01 Executive Summary — Ezekiel (English → Mandarin)

The bottom line: Ezekiel is translatable into Mandarin at high doctrinal fidelity using the existing CUV (和合本, 神版) tradition and the Romans baseline Language Package — but it carries substantially more Critical/High collision risk than Romans did, concentrated in a handful of sharp, well-defined worldview confrontations with Chinese folk religion, Confucianism, and Daoism.

---

Why it matters

Ezekiel is Hebrew OT prophecy, not NT epistle — a different genre, source language, and risk profile than the Romans package this pipeline was originally built around. Its core passage (37:1-14, the Valley of Dry Bones) sits at the exact intersection of the book’s hardest translation problem (רוּחַ/ruach’s untranslatable wind-breath-Spirit wordplay) and its most theologically load-bearing doctrine (national restoration typologically anticipating resurrection). Get this passage wrong, and the curriculum’s anchor lesson either flattens into folklore (还魂/ghost reanimation), imports reincarnation (轮回/投胎), or collapses OT corporate restoration into NT individual resurrection without the needed distinction. Six other doctrines carry equally serious, equally specific collision risks with dominant Chinese religious traditions.

---

Key findings

• Full-book coverage complete: all 48 chapters reviewed; every chapter mapped to a doctrine and risk tier, with “reviewed — no new terms” chapters explicitly logged rather than omitted (see analysis/11_doctrine_analysis.md, Part C).
• 68 terms now populate the Ezekiel-specific translation memory (assets/translation_memory.json): 21 inherited exactly from the Romans baseline (some risk-elevated for this book, per established prior analysis) and 47 newly introduced by Ezekiel.
• Doctrine risk registry totals (assets/doctrine_risk_registry.json): Critical = 6, High = 11, Medium = 5, Low = 2 across 24 sub-doctrines, grouped into the curriculum’s 7 named doctrines.
• 17 of 24 sub-doctrines (71%) require mandatory human theologian review — nearly triple the Romans baseline’s proportional theologian-review load — reflecting Ezekiel’s unusually dense concentration of OT-specific, culturally load-bearing imagery (glory-throne visions, temple architecture, national-resurrection metaphor, dragon/sea-monster judgment imagery).
• Core-glossary term-level risk (analysis/08_core_glossary.md): of the book’s newly introduced/extended terms, Critical = 11, High = 15, Medium = 14, Low = 8.
• The single highest-risk lexical item in the whole curriculum is רוּחַ (ruach) at Ezekiel 37:9, where one Hebrew word spans wind/breath/Spirit within one verse — no Mandarin term can carry all three senses, and mandatory teacher notes (not a translation fix) are the only mitigation.

---

Risks

1. רוּחַ (ruach) wordplay collapse — a translator or automated system defaulting to one uniform Mandarin word (气息, 风, or 灵) across Ezekiel 37 destroys the passage’s central theological argument. Absolute forbidden term: 气 (qì), the Daoist/qigong vital-energy concept.
2. Ancestral guilt-transfer collision (Ezekiel 18:20) — the single sharpest worldview confrontation in the curriculum; Chinese ancestor-veneration and inherited-merit/demerit folk logic (阴德-adjacent) directly contradicts “the son shall not bear the guilt of the father.” Under-translation here leaves the curriculum’s most culturally significant verse ambiguous.
3. 龙 (dragon) temptation at Ezekiel 29/32 — low-probability, catastrophic-severity risk: using 龙 for Pharaoh’s tannim imagery would invert a judgment-on-pride oracle into an accidental endorsement of China’s supreme national/imperial symbol (“descendants of the dragon”). Absolute prohibition required, not a preference.
4. New Heart/New Spirit self-cultivation drift — high probability that a well-meaning teacher reaches for 脱胎换骨 (Daoist alchemical transformation) or 修身/修炼 (self-cultivation) to explain 石心/肉心 and 新心/新灵, silently converting a sovereign, one-time divine gift into an earned technique.
5. Corporate-vs-individual resurrection conflation — 复活 is Critical-risk-anchored elsewhere to individual, literal, bodily resurrection; Ezekiel 37’s primarily national/corporate referent risks either erasing that distinction or, worse, drifting toward a reincarnation-adjacent reading if mishandled.
6. Glory-narrative fragmentation — 荣耀’s departure (chs. 8-11) and return (ch. 43, 48) is a single 48-chapter narrative arc; treating each occurrence as an isolated vocabulary item (rather than a continuous story) undercuts the God’s-Presence-Departing-and-Returning doctrine.
7. Temple-vision cultural mapping — 圣所/殿 (chs. 40-48) risk default assimilation to existing Chinese sacred-space categories (寺/道观/庙) and their associated idol/ancestor-veneration practices, none of which the vision’s holiness regulations permit.

---

Opportunities

• Built-in teaching moments already exist inside the text itself. Ezekiel 13’s false prophets pronouncing hollow “peace” (平安) is a ready-made in-text illustration of the exact 平安-as-amulet-blessing risk the Romans baseline already flags — the curriculum can teach the risk using the Bible’s own critique of it.
• Strong positive cross-tradition bridges exist and can be leveraged in teaching material, per analysis/04_comparative_theology.md: Islamic theology’s “no soul bears another’s burden” closely parallels Ezekiel 18; the Quran’s own dry-bones-to-life imagery (2:259; 36:78-79) offers a genuine, respectful bridge for Ezekiel 37 with Muslim-background learners; Jewish Shekhinah theology offers rich continuity for the Presence-Departing-and-Returning doctrine.
• Over a century of settled CUV vocabulary (救恩, 复活, 耶和华, 圣所) means most of Ezekiel’s translation work is defense of already-good vocabulary against three well-mapped sets of superficially similar borrowed meanings (Confucian, Buddhist, Daoist/folk) — not invention of new terms from scratch.
• The core passage’s very difficulty is pedagogically useful: explaining why no single Mandarin word can carry ruach’s full range is itself a teaching opportunity about the richness of the original text, not merely a translation limitation to apologize for.

---

Recommended actions

1. Route all Ezekiel 37:1-14 segments to human theologian review without exception, regardless of whether an individual segment appears to contain a flagged term — the core passage’s risk is structural (the ruach wordplay, the resurrection-typology boundary), not merely lexical.
2. Hard-code 龙 as an absolute forbidden term in Phase 2 validation (not merely a flagged preference) for any Pharaoh/Egypt-oracle content (chs. 29, 32).
3. Require the mandatory ruach teaching note at Ezekiel 37:1, 9, and 14 in every derived teaching artifact, not just the base translation — this is a compensating pedagogical structure, not a one-time footnote.
4. Pair every occurrence of 义人/回转/悔改 in chs. 14, 18, 33 with the standing “non-transferable, not karma-purifying” doctrinal fence established in analysis/06_linguistic_gap_analysis.md §3.
5. Resource Phase 2 review capacity for a ~71% theologian-review load (17 of 24 sub-doctrines) — significantly higher than Romans’ comparable proportion — before beginning bulk segment translation.
6. Load assets/translation_memory.json (Ezekiel) together with, never instead of, the Romans baseline for every Phase 2 session, per the updated pre-flight checklist in analysis/12_ai_translation_requirements.md v1.28.

---

This summary synthesizes analysis/04_comparative_theology.md, analysis/06_linguistic_gap_analysis.md, analysis/07_semantic_analysis.md, analysis/08_core_glossary.md, analysis/11_doctrine_analysis.md, assets/bible_term_registry.json, and assets/doctrine_risk_registry.json. It must be read alongside, not instead of, those full artifacts before Phase 2 begins.