Executive Summary: Acts — English → Ndebele Language Package

Why it matters

Acts is the highest-risk book in this language pair produced so far. It inherits every mediated-access risk already documented in the Romans baseline (uNkulunkulu approached through amadlozi, not directly) and adds a doctrine absent from Romans entirely: direct narrative confrontation with occult and spiritual powers (Simon Magus, Elymas, the python-spirit slave girl, the sons of Sceva, Artemis of the Ephesians). Getting Acts 2:1-41 — Pentecost, and its salvation formula “call on the name of the Lord” — right in Ndebele is the theological anchor for teaching Christ as sole mediator to an audience whose traditional religion normally requires an ancestral intermediary.

Key findings

• 20 doctrines identified across all 28 chapters of Acts (full-book coverage confirmed, no chapter silently skipped); 10 are Critical-tier, up from 9 in the Romans baseline — the increase is driven almost entirely by the new Occult and Spiritual Powers doctrine.
• 112 terms now under enforcement in the expanded translation memory: 41 terms inherited exactly from the Romans baseline, plus 71 new Acts-specific terms and phrases.
• New terms alone contain 15 Critical and 15 High risk items (per 08_core_glossary.md risk tier summary) — nearly matching the entire Romans baseline’s Critical+High count (18) in new vocabulary alone.
• The single highest-stakes new collision is “unknown god” (Acts 17:23), which sits closer to popular Ndebele theology (a remote uNkulunkulu) than any term in the Romans baseline; Paul’s own rhetoric must be used against that framing, not read as confirming it.
• 17 of 20 Acts doctrines require mandatory human theologian review (Critical + High combined); only 2 require native speaker review and 1 is automated-only — a heavier oversight load than Romans’ comparable 18-of-24 baseline ratio.

Risks

• Occult/spiritual-power vocabulary cluster (ubuthakathi, umoya wokubhula, umoya omubi, uSathane) is both a doctrinal and a pastoral-safety risk: witchcraft-accusation language carries real, sometimes violent, social consequences in contemporary Ndebele communities, so imprecise translation is not a purely academic error.
• Repentance and baptism (Critical-tier, recurring in 9 of 28 chapters) sit immediately adjacent to traditional ritual appeasement and cleansing practices; this is the most frequently recurring Critical risk in the book.
• Evil/unclean spirit vs. amadlozi is the single most dangerous category-collapse point in the entire two-book package: conflating them either wrongly demonizes ancestor-veneration wholesale or wrongly domesticates a hostile biblical category.
• Elders (abadala) and patriarch (ukhokho) both reuse ordinary Ndebele words for customary/ancestral roles carrying real traditional authority, requiring careful fencing in every occurrence.

Opportunities

• Ndebele’s established Bible translation tradition (rooted in Zulu-shared vocabulary, first printed complete in 1971/1978) means almost no term requires invention from scratch — the work is disambiguation and fencing, not lexical gap-filling.
• Ndebele kinship vocabulary (household baptism, indlu/umuzi) gives the Acts 16 household-conversion narratives unusually strong, positive cultural resonance when handled correctly.
• Acts’ repeated, decisive gospel-versus-occult-power confrontations, if taught with precision, directly and repeatedly reinforce the Romans baseline’s central corrective (Christ alone mediates, no ancestral intermediary needed) more concretely than Romans’ own more abstract argumentation.

Recommended actions

1. Route all 17 Critical/High-risk Acts doctrines to mandatory human theologian review before Phase 2 translation is finalized, per doctrine_risk_registry.json.
2. Treat the Occult and Spiritual Powers cluster (§2.2 of 06_linguistic_gap_analysis.md) as a single review block, not term-by-term, given how tightly its vocabulary interlocks.
3. Enforce the expanded translation_memory.json and the extended 12_ai_translation_requirements.md forbidden-substitution list without exception in all Phase 2 segment translation.
4. Prioritize theologian sign-off on Acts 2:21/4:12 (call on the Lord’s name / no other name), Acts 17:23-28 (unknown god), and Acts 19:12-16 (evil spirit vs. amadlozi) as the three highest-severity ambiguities identified in this Language Package.

---