Executive Summary — Acts (Handelingen) | English → Dutch Language Package

Why it matters

The Netherlands’ Bible-translation landscape carries more live, active denominational fault lines than most destination languages in this program — a historic Reformed/Catholic north-south divide, a Doopsgezinden (Mennonite) believer’s-baptism tradition contesting mainstream infant baptism, a growing Pentecostal/charismatic minority reading tongues differently than mainstream Reformed churches, and a large secular/ietsisme majority for whom Pentecost and Ascension survive mainly as holiday weekends. Acts narrates the church’s own founding events — Pentecost, the first baptisms, the first Jew-Gentile boundary crossing, the first ordained offices — which means this curriculum doesn’t just reuse the Romans baseline’s fault lines, it activates every one of them at their point of origin, and adds one entirely new one (baptism) with no Romans precedent.

Key findings

• 84 terms are now fixed in assets/translation_memory.json: 45 inherited byte-for-byte from the Romans baseline, 39 newly introduced by Acts.
• 42 of those 84 terms (50%) carry Critical or High doctrine risk and require mandatory human theologian review at every occurrence — 14 Critical, 28 High.
• 22 of the 35 doctrines tracked in assets/doctrine_risk_registry.json require human theologian review (10 Critical + 12 High); 10 require native speaker review (Medium); only 3 are automated-only (Low).
• The core passage, Acts 2:1–41, is the single highest-density concentration of Critical-risk content in the book: Pentecost, the Spirit’s outpouring, the “Lord and Christ” confession, the sovereignty-and-culpability statement of 2:23, and the baptism/repentance formula of 2:38 all sit inside these 41 verses.
• Baptism (doop) is a qualitatively new fault line absent from Romans: it directly engages the historic Doopsgezinden believer’s-baptism tradition versus Reformed/Catholic infant baptism (kinderdoop), with eight distinct occurrences across the book (2:38; 8:36-38; 9:18; 10:47-48; 16:15,33; 18:8; 19:3-5).
• The Holy Spirit’s risk tier is raised from Medium (Romans) to Critical (Acts) because 5:3,9 makes the Spirit’s personal deity explicit (“lying to the Holy Spirit”) in a way Romans never states directly.
• Existing Dutch Bible versions genuinely diverge at doctrinally consequential points (HSV vs. NBV21 at Acts 2:38, 13:39, 20:28); this curriculum fixes HSV-aligned renderings and explicitly rejects NBV21’s more resolved phrasing at each flagged verse.
• One sectarian rendering pattern (Jehovah’s Getuigen’s NWT-Dutch) poses an active, ongoing contamination risk specifically at Acts 2’s Lord-Christ identification and requires explicit guardrails.

Risks

• Denominational silent resolution: the single greatest risk across this entire curriculum is that a translation choice — not a theological argument — quietly settles a live Dutch church debate (baptism mode/timing, tongues nature, episcopal vs. eldership polity, Christian Zionism’s reading of “Israel”/“throne of David”). Every Critical/High escalation rule in 12_ai_translation_requirements.md exists to prevent this.
• Institutional false-cognates: “bisschop” (episkopos) and “Hoge Raad” (Synedrion) are both real, active Dutch institutions with meanings that would corrupt the text if substituted for the biblical referent.
• Secularized holiday-emptying: both “Pinksteren” and “Hemelvaart” are now primarily long-weekend holiday markers in Dutch civic life; without explicit re-grounding, their entire theological content risks silent loss for the majority-secular audience.
• Register over-correction: Acts 27’s physical-rescue use of the same Greek word as doctrinal “salvation” creates a real risk of over-doctrinalizing plain narrative language — the opposite failure mode from the usual under-translation risk.

Opportunities

• Acts’ concrete narrative texture (real councils, real offices, real controversies) gives Dutch Reformed readers an unusually direct biblical-narrative anchor for their own still-active church polity (ouderling, diaken, opziener) — a stronger connection than Romans offers.
• The Areopagus speech (ch.17) and its “unknown god” altar offer a built-in, textually authentic apologetic bridge to the large ietsisme secular-spiritual segment of the Dutch population, without requiring the curriculum to import extra-biblical sociological vocabulary.
• The book’s Jerusalem-to-Rome structural arc, culminating in the triumphant “unhindered” (28:31), gives teaching material a strong, textually grounded answer to any residual discomfort around Dutch colonial-era “zending” history: the text’s own trajectory is proclamation, not conquest.

Recommended actions

1. Load assets/translation_memory.json, assets/bible_term_registry.json, assets/doctrine_risk_registry.json, and analysis/12_ai_translation_requirements.md (all Acts) together with the full Romans baseline package before any Phase 2 segment processing begins.
2. Route all 14 Critical-term and 28 High-term occurrences, and all 22 Critical/High doctrines, to human theologian review per the escalation rules; route the remaining 10 Medium-risk doctrines to native speaker review.
3. Apply the baptism-ambiguity-preservation rule (retain “tot vergeving van zonden,” never NBV21’s more resolved phrasing) at every one of the eight baptism-linked passages without exception.
4. Enforce the “opziener(s)/herder(en), never bisschop(pen)” and “Sanhedrin, never Hoge Raad” naming rules at the validation layer, not just the style-guide layer, given how easily either substitution could pass an otherwise fluent Dutch review.
5. Brief all Phase 2 reviewers on the Jehovah’s Getuigen NWT-Dutch contamination risk specifically for Acts 2:20-21, 34, 36, given the group’s continued visible public presence in the Netherlands. ---END===