Executive Summary: Revelation (English → Javanese) Language Package

Phase 1 Complete — Translation Requirements Interface

Why it matters: Revelation is the highest doctrinal-density curriculum this pipeline has processed for Javanese. Its apocalyptic genre collides with Javanese cultural vocabulary far more extensively than Romans’ epistolary argument did — not primarily with normative Islam, but with the syncretic kejawen mystical layer (protective serpent-guardian mythology, the Jayabaya “hidden knight” restorer prophecy, aristocratic polygamous marriage custom, and a purgative-testing-crater cosmology) that runs beneath Java’s formal Muslim-majority identity. Getting this Language Package right determines whether Javanese learners receive Revelation’s promises of Christ’s return, final judgment, and the new creation — or an unintentional syncretic substitute.

Key findings:

• The book’s own fixed citation title, “Wahyu,” is a load-bearing collision with the baseline’s own forbidden-term list (wahyu = impersonal royal-mandate light). Resolved via a split policy: retain “Wahyu” only as the immutable citation title; use kawiyakan for the common noun everywhere else.
• Full-book coverage (all 22 chapters) is complete: every chapter is mapped to its load-bearing doctrine(s) and vocabulary; chapters contributing no new terms (8, 10, 18) are explicitly logged as reviewed, not silently skipped.
• The core passage, Revelation 21:1–8, sits at the intersection of five Critical-tier doctrines simultaneously (New Heaven and New Earth, New Jerusalem as God’s Dwelling, Second Death, Adoption/Inheritance of Overcomers, Sovereignty of God over History) and is now confirmed as the single highest-density review passage in the curriculum.
• 24 baseline Romans terms carry forward unchanged into Revelation (Gusti Allah, Gusti Yesus, sih-rahmat, kaslametan, Sang Mesias, etc.); 78+ new terms have been catalogued, defined, risk-tiered, and seeded into assets/translation_memory.json.

Risks (quantified for theologian-oversight planning):

• Doctrine level (doctrine_risk_registry.json): 35 doctrines tabled — 15 Critical, 13 High, 6 Medium, 1 Low among Revelation-specific doctrines; combined with baseline-inherited doctrines retaining Critical/High tiers (Grace, Salvation, Righteousness, Justification, Resurrection, Power of God, Messiah, Sonship of Christ, Incarnation), 28 doctrines require mandatory human theologian review, 6 require native speaker review, 1 is automated-only.
• Term level (bible_term_registry.json + translation_memory.json): of the ~96 terms in the core Revelation registry, 29 are Critical-risk and 38 are High-risk — 67 terms requiring theologian-level scrutiny at every occurrence, before counting the additional ~40 supplementary full-book-coverage terms seeded in Step 9.
• Single highest-risk ambiguity: unqualified naga (dragon) — Java’s naga is typically a protective, venerated guardian figure (keraton gateway serpents, Sang Hyang Anantaboga), the functional opposite of Revelation 12’s meaning. A single unglossed occurrence could invert the doctrine of Satan’s defeat.
• Second-highest: garwanipun Sang Cempe (Bride/wife of the Lamb) — collides with the still-living Javanese aristocratic custom of a nobleman holding multiple consorts (garwa padmi/ampeyan), directly undermining the required singular, exclusive covenant reading.
• Structural risk: the Wahyu/kawiyakan split must be enforced perfectly in every Phase 2 document; any slippage reactivates the baseline’s most tightly guarded forbidden term inside the very book named after it.

Opportunities:

• Javanese already has an established, continuous Bible-translation tradition (Kitab Sutji, LAI 1980/1994) supplying settled, non-negotiable renderings for the highest-risk terms (Gusti Allah, Gusti Yesus, Roh Suci, kaslametan) — this curriculum does not need to invent core vocabulary, only extend it carefully.
• Several genuinely missing-vocabulary items (Alpha and Omega, Lamb-as-royal-title, second death, martyr distinct from syahid) have been resolved through disciplined transliteration/coinage decisions consistent with the baseline’s own precedent (Abba), giving Phase 2 a settled, defensible rationale for each rather than an ad hoc choice.
• The doctrine of Judgment “according to works” (Rev 20) has been proactively harmonized with the baseline’s justification-by-faith doctrine via a mandatory translator note, preventing what would otherwise be the curriculum’s most likely source of apparent internal contradiction with Romans.
• The exclusivity-of-worship doctrine’s deliberate reuse of nyembah for both true and counterfeit worship (ch. 13 vs. chs. 4–5, 19) is preserved rather than smoothed over, keeping intact a rhetorical device the source text itself relies on.

Recommended actions:

1. Load assets/translation_memory.json (this package) alongside the baseline Romans file for every Phase 2 segment; treat any divergence in shared terms as a packaging error that halts translation.
2. Route all 28 theologian-mandatory doctrines and 67+ theologian-mandatory terms through human theologian review before any segment is marked approved, per analysis/12_ai_translation_requirements.md’s escalation rules.
3. Apply the mandatory translator-note requirement without exception at every occurrence of: kaslametan (doxological uses especially), pati kang kaping pindho, naga, garwanipun Sang Cempe, and kaadili miturut pandamelanipun.
4. Enforce the Wahyu/kawiyakan title split via automated pre-submission validation (regex-style check for lowercase “wahyu” outside citation contexts) in addition to human review.
5. Carry the cross-document consistency mandates (Setya tuhu lan sanyata at 3:14/19:11/21:5; Sang Cempe at all ~28 occurrences; Revelation 21:1–8 as fixed core-passage language) into every Phase 2 lesson file without exception, mirroring the rigor the baseline established for Romans 1:16–17, 8:28, and 10:9–10.