Executive Summary: Titus (English → Urdu Language Package)

Why it matters

Titus compresses this pipeline’s sharpest doctrinal-collision material into three short chapters. It contains the single most explicit deity-of-Christ statement in this curriculum (2:13’s Granville-Sharp “our great God and Savior Jesus Christ” — more direct than Romans 9:5), the single most unqualified salvation-by-grace statement (3:5’s blanket denial that any human righteous deed contributes to salvation), and this pipeline’s first-ever regeneration doctrine (3:5-6), whose most natural Urdu “washing” word (غسل) happens to name a currently, universally practiced Islamic ritual. Titus also introduces a risk category the Romans/Galatians baseline flagged only in passing: collision with live, currently enforced Islamic legal and devotional categories — Pakistan’s blasphemy law, the ulī’l-amr political-obedience doctrine, hadith-authentication vocabulary, and Sufi shrine-veneration vocabulary for “elder” — not only abstract creedal claims. Getting Titus right matters more for translator/reviewer safety, not just doctrinal accuracy, than any other book processed so far in this family.

Key findings

• 71 terms are now recorded in the Titus translation memory: 26 inherited exactly from the Romans/Galatians baseline and 45 newly seeded for this book. Of the 71, 45 (19 Critical + 26 High) require mandatory human theologian review before Phase 2 translation proceeds.
• 13 doctrines are documented in the doctrine risk registry: 5 Critical, 8 High, 3 Medium, 0 Low — meaning 13 of 13 non-Low doctrines route to either human theologian (13) or native speaker (3) review; there is no doctrine in this book cleared for automated-only review.
• Titus’s risk pattern is structurally different in kind from Romans/Galatians: most of its risk is not missing vocabulary (Urdu has adequate words for nearly everything Titus says) but semantic overcrowding — the correct Urdu word sits inside a semantic field already dominated by a specific, named, currently-practiced Islamic technical term (غسل for ritual washing, اطاعت for political obedience, صحیح for hadith authentication, بدعتی for doctrinal innovation, بزرگ for Sufi sainthood, پاک/ناپاک for fiqh ritual purity).
• Two genuinely new terms with no prior baseline anchor were coined: نئی پیدائش (regeneration) and نئی پیدائش کا دھلاؤ (washing of regeneration) — the seed vocabulary for a doctrine (Regeneration by the Holy Spirit) that does not appear anywhere in the Romans/Galatians curriculum.
• The σωτήρ (Savior) title recurs six times (1:3, 4; 2:10, 13; 3:4, 6), deliberately alternating between God the Father and Jesus Christ — a Trinitarian data-point that must be tracked as one linked unit across Phase 2’s parallel-processing pipeline, not six independent segments.
• The good-works basis/fruit distinction (3:5 excludes works from salvation’s basis; 3:8 commands them emphatically as fruit) recurs six times (1:16; 2:7, 14; 3:1, 8, 14) and must be re-flagged at every individual occurrence, since a reader encountering any single occurrence in isolation could otherwise draw the opposite conclusion from either half of the pair.

Risks

• Highest risk in the book, and in this Language Package family to date: Titus 2:13 identifies Jesus Christ directly as “the great God and Savior” — a grammatical identity-claim (not two coordinate titles) colliding at the sharpest possible point with Qur’an 5:72-75 and 4:171. No footnote can soften this without falsifying the text; mishandling here is both a translation-accuracy risk and a reader-safety/reception risk.
• Live legal-safety risk, unique to this book: rendering βλασφημέω (2:5, 8, 10; 3:2) with كفر-family vocabulary would import Pakistan’s criminally-enforced blasphemy-law category into a verse about believers’ reputational conduct — a real-world danger, not merely a doctrinal imprecision.
• Live ritual-collision risk, unique to this book: the single most natural-sounding Urdu word for Titus 3:5’s “washing” is غسل, the specific Islamic major-ablution term; using it would present regeneration as a ritual rite parallel to a currently practiced sunnah, precisely the syncretism this Language Package exists to prevent.
• Live political-doctrine risk, unique to this book: rendering ὑποτάσσω (2:5, 9; 3:1) with اطاعت would present Paul’s gospel-witness-motivated civic instruction as an endorsement of the specific Qur’anic ulī’l-amr obedience doctrine (4:59).
• Cumulative, not single-point, risk: the good-works basis/fruit distinction and the six-fold σωτήρ pattern are structurally vulnerable to segment-by-segment parallel processing, where a technically correct individual rendering can still lose a book-level doctrinal pattern if reviewed in isolation.

Opportunities

• Several terms offer genuine, low-friction bridge points worth using pastorally: ظاہر (God’s manifestation, echoing al-Ẕāhir), خدا کا بندہ (bandah-e-khuda, positively resonant servant-of-God language), رحمت (shared divine-mercy vocabulary), and the Qur’an’s own affirmed expectation of Isa’s return (a genuine point of contact for teaching 2:13’s blessed hope, provided its very different content is taught alongside it).
• Titus’s household and civic-order material (2:2-10; 3:1-2) is, on the whole, the lowest-friction content in the book for this audience — an opportunity to build reader trust and comprehension before addressing the book’s Critical-risk material.
• The basis/fruit good-works structure, taught explicitly and repeatedly, offers a constructive, non-confrontational way to engage the mīzān deeds-at-judgment framework familiar to Urdu readers without requiring a polemical framing.

Recommended actions

1. Route Titus 2:13, 2:14, and 3:4-7 (as a single linked unit) to mandatory human theologian review with explicit pastoral-framing sign-off before any Phase 2 translation of these verses is marked complete.
2. Implement automated (regex-style) validation at Phase 2 QA specifically for the two binary forbidden-substitution risks most likely to be silently reintroduced by translator instinct: غسل (never for “washing of regeneration”) and اطاعت (never for “submission to authority”).
3. Configure Phase 2’s segment-caching to process the six σωτήρ occurrences and the six نیک کام occurrences as linked cross-verse review units routed to a shared reviewer/reviewer pool, not as independent parallel-processed segments.
4. Run one consolidated theologian review pass across all purity-vocabulary occurrences (1:15, 2:5, 3:5) rather than reviewing each in isolation, given four distinct but easily-conflated Greek purity roots competing for the same Urdu پاک-root vocabulary.
5. Brief all Phase 2 reviewers on the two book-specific safety dimensions unique to Titus — Pakistan’s blasphemy-law collision (2:5, 8, 10; 3:2) and the ulī’l-amr political-obedience collision (3:1) — as reviewer-safety issues, not only translation-accuracy issues.

---

Prepared per PRD Phase 1 Step 11. Fully consistent with, and derived from, assets/translation_memory.json, assets/bible_term_registry.json, assets/doctrine_risk_registry.json, analysis/11_doctrine_analysis.md, analysis/04_comparative_theology.md, analysis/06_linguistic_gap_analysis.md, and analysis/12_ai_translation_requirements.md (all Titus, v1).