Executive Summary: Titus → Maithili Translation Requirements

Why it matters

Titus is short — three chapters, 46 verses — but doctrinally dense: nearly every section carries load-bearing theological weight, and it introduces the single highest-risk lexical decision found anywhere in this pipeline’s Maithili work so far. The letter’s core passage (2:11–3:8) compresses grace’s appearing, Christ’s deity, salvation apart from works, and Spirit-wrought regeneration into 22 verses, sitting directly on top of Maithili’s Hindu-majority karma-merit and rebirth vocabulary. Getting this book right sets precedent for how “new birth,” “good works,” and “elder qualifications” will read across the whole future Maithili curriculum — there is no existing settled Maithili Christian tradition to fall back on, so these choices function as first-draft doctrine, not routine translation.

Key findings

• 7 new theological terms required outright coinage (no existing Maithili word was adequate without creating unwanted syncretism): epiphany/appearing, training, godliness, ungodliness, good works, regeneration, and the elder/overseer office pair — each documented in analysis/06_linguistic_gap_analysis.md.
• Regeneration (Titus 3:5) is the single highest-risk term in the entire Titus package. The Greek παλιγγενεσία is literally “again-birth”; a naive gloss lands directly on पुनर्जन्म, the Hindu reincarnation term already permanently forbidden in the Romans baseline. नवजन्म was coined specifically to break this collision.
• The translation memory now contains 89 enforced terms for Titus (17 inherited exactly from the Romans baseline, 72 new). Of these, 20 are Critical risk and 29 are High risk — 49 terms total requiring mandatory human theologian review before any Phase 2 segment reaches native-speaker or automated review.
• The doctrine risk registry adds 15 named doctrines for Titus (7 Critical, 8 High/Medium combined per the registry’s own risk summary), bringing 15 doctrines requiring mandatory human theologian review and 1 requiring native-speaker review — with zero doctrines cleared for automated-only review in this book.
• Titus 2:13 (“our great God and Savior Jesus Christ”) is flagged as one of the New Testament’s clearest single-verse affirmations of Christ’s full deity; Maithili’s grammar must preserve the single-article construction so the phrase cannot be misread as naming two beings.
• Two culturally load-bearing collision points are specific to this region and this book: “husband of one wife” (1:6) directly confronts documented historic Maithil Brahmin kulin polygamy custom, and “genealogies” (3:9) directly names Panjikaran, Mithila’s own living genealogical-registry institution — neither is a hypothetical risk, both are active regional social practice.

Risks

• Karma-merit re-importation is the book’s dominant systemic risk: “good works” (नीक काज, six occurrences), “regeneration” (नवजन्म), and the 3:5 works-of-righteousness negation all sit on the same fault line, where a single inconsistent rendering at any occurrence could quietly reintroduce a merit-accumulation logic the letter’s own argument explicitly excludes.
• Guru-śiṣya (guru-disciple) absorption risk for “grace that trains” (2:11-12): without careful syntactic preservation, Maithili readers could hear grace’s training work through the region’s own live teaching-lineage (paramparā) framework, making a human teacher rather than grace itself the implied agent.
• Avatar-descent absorption risk for “epiphany/appearing” (2:11, 2:13, 3:4): Mithila’s own geographic and narrative ownership of the Ramayana makes this collision more emotionally resonant here than in most other languages in this pipeline.
• Softening risk on two culturally sensitive qualifications: both “husband of one wife” and the household-slave instructions (2:9-10) carry real potential to be diplomatically vague-ified by well-meaning translators wanting to avoid social friction — precisely the failure mode the requirements document is built to prevent.

Opportunities

• Mithila’s shared Sanskrit tatsama vocabulary with Hindi allows disciplined adaptation of well-tested Hindi-precedent term choices rather than pure invention, per the Translation Landscape findings — the task is informed adaptation, not starting from zero.
• The region’s “Atithi Devo Bhava” guest-honor custom offers a genuinely positive cultural bridge for the elder qualification of hospitality (φιλόξενος), usable without doctrinal distortion.
• Mithila’s strong patrilineal inheritance customs give “heir” (उत्तराधिकारी, 3:7) unusually strong positive resonance for teaching full spiritual inheritance rights, mirroring the “adoption” bridge already established in the Romans baseline.
• Titus’s own health-metaphor for “sound doctrine” (स्वस्थ शिक्षा) is a memorable, teachable image worth preserving deliberately rather than flattening to generic “correct teaching.”

Recommended actions

1. Route all 15 Critical/High Titus doctrines and all 49 Critical/High Titus terms through mandatory human theologian review before Phase 2 translation is considered complete — no exceptions, per assets/doctrine_risk_registry.json.
2. Apply the mandatory translator notes at every occurrence of regeneration, epiphany/appearing, and Titus 2:13’s deity-of-Christ construction, with zero-exception enforcement built into the Phase 2 validation checklist.
3. Commission joint theologian + native-speaker review specifically for “husband of one wife” (1:6) and the household-slave instructions (2:9-10), given their direct collision with documented regional social custom.
4. Enforce byte-for-byte consistency of नीक काज across all six occurrences (1:16; 2:7,14; 3:1,8,14) as a standing Phase 2 validation rule, since repetition is itself part of this doctrine’s pedagogical design.
5. Carry forward assets/translation_memory.json (Titus) and analysis/12_ai_translation_requirements.md (Titus extension) as living documents for any future Pastoral Epistles (1–2 Timothy) added to this same Language Package, since several Titus-specific formulas (e.g. “the saying is trustworthy”) are explicitly designed for reuse there.