01 Executive Summary — Luke (English → Vietnamese)

Why it matters

Luke is the Gospel of universal reach — Savior for Jew and Gentile, rich and poor — proclaimed in a destination culture whose three dominant traditions (Mahayana Buddhism, Confucian ethics, ancestor veneration) each supply a “natural-sounding” but doctrinally wrong Vietnamese word for nearly every core term in this curriculum. The core passage, Luke 4:16-21, concentrates more Critical-risk vocabulary in six verses than any comparable stretch of Romans: anointing, release/liberty, good news to the poor, the Jubilee year, and prophetic fulfillment all converge in Jesus’ own inaugural sermon. Getting this passage — and the doctrines it anchors — right in Vietnamese is the single highest-leverage translation decision in the whole curriculum.

Key findings

• 35 doctrines analyzed across all 24 chapters (full-book coverage confirmed, no chapter silently skipped): 14 Critical, 12 High, 8 Medium, 1 Low. 26 doctrines require mandatory human theologian review; 8 require native speaker review; 1 is automated-only.
• 92 total glossary terms now governed: 47 inherited unchanged from the Romans baseline package, plus 45 new terms introduced by Luke. Of these, 18 terms are Critical risk and 32 are High risk — 50 terms (54% of the glossary) require theologian-level enforcement.
• The single highest-frequency new risk in the curriculum: the Greek word ἄφεσις carries two senses in Luke — “release/liberty” (4:18) and “forgiveness of sins” (1:77; 3:3; 5:20-24; 24:47) — and both independently forbid the same Vietnamese word, sự giải thoát, already the baseline’s most important forbidden substitution for “salvation.” This single collision point now recurs at least five times across the Gospel.
• Four additional new forbidden-substitution clusters identified beyond the ἄφεσις collision: repentance (never sám hối), compassion (never bare từ bi), Paradise (never bare thiên đàng), and Hades/Gehenna (never left unqualified against the folk-Taoist multi-court hell system).
• The core passage (Luke 4:16-21) is the single densest risk zone in the whole curriculum: five of its terms (anoint, release/liberty, poor, Jubilee year, fulfilled) are independently Critical or High risk, and the passage is now flagged for mandatory whole-unit theologian review rather than term-by-term review alone.

Risks

• Syncretism risk is structural, not incidental. Vietnamese Buddhism, Confucianism, and ancestor veneration each supply positive, respected, commonly-used vocabulary that is doctrinally opposite to Luke’s meaning in at least one major doctrine each: Buddhism (liberation/merit/compassion/afterlife), Confucianism (filial obedience vs. Christ’s absolute priority), ancestor veneration (prayer as ancestor petition, Hades as a negotiable underworld).
• Repetition multiplies exposure. Repentance/forgiveness vocabulary alone recurs at eleven separate passages; a single wrong lexical habit (e.g., reaching for sám hối or giải thoát) would propagate across the entire Gospel rather than remaining a localized error.
• “Today” and “fulfilled” are easy to under-translate. Both are ordinary-sounding Vietnamese words (hôm nay, được ứng nghiệm) that carry unusually heavy theological freight in Luke; without deliberate cross-document tracking, translators may treat each occurrence as a generic time-stamp and quietly lose Luke’s distinctive “already-arrived” Kingdom theology.

Opportunities

• Vietnam’s century of settled Protestant Bible vocabulary is a strong asset. Đức Chúa Trời, Đức Chúa Giê-xu, Đức Thánh Linh, and sự cứu rỗi are already fixed and non-ambiguous; the work here is disciplined extension, not invention from scratch.
• Several doctrines have genuine bridging vocabulary available for teaching, not just fencing: Jewish and Islamic repentance/forgiveness concepts, Jewish Jubilee/tzedakah roots for “good news to the poor,” and Sufi/guru-disciple total-submission models for discipleship cost can all be used pastorally as partial points of contact, provided the translation itself never substitutes the competing term.
• The transliteration strategy already validated for Đấng Mê-si-a extends cleanly to Paradise (Ba-ra-đi), giving translators a proven pattern for marking a term as an imported theological category requiring explanation, rather than inventing a new solution.

Recommended actions

1. Adopt assets/translation_memory.json (v1, 92 terms) and assets/bible_term_registry.json before any Phase 2 segment translation begins, loaded alongside (never instead of) the Romans baseline files.
2. Route Luke 4:16-21 for mandatory whole-passage theologian review, not just individual flagged terms, given its unmatched risk density.
3. Enforce the ἄφεσις sense-determination step (release vs. forgiveness) as a distinct, logged decision in every segment where the word occurs, per analysis/12_ai_translation_requirements.md.
4. Brief all human theologian reviewers on the five new forbidden-substitution clusters (ἄφεσις/giải thoát, sám hối, từ bi, bare thiên đàng, unqualified âm phủ/hỏa ngục) before review work begins, since these are new to this curriculum and not covered by Romans-era reviewer training.
5. Track “today” (hôm nay) and “fulfilled” (được ứng nghiệm) as linked cross-reference sets (2:11, 4:21, 19:9, 23:43; 4:21, 24:44) across all Phase 2 output to preserve Luke’s inaugurated-kingdom theology consistently.