Executive Summary: Jude Translation Requirements (English–Kannada)

Why it matters

Jude is short — 25 verses, one chapter — but it is the most doctrinally concentrated book yet processed in this pipeline for Kannada. Almost every verse carries Critical- or High-risk theological freight, and Jude’s argument depends on Kannada readers correctly distinguishing biblical categories from Karnataka’s three living devotional traditions — Haridasa Krishna-bhakti, Lingayat/Veerashaiva theology, and mainstream Shakta devotion — each of which offers vocabulary that sounds right term-by-term while pointing toward a different theological destination (most acutely, Aikya: the devotee’s identity dissolving into Shiva). Jude introduces two new, letter-specific translation challenges the Romans baseline did not need to solve: a paired double-title for Christ’s lordship (δεσπότης + κύριος, both denied in one clause, 1:4) and a five-times-repeated structural keyword (“kept,” τηρέω) whose consistent rendering carries the letter’s central argument.

Key findings

• 78 terms are catalogued in the Jude-specific term registry (assets/bible_term_registry.json); of these, 17 are Critical risk and 21 are High risk — 38 terms (49%) require mandatory human theologian review before Phase 2 translation, versus 23 Medium (native speaker review) and 17 Low (automated review).
• 23 doctrines are catalogued in the Jude-specific doctrine registry; 9 are Critical and 12 are High — 21 doctrines (91% of all doctrines identified) require human theologian review, versus 3 Medium (native speaker review) and 0 Low. No doctrine in Jude qualifies for automated-only review.
• Jude’s single highest-priority translation-mechanics risk is not a wrong word but wrong consistency: the τηρέω/φυλάσσω “kept” word-family (1:1, 1:6 x2, 1:13, 1:21, 1:24) must use one Kannada root throughout, or the letter’s core argument (the God who reserves judgment for the disobedient is the same God who keeps his own people secure) becomes invisible to the reader.
• Jude 1:1’s self-designation ದಾಸ (servant/bondservant) is the exact root of “Haridasa,” Karnataka’s entire Vaishnava-bhakti poet-composer tradition — a genuine bridging opportunity requiring careful doctrinal fencing, not avoidance.
• Jude’s doxology (1:24-25) concentrates more Critical-risk terms per verse than any other passage in the letter (ಒಡೆಯ-adjacent ಕρ್ತ, ರಕ್ಷಕ, ಏಕಮಾತ್ರ, ನಿರ್ದೋಷಿಗಳಾಗಿ ನಿಲ್ಲಿಸು, the locked doxological quadruple, and the ಯುಗ-bracketing time formula) — exactly where a translator’s instinct toward elevated, “exalted-sounding” Kannada religious register is most likely to reach for Aikya-adjacent vocabulary.

Risks

• Aikya-drift in the doxology and “kept/blameless” language (Critical): God’s preserving power, presented blamelessness, and eternal life are the passages most vulnerable to unconsciously borrowed Veerashaiva mystical-union phrasing, precisely because the register is most elevated there.
• δεσπότης/κύριος collapse (Critical): defaulting either title to ಸ್ವಾಮಿ (guru/Jangama honorific) in Jude 1:4 would reframe Christ’s undivided sovereignty as one honored teacher’s authority among several — worse here than in Romans, since both titles appear in the same clause.
• ಭಕ್ತಿಹೀನ (“ungodly”) under-translation (High): Karnataka’s bhakti-saturated devotional culture risks this being heard as “insufficiently devout” rather than Jude’s actual charge of settled rebellion against God’s authority.
• ಕರುಣೆ (“mercy”) softened to generic devotional compassion (High): without judicial/rescue anchoring, this risks blending with Shakta-tradition compassion language (e.g., Chamundeshwari as Karuṇāmayi).
• ψυχικοί (“soulish”) rendered with ಆತ್ಮ-root vocabulary (High): would entangle Jude’s negative anthropological diagnosis with Vedantic ātman/paramātman categories, worsened by ಪವಿತ್ರಾತ್ಮ appearing one verse later.
• Enoch citation (1:14-15) misunderstood (High): risk of either implying 1 Enoch’s canonicity or, conversely, flattening it to the same status as Sharana vachana or Haridasa kirtane literature.

Opportunities

• ದಾಸ (servant) offers genuine, positive cultural bridging for teaching total, joyful belonging to Christ — provided the doctrinal distinction (purchased by the one Lord vs. chosen devotion to an iṣṭadevatā) is made explicit at first use.
• The letter’s universal, no-caste “common salvation” (1:3) resonates authentically with the Lingayat reform movement’s own historic rejection of caste hierarchy — a legitimate point of shared ground to acknowledge before drawing the necessary doctrinal distinctions.
• Jude’s vivid nature imagery (waterless clouds, fruitless trees, wandering stars) translates into plain, memorable Kannada without technical vocabulary risk, offering an accessible entry point for readers at the target Class 8-10 reading level.

Recommended actions

1. Lock the τηρέω/φυλάσσω consistency rule as a mandatory whole-document validation pass in Phase 2, run after all segments are translated, not merely per-segment.
2. Require human theologian sign-off on all 17 Critical and 21 High risk terms (38 total, 49% of the Jude term registry) before any Phase 2 segment reaches translation memory.
3. Require mandatory [TRANSLATOR NOTE] insertion at the eight flagged first-occurrence points identified in analysis/12_ai_translation_requirements.md (ದಾಸ, ನಂಬಿಕೆ’s dual sense, ಒಡೆಯ/ಕρ್ತ pairing, ಮಹಿಮೆಯುಳ್ಳವರು, the Enoch citation, ಪ್ರಾಪಂಚಿಕ ಸ್ವಭಾವದ, ಸಂದೇಹಪಡುವವರು, and the 1:25 doxological quadruple).
4. Brief all assigned theologian reviewers explicitly on the Aikya-drift risk before they review the doxology (1:24-25) and the “kept/eternal life/presented blameless” cluster, since this is the single highest-consequence, hardest-to-catch failure mode identified across this Language Package.
5. Carry forward all baseline Romans forbidden substitutions unchanged and add the seven new Jude-specific forbidden substitutions (ಸ್ವಾಮಿ for δεσπότης; ಸಾಧನೆ/ಸಾಧನ for “contend”; bare ಆತ್ಮ-root for “soulish”; ಕರ್ಮಫಲ for “judgment”; dropped ಒಮ್ಮೆಗೇ; ಅಲೆದಾಡುವವರು for “wavering”; bare ಯುಗ for the doxology’s time formula) into the Phase 2 system prompt without exception.

*Full detail: `analysis/07_semantic_analysis.md`, `analysis/08_core_glossary.md`, `analysis/06_linguistic_gap_analysis.md`, `analysis/11_doctrine_analysis.md`, `assets/translation_memory.json`, `assets/bible_term_registry.json`, `assets/doctrine_risk_registry.json`, `analysis/12_ai_translation_requirements.md`.*