Executive Summary — Acts (Elçilerin İşleri) Turkish Language Package

Why it matters

Acts is the theological hinge of the whole nine-book Turkish Language Package: it narrates, rather than argues, everything Romans through John have established doctrinally — the Spirit’s full personhood, Christ’s deity and Lordship, justification apart from law-works, and the gospel’s move from Jew to Gentile — while introducing this pipeline’s single most dangerous new false-friend (şehit for “witness”) and its densest concentration of Critical risk in one chapter (the Jerusalem Council, ch. 15). Getting Acts right or wrong shapes how every downstream Turkish reader understands the church’s own founding story.

Key findings

• 25 doctrines, 12 Critical / 11 High / 4 Medium, 0 Low — 23 doctrines require mandatory human theologian review; 4 require native-speaker review; none are automated-only (assets/doctrine_risk_registry.json).
• ~105 glossary terms tracked, 66 rated Critical or High (25 Critical, 41 High) — every one requires theologian sign-off before Phase 2 promotion (assets/bible_term_registry.json).
• 59 brand-new terms never before analyzed in this Language Package originate in Acts, spanning all 28 chapters — from Pentecost’s “tongues as of fire” to the book’s final “preaching unhindered.”
• The core passage, Acts 2:1-41, alone concentrates 7 of the 25 registry doctrines (Holy Spirit, Resurrection, Deity/Lordship of Christ, Repentance and Baptism, Election, Gospel to Gentiles, Church as Community) — every segment drawn from it is auto-escalated.
• Full-book coverage confirmed: all 28 chapters analyzed; five (14, 18, 21, 25, 27) explicitly marked “reviewed, no new doctrinal load” per the PRD’s no-silent-omission mandate.

Risks

• Şehit vs. tanık (“witness”): the single highest-priority forbidden substitution in this package. Şehit’s Islamic martial/state-martyrdom resonance would import an entirely wrong frame onto Stephen’s death and the book’s whole persecution theme.
• Kulu (“Servant,” Acts 3:13, 26) left unqualified: risks silently flattening Jesus into “Allah’ın kulu,” an ordinary devout servant, undercutting established deity doctrine unless every occurrence carries an explicit Isaiah 53 qualifier.
• Double-stacked Critical risk at the Jerusalem Council (ch. 15): amel/ameller (works of the law) and the Sünnet/Sünnet homonym collision both peak in the same passage — Acts’ own central narrative crisis.
• κοινός’s two unrelated senses inside one book: “shared” (2:44/4:32) vs. “unclean” (10:14-15/11:8) risk conflation by any translator working quickly across chapters.
• Acts 2:23’s “determined plan and foreknowledge”: the passage in this entire Language Package where a kader-style misreading (removing human responsibility) is most tempting.
• Acts 2:36’s “God made him Lord and Christ”: “yaptı” risks an adoptionist misreading unless taught as redemptive-historical vindication of Christ’s pre-existing deity.

Opportunities

• Turkish “dil” naturally carries both “tongue” and “language,” preserving the Greek glōssa wordplay at Pentecost — a genuine translation asset unavailable in many other target languages.
• Ottoman legal/commercial vocabulary (mühür, kapora, senet) already established in prior books extends naturally into Acts’ own legal-narrative register (Sanhedrin proceedings, Roman citizenship).
• The Areopagus speech (ch. 17) offers a ready-made, textually authentic model for respectful engagement with Turkey’s own religious pluralism, paralleling the interfaith discipline already proven for John 14:6.
• Tövbe’s genuine positive Islamic resonance can be affirmed honestly at every repentance-and-baptism passage, provided each occurrence is explicitly anchored to Christ’s atoning work.

Recommended actions

1. Load translation_memory.json (v1, 105 terms) and enforce the şehit/Kulu/Hades/amel/kader forbidden-substitution list without exception in every Phase 2 segment.
2. Route Acts 2:1-41, chs. 5, 7, 10-11, 15, 9/22/26, and 17 to human theologian review automatically, per 12_ai_translation_requirements.md’s Acts Additions escalation rules.
3. Lock a single consistent rendering of παρρησία (“boldness”) across all nine-plus occurrences (4:13 through 28:31) before any parallel Phase 2 batch processing begins, to preserve Luke’s deliberate structural inclusio.
4. Brief all native-speaker reviewers explicitly on the κοινός double-sense fence and the Sünnet/Sünnet homonym risk before assigning any ch. 2-4, 7, 10-11, or 15 segments.
5. Pair every “no other name” (4:12) and Areopagus (ch. 17) segment with mandatory pastoral/interfaith teaching-note scaffolding, matching the discipline already proven for John 14:6.