Executive Summary — Titus — English → Somali Language Package

Why it matters

Titus extends the Romans Language Package into fresh, higher-stakes terrain: a short pastoral letter that grounds church order, leadership qualification, and household ethics directly in grace-alone salvation. Two of its features raise the doctrinal-risk profile above Romans’ already-substantial baseline: (1) Titus applies the title “Savior” identically to God the Father and to Jesus Christ six times, climaxing in 2:13’s single-article Greek construction naming Christ as “our great God and Savior” — one of the New Testament’s most direct grammatical assertions of Christ’s full deity, landing in maximal collision with tawhid; and (2) Titus 1:6’s elder qualification, “husband of one wife,” is not merely a conceptual collision but a practical one, directly confronting socially normal, legally endorsed Somali polygamy among leadership-eligible men. No comparable practical-behavioral collision exists anywhere in the Romans baseline.

Key findings

• 75 total terms now populate the Titus translation memory (22 inherited exactly from the Romans baseline, 53 newly identified for this book), of which 13 are Critical risk and 28 are High risk — 41 terms combined requiring mandatory human theologian review at every occurrence.
• 17 doctrines are documented for full-book coverage (4 Critical, 10 High, 3 Medium, 0 Low) — 14 require human theologian review, 3 require native speaker review, 0 are automated-only, a notably higher theologian-review burden proportionally than the Romans baseline given Titus’s shorter length.
• The core passage (Titus 2:11–3:8) alone carries five of the book’s thirteen Critical-risk terms (Savior, the great-God-and-Savior construction, husband of one wife — introduced just before the core passage at 1:6 but functionally load-bearing throughout — good works’ dual function, and regeneration), confirming it as the theological anchor the curriculum parameters intend, while chapters 1 and 3:9-15 outside the core passage still contribute substantial independent doctrinal load (elder/overseer qualifications, divisive-controversies vocabulary) that could not be safely omitted.
• Somali’s existing tarbiyad (child/moral formation), cibaado (ritual worship), camal (deeds-ledger), caqiido (systematic creed), tawba (repentance), bidco/bidcase (heretical innovation), and waayeel/odayaal (clan-elder institution) vocabularies each sit in a crowded semantic neighborhood already claimed by Islamic doctrinal or clan-social usage; none can be avoided outright without losing readability, so each requires deliberate, documented “fencing” through mandatory first-use teaching notes rather than lexical substitution alone.
• Genuine, positive cross-tradition resonance exists at two points — God’s mercy/kindness (Naxariis, Roonaanta; strong shared vocabulary with Rahman/Rahim-adjacent Islamic devotional language) and adoption/inheritance (via the Somali sheegad clan-incorporation custom) — both flagged as teaching bridges, not translation shortcuts.

Risks

• Christological risk (Critical): Splitting Titus 2:13’s single-referent “great God and Savior” construction into two separate, ranked entities during translation would silently erase, not merely weaken, the verse’s deity-of-Christ claim.
• Practical/behavioral risk (Critical): A faithful rendering of “husband of one wife” will functionally disqualify otherwise well-regarded, currently-practicing-polygamist candidate leaders in real congregations — a consequence with pastoral and social stakes beyond the translation itself.
• Terminological drift risk (Critical): “Good works” (Shuqullada Wanaagsan) must carry two theologically opposite functions — excluded as salvation’s ground (3:5), commanded as its fruit (2:14; 3:1, 8, 14) — using the identical term throughout; any drift toward two different Somali words, or toward the rejected camal deeds-ledger term, would either blur or reintroduce the very works-righteousness framework Titus 3:5 explicitly excludes.
• Regeneration/repentance conflation risk (Critical): Without sustained teaching, “Dib-u-dhalasho” (regeneration) risks silent substitution in readers’ minds with the familiar, human-initiated tawba (repentance) — a different causal structure (God acting on the person vs. the person acting toward God) that no lexical choice alone can prevent.
• Political/authority risk (High): “Rulers/authorities” (3:1) and household submission (2:5, 9) must be taught with explicit sphere-qualifiers in a country with decades of contested or absent central state authority, to avoid either sounding hollow or being heard as endorsing a specific armed faction’s legitimacy claim.

Opportunities

• Somali’s genealogical consciousness (abtirsiimo) and clan-incorporation custom (sheegad), already identified as bridges in the Romans baseline, extend naturally into Titus’s heirs/inheritance (3:7) and genealogies-avoidance (3:9) material — provided the “unprofitable” qualifier is never dropped from the latter.
• Titus’s mercy/kindness vocabulary (3:4-5) offers the single strongest positive cross-tradition resonance point in the book, given the devotional centrality of divine mercy (Rahman/Rahim) in Somali Islamic practice — a genuine, low-risk teaching bridge distinct from the book’s Critical-risk material.
• The letter’s household and civic virtue lists (1:6-9; 2:2-10; 3:1-2) read naturally in list form for Somalia’s strong oral-transmission culture, offering a low-friction on-ramp into the letter’s harder doctrinal core.

Recommended actions

1. Load assets/translation_memory.json (Titus) together with the Romans baseline TM before any Phase 2 segment processing; treat any apparent conflict as resolved in favor of the Romans baseline for genuinely shared concepts, and as a deliberately distinct new entry where Titus usage differs (e.g., Naxariis correctly serving eleos in 3:5).
2. Route Titus 2:11–3:8 as a single indivisible human-theologian review unit, not segment-by-segment, given its unmatched concentration of Critical-risk material.
3. Attach mandatory first-use-per-lesson teaching notes (not merely translator notes) for: the Savior title’s shared Father/Son application, the one-office elder/overseer clarification, grace-not-law as the trainer in 2:12, and the regeneration/tawba distinction in 3:5.
4. Never abbreviate “husband of one wife” and never drop the “unprofitable” qualifier from “genealogies” in any occurrence, in any document, at any point in Phase 2 or beyond.
5. Proceed to Phase 2 segment translation only after confirming all 75 translation-memory entries and all 17 doctrine-registry entries are loaded and version-logged per the enforcement instructions in analysis/12_ai_translation_requirements.md.

---

See assets/translation_memory.json and assets/bible_term_registry.json for the full term-level database. See assets/doctrine_risk_registry.json and analysis/11_doctrine_analysis.md for the full doctrine-level matrix. See analysis/04_comparative_theology.md for cross-tradition interpretive detail underlying every risk cited above.