01 Executive Summary — Acts 1–28 Language Package (English → Vietnamese)

Why it matters

Acts is the highest-collision-risk curriculum yet built on this Vietnamese Language Package. Where Romans dealt mainly with abstract doctrine (grace, justification, resurrection), Acts narrates spirit-phenomena — tongues, visions, exorcism, sorcery, casting lots — that sit within a few words of living, currently practiced Vietnamese folk religion (đồng bóng spirit-mediumship, thầy bói fortune-telling, bùa chú charms, cúng ancestor-offerings). Get the fencing wrong here and a Bible study lesson can accidentally validate the very practices the text condemns, or make Pentecost read like a possession-trance account. The core passage, Acts 2:1-41 (Pentecost and Peter’s sermon), is the single densest concentration of these risks in the whole book and anchors the entire curriculum’s doctrinal load-bearing.

Key findings

• Full-book coverage confirmed. All 28 chapters carry identifiable doctrinal content; none required a “no new doctrine” placeholder (see 11_doctrine_analysis.md Part 2).
• 41 doctrines identified (13 Critical, 13 High, 13 Medium, 2 Low), clustering under the curriculum’s 9 named doctrines.
• Translation memory now totals 112 terms: 44 inherited unchanged from the Romans baseline, plus 68 new Acts-specific terms.
• 52 terms are Critical or High risk (18 inherited + 34 new), meaning nearly half the whole-book vocabulary load requires mandatory human theologian sign-off before Phase 2 processing — up from the Romans baseline’s comparatively lower Critical/High density.
• One risk tier upgrade discovered mid-analysis: “held all things in common” (Acts 2:44-45; 4:32-35) escalated from Medium to High after identifying a Vietnam-specific collision with 20th-century agricultural-collectivization memory (hợp tác xã) that the Romans baseline had no occasion to flag.
• A new dual-rendering rule (Đấng Mê-si-a vs. Đấng Christ) must be applied consistently across all 28 chapters — a scale of consistency-management Romans never required.

Risks

• Critical: Spirit-filling/tongues (Acts 2, core passage). Nearest cultural analogue is đồng bóng/hầu đồng spirit-possession, where the medium’s will is displaced. Biblical filling must be taught as heightened, not displaced, agency; every occurrence requires theologian review.
• Critical: “The Way” (Ch. 9, 19, 22, 24). Bare “Đạo” reads as one path among Vietnam’s normal religious plurality (Đạo Phật, Đạo Khổng, đạo ông bà), directly undermining the exclusivity claims elsewhere in the book.
• Critical: Sorcery vs. apostolic miracle (Ch. 8, 13, 16, 19). Direct, repeated collision with live occult practice; mishandling risks presenting magic and miracle as equivalent “spiritual technologies.”
• Critical: “No other name” (4:12). Absolute salvation exclusivity runs directly against default Vietnamese religious-pluralism assumptions.
• High: Repentance vs. sám hối (9+ chapters). Very frequent term sitting immediately beside a Buddhist self-effort merit-clearing concept.
• High: Jerusalem Council food/blood provisions (Ch. 15). Direct, ongoing collision with ancestor food-offerings (cúng) and blood-based cuisine (huyết, tiết canh); requires pastoral notes, not silent translation.
• Newly surfaced: “Held all things in common” (Ch. 2, 4, 11). Risk of misreading as state-mandated collectivization rather than voluntary grace-generosity — a collision unique to Vietnam’s 20th-century history, absent from the Romans baseline entirely.

Opportunities

• Vietnamese Protestant Bible vocabulary is a century deep and already settled for most core terms (Đức Thánh Linh, sự cứu rỗi, sự sống lại) — the work is disciplined reuse and fencing, not invention.
• Acts’ own narrative structure (Ch. 8, 13, 16, 19’s sorcery-vs-miracle contrasts; Ch. 15’s Council) gives built-in teaching moments to proactively address syncretism risk rather than merely avoid it.
• The Đấng Mê-si-a / Đấng Christ distinction, once locked into translation memory and QA tooling, resolves an ambiguity that would otherwise recur roughly 40+ times across the book.

Recommended actions

1. Load assets/translation_memory.json (Acts version) and analysis/12_ai_translation_requirements.md before any Phase 2 segment work begins — both extend, and never contradict, the Romans baseline.
2. Route all 52 Critical/High terms and 26 Critical/High doctrines to mandatory human theologian review (per assets/doctrine_risk_registry.json risk_summary: total_requiring_theologian_review = 26 doctrines; term-level count = 52 of 112 total terms).
3. Run a dedicated Đấng Mê-si-a/Đấng Christ consistency audit across all Phase 2 output before Step 17 sign-off, given the scale at which this rule applies in Acts versus Romans.
4. Attach mandatory pastoral teaching notes (not silent translation) at every occurrence of “held all things in common,” “things sacrificed to idols,” “blood,” and “the Way” — these four terms carry cultural-collision risk that translation alone cannot resolve.
5. Prioritize 100% theologian-review sampling on the ranked top-tier ambiguities identified in analysis/06_linguistic_gap_analysis.md Section E (items 1-3, 6) during Phase 2 Step 17 QA.