Executive Summary: Revelation Language Package (Odia)

Why it matters

Revelation is the highest doctrinal-risk book this pipeline has processed for Odia. It intensifies every syncretism risk the Romans baseline already flagged (Lord/ପ୍ରଭୁ vs. Jagannath’s ଠାକୁର; resurrection vs. rebirth; salvation vs. moksha) and adds a genuinely new, Odisha-specific collision cluster: a cosmic serpent-antagonist that collides with the revered Ananta Śeṣa, a “beast” antagonist that collides with Vishnu’s revered animal avatāras, and — most acutely — an “image of the beast” that shares its ordinary Odia word (ପ୍ରତିମା) with Jagannath’s own consecrated, periodically-renewed sacred image. The core passage (Revelation 21:1-8) alone carries five of the fourteen Critical-risk doctrines simultaneously, making it the theological anchor the entire curriculum’s risk-management strategy is built around.

Key findings

• 99 glossary terms are enforced in translation_memory.json, of which 71 (72%) are Critical or High risk — 25 Critical, 46 High, 23 Medium, 5 Low. This is a markedly higher risk concentration than the Romans baseline.
• 29 doctrines are tracked in doctrine_risk_registry.json: 14 Critical, 13 High, 2 Medium, 0 Low. 27 of 29 doctrines (93%) require mandatory human theologian review; only 2 require native-speaker-only review, and none are automated-only — a sharp contrast to Romans, where automated-only review was possible for some Low-risk doctrines.
• The single highest-stakes term-collision identified across both the Romans and Revelation packages combined is “image of the beast” (ପଶୁର ପ୍ରତିମା), given its surface resemblance to Nabakalebara image-consecration language.
• Five passages are recommended for locked, verbatim cross-document rendering: Revelation 21:5 (New Creation thesis), 12:11 (interpretive key for “overcoming”), 11:15 (Reign of Christ thesis), 19:16 (Lordship title), and 22:20-21 (closing prayer/benediction).
• Four items were left open in prior analysis steps for explicit human theologian ruling; all four have now been resolved in assets/bible_term_registry.json and carried into assets/translation_memory.json: παντοκράτωρ → ସର୍ବସାମର୍ଥ୍ୟବାନ (never ସର୍ବଶକ୍ତିମାନ); δράκων/θηρίον → ମହାସର୍ପ/ପଶୁ with mandatory disambiguating notes; εἰκὼν τοῦ θηρίου and “no temple” (21:22) → fulfillment-framed pastoral treatment, not polemic; κρίνω κατὰ τὰ ἔργα → prefer ଆଚରଣ over କର୍ମ wherever syntax allows.

Risks

• Critical, Odisha-specific: the dragon/serpent, beast, and image-of-the-beast vocabulary cluster (chs. 12-13, 17, 19-20) risks either (a) inadvertently validating Jagannath devotional imagery as compatible with the text’s antagonists, or (b) being read as a direct, inflammatory attack on Odia religious heritage if pastoral framing is omitted. Both failure modes are theologically and pastorally serious given the historical sensitivity of Hindu-Christian difference in this region (2008 Kandhamal riots).
• Critical, cross-cutting: 21:22’s “no temple” statement, set against the centrality of the Puri Jagannath Mandira to Odia identity, is the single passage most likely to be misread as polemical if translated without its accompanying fulfillment-framing note.
• High, systemic: the worship verb προσκυνέω (~24 occurrences) requires flawless object-sensitive discipline (ଉପାସନା vs. ପ୍ରଣାମ) across the entire book; a single inconsistent rendering would blur the book’s central moral fork between true and false worship.
• High, systemic: κ୍ଲେଶ/tribulation and କର୍ମ/“judged according to works” both carry live, competing technical meanings in mainstream Odia religious philosophy (Yogic kleśa; karmic rebirth-ledger) that must be actively fenced in every occurrence, not merely translated.

Opportunities

• Odisha’s existing, mature transliteration-preferring convention (established by the baseline’s ପରିତ୍ରାଣ-not-ମୋକ୍ଷ precedent) extends cleanly to Revelation’s new collision terms (ହାଦେସ୍, ଅଗାଧ ଗର୍ତ୍ତ, ଶୟତାନ), giving translators a proven pattern to apply rather than requiring invention from scratch.
• The Ananda Bazaar caste-crossing mahaprasad tradition and the “no distinction” universal-scope doctrine offer a genuine, if imperfect, bridge point for teaching Revelation’s fully inclusive redeemed-multitude vision (5:9; 7:9; 21:24-26) — a rare case where local practice partially anticipates, rather than merely risks confusing, the text’s own claim.
• Five thesis-level verses recommended for locked cross-document consistency give Phase 2 a small, high-leverage set of verses where getting the rendering right once pays dividends across every downstream lesson and document.

Recommended actions

1. Route all Critical and High risk segments (71 of 99 glossary terms; 27 of 29 doctrines) to mandatory human theologian review in Phase 2 Step 17 — plan reviewer capacity for a substantially higher review load than the Romans baseline required.
2. Attach the five mandatory pastoral/disambiguating translator notes (dragon/serpent, image of the beast, no temple, idolater, tribulation/kleśa) as non-optional Phase 2 output requirements, not discretionary footnotes.
3. Lock the five cross-document thesis verses (21:5; 12:11; 11:15; 19:16; 22:20-21) verbatim before any Phase 2 lesson content is drafted, to prevent downstream inconsistency.
4. Brief all Phase 2 human theologian reviewers on the Kandhamal riots history (analysis/04_comparative_theology.md) before assigning cosmic-conflict and idolatry-vocabulary segments, so pastoral framing decisions are made with full awareness of the stakes for the reading community.
5. Carry the four now-resolved open items (Almighty root; dragon/beast disambiguation; image-of-the-beast/no-temple pastoral framing; κρίνω κατὰ τὰ ἔργα wording) into Phase 2 as settled, non-negotiable defaults — do not reopen them mid-translation.