真道 read as ordinary, low-stakes temple-switching (轉拜另一個神) — a normal pattern in Hong Kong’s religiously plural environment — would strip the warning passages (6:4–8, 10:26–31, 12:15–17) of their intended weight.
• Mercy Seat / throne-of-grace transactionalism (Critical): 施恩座 and 施恩嘅寶座 read as a Wong Tai Sin-style wish-granting seat would reintroduce exactly the grace-as-transaction error the Romans baseline already fought to exclude, now embodied in a physical cultic object.
• Assurance/warning imbalance (Critical): Hebrews 10 places its strongest assurance language beside its strongest apostasy warning; translation must not resolve that tension by softening either side.

Opportunities

• Productive cultural friction, used deliberately: 無族譜 (Melchizedek’s “without genealogy”) directly confronts Pearl River Delta clan culture’s genealogy-based legitimacy — this can be taught as a positive contrast (God’s own oath outranks documented ancestry) rather than merely fenced off as a risk.
• Strong existing CUV vocabulary base: most of Hebrews’ priesthood/sacrifice terms (大祭司, 中保, 血, 約, 施恩座) already have established, cross-checked Traditional-character CUV renderings, giving this curriculum a firmer inherited-vocabulary foundation than a from-scratch coinage effort would.
• Direct apologetic value: the Access-to-God-through-Christ’s-Blood doctrine (親近/draw near) offers an unusually vivid, culturally resonant contrast point for Hong Kong readers, since it directly answers the felt restriction of not being able to approach a deity’s inner shrine without a ritual specialist.

Recommended actions

1. Lock assets/translation_memory.json (105 terms) and assets/doctrine_risk_registry.json (32 doctrines) as Phase 2’s enforced baseline; do not permit ad hoc term substitution during translation.
2. Route all 26 Critical/High doctrines and 56 Critical/High terms to mandatory human theologian review per analysis/12_ai_translation_requirements.md’s escalation rules — with the core passage, Hebrews 9:11–28, requiring theologian review in its entirety regardless of individual term risk tier.
3. Apply the two new bias-screening checks (SDA sanctuary doctrine; NWT deity-of-Christ softening) to every Hebrews 1, 8, and 9 segment before approval.
4. Use analysis/05_translation_landscape.md’s recommended YouVersion citation convention (version ID 46, CUV, book code HEB, 希伯來書) for every cross-reference generated in Phase 2.
5. Proceed to Phase 2 segment translation with this Phase 1 package as the complete, locked language authority for the Hebrews curriculum.

End of 01_executive_summary.md