Executive Summary — Acts Language Package (Portuguese)

Curriculum: Acts 1–28 · Core passage: Acts 2:1–41 · Destination: Portuguese (Brazil-anchored, pan-Lusophone register) · Extends: baseline Romans Language Package (never contradicted)

---

Why it matters

Acts narrates the Spirit’s outpouring, the church’s birth, and the gospel’s advance through miracle, mission, and confrontation with rival spiritual power — every one of which lands on the single most crowded religious vocabulary field in this pipeline’s destination-language footprint. Brazil hosts the world’s largest self-identified Spiritist (Kardecist) population and a large, organized Afro-diasporic religious sector (Umbanda/Candomblé) alongside its historic Catholic majority and fast-growing Evangelical/Pentecostal movements. Acts’ own subject matter — spirit-filling, tongues, healing, exorcism, visions, sorcery, resurrection, and an exclusivist salvation claim — is not an abstract comparative-religion exercise here; it recurs on nearly every chapter and activates live, mainstream, organized rival doctrinal systems rather than fringe folk belief. Getting this translation right is materially higher-stakes than the Romans baseline it extends.

Key findings

• 99 total glossary terms now govern Phase 2 translation: 47 inherited verbatim from the Romans baseline, 52 new terms introduced by Acts’ full 28-chapter scope.
• 58 terms (29 Critical + 29 High) require absolute or near-absolute enforcement with mandatory theologian-level scrutiny — more than half the entire glossary.
• 22 doctrines are registered for Acts (doctrine_risk_registry.json): 10 Critical, 10 High, 2 Medium, 0 Low — 20 of 22 doctrines require human theologian review; only 2 require native-speaker review; none are automated-only.
• The single most acute new risk cluster is the Holy Spirit and Pentecost (spirit-filling, tongues, Spirit-baptism, the gift of the Spirit): its standard Evangelical/Pentecostal Portuguese vocabulary (“cheio do Espírito Santo”) sits directly beside Kardecist/Umbanda “incorporação” and “possessão mediúnica” vocabulary describing near-identical surface phenomena with an opposite theology of personhood and agency.
• A newly identified Critical gap, absent from the Romans baseline entirely, is “imposição de mãos” (laying on of hands) — gesturally and lexically near-identical to the Kardecist “passe” healing practice performed routinely at Brazilian Spiritist centers.
• Acts’ occult-confrontation material (Simon Magus, Elymas, the Philippian spirit of divination, the sons of Sceva) is the single most directly relevant doctrinal cluster in the whole curriculum for Brazil, where divinatory mediumship and spirit consultation are mainstream and organized, not marginal.
• Acts 4:12 (“no other name… no salvation”) and Acts 16:31 (“believe in the Lord Jesus and you will be saved”) intensify, rather than merely repeat, the exclusivity risk already flagged as Critical in Romans — Acts states it with even less qualification than Romans does.

Risks

• Syncretistic softening of Spirit-filling language could read Acts 2, 4, 9, and 13 as endorsing mediumship or spirit possession rather than describing a personal, sovereign, non-displacing empowerment.
• Resurrection-adjacent vocabulary drift (“ressurreição,” “ascensão”) toward “reencarnação”/“desencarnação” would compromise the doctrinal center of Peter’s Pentecost sermon and the book’s opening scene alike.
• Softened exclusivity in Acts 4:12, 15:11, or 16:31 would collapse the curriculum’s clearest statements of the gospel’s uniqueness into Brazilian religious pluralism’s default “many valid paths” assumption.
• Anachronistic ecclesiology (“bispos” unqualified) risks importing later monarchical-episcopal or sacerdotal weight onto Acts’ functional, emergent local leadership (chs. 6, 14, 20).
• Occult-material mishandling in either direction — sensationalizing operational detail or flattening real spiritual conflict into mere psychology/fraud — would undercut both pastoral care and doctrinal accuracy in the curriculum’s highest-density collision chapters (8, 13, 16, 19).

Opportunities

• Acts supplies the New Testament’s sharpest available textual proof-texts for doctrines the Romans package could only state abstractly: the Spirit’s personhood (5:3-4, lying to the Spirit), bodily resurrection against soul-migration (2:27-31’s “no corruption”), and decisive two-kingdom conversion (26:18) — all of direct, concrete apologetic value against specifically Brazilian rival frameworks.
• The book’s own internal safeguards (Simon’s rebuked purchase-attempt at 8:18-20; the sons of Sceva’s failed magic-formula use of Jesus’ name at 19:13-17) can be taught as the text’s own built-in refutation of common syncretistic misreadings, reducing reliance on external polemic.
• Consistent, locked cross-reference verses (2:38; 4:12; 15:11; 16:31; 28:31) give Phase 2 a small set of maximally load-bearing anchor verses whose correct, consistent rendering disproportionately protects doctrinal fidelity across the whole 28-chapter curriculum.

Recommended actions

1. Merge assets/translation_memory.json (v1, 99 terms) into the Phase 2 enforcement pipeline before any segment translation begins; confirm no worker proposes ad hoc alternatives for any of the 58 Critical/High terms.
2. Route all 20 theologian-tier doctrines’ occurrences through mandatory human theologian review per assets/doctrine_risk_registry.json; do not permit automated-only approval for any Critical or High segment.
3. Prioritize back-translation QA on the five locked cross-reference verses (Acts 2:38; 4:12; 15:11; 16:31; 28:31) given their disproportionate doctrinal weight and citation frequency across all planned curriculum materials.
4. Brief all Phase 2 translators and reviewers specifically on the Kardecist/Umbanda/Candomblé collision vocabulary catalogued in Section 4 of analysis/06_linguistic_gap_analysis.md, with particular attention to the newly identified “imposição de mãos” / Kardecist “passe” collision, which has no Romans-baseline precedent.
5. Carry forward all baseline Romans conventions (register, script, citation format, escalation logic) unchanged into Acts Phase 2 processing, per analysis/12_ai_translation_requirements.md v2.0.