Executive Summary: Titus Language Package (English → Korean)

Why it matters

Titus packs the curriculum’s seven target doctrines — grace that trains for godliness, elder qualifications, sound doctrine and good works, salvation by grace not works, regeneration by the Spirit, submission to authority, and avoiding divisive controversies — into three short chapters, with the densest concentration in the core passage (Titus 2:11–3:8). Korea is the best-positioned language in this pipeline for doctrinal accuracy, given over a century of settled, indigenous Protestant Bible vocabulary and a majority-Christian population among believers. But that same maturity means the risk here is not missing words — it’s quiet syncretistic drift: well-worn Korean words that already carry a strong competing meaning from shamanism (무속신앙), Korean Buddhism, or Confucian social ethics can be reached for on autopilot and silently reshape the text’s meaning.

Key findings

• One true lexical collision, not just a connotation risk: 중생 (regeneration, Titus 3:5) is a hard homophone with 衆生, one of Korean Buddhism’s most common words (“all sentient beings”). This is the single highest-priority translation risk in the entire book and requires a structural rule (never standalone; always 중생의 씻음), not a discretionary word choice.
• Titus intensifies, rather than introduces, the Romans baseline’s sharpest risk term: 은혜 (grace) is grammatically personified as an active “trainer” (2:12) and set directly against a negated works-clause (3:5) — the clearest anti-기복신앙 (fortune-seeking faith) statement anywhere in the curriculum.
• Titus 2:13 is the clearest single deity-of-Christ statement in the letter (“our great God and Savior Jesus Christ”), and the Korean rendering must preserve its single-referent grammar or the affirmation weakens.
• Titus 2:5’s wife-submission instruction is the sharpest culturally-specific collision in the book, sitting directly against Korea’s historic, now-repudiated Confucian “three obediences” doctrine (삼종지도).
• Two terms carry inverted-valence risk: 족보 (genealogies, 3:9) is positively loaded in Korean culture where Paul is dismissive; 이단 (divisive person, 3:10) is far more institutionally loaded in modern Korean church discourse than Paul’s narrow, local, two-warning scenario.
• Full-book coverage confirmed: all three chapters contribute load-bearing vocabulary; no section was silently omitted (see 11_doctrine_analysis.md Part C).

Risks

Of 24 doctrines identified for Titus: 6 Critical, 9 High, 8 Medium, 1 Low — 15 requiring mandatory human theologian review, 8 native speaker review, 1 automated-only (doctrine_risk_registry.json). Of the terms newly introduced by Titus (not inherited from Romans): 4 Critical, 13 High, ~27 Medium, ~6 Low, plus 7 low-risk proper names, layered on top of 7 Critical, 12 High, 5 Medium, 3 Low terms inherited unchanged from the Romans baseline and directly relevant to Titus. In total, 11 Critical-risk terms and 25 High-risk terms in the combined Titus translation memory require mandatory human theologian review on every occurrence — the highest concentration falling inside the six-verse span of Titus 3:4–3:8, the doctrinal climax of the letter.

Opportunities

• Korea’s existing Bible-translation maturity means paraphrase-plus-fencing beats new coinage everywhere in this book except proper names — no term requires an invented word, only a disambiguating teaching note.
• The Korean 아빠/Abba and 언약/covenant-contract distinctions that already serve the Romans curriculum well continue to serve Titus without modification.
• Titus’s compact length makes it practical to schedule the entire core passage (2:11–3:8) as a single concentrated theologian-review unit rather than distributing it across ordinary batch processing, maximizing review quality where risk is highest.
• The letter’s grace-inclusio (1:4/3:15) and Savior-title alternation (six occurrences) give Phase 2 a built-in internal consistency check: if 은혜 and 구주 render identically at every occurrence, translation-memory discipline across the whole book can be verified at a glance.

Recommended actions

1. Encode the 중생 structural rule (never standalone; always 중생의 씻음) as a hard automated pre-submission validation check in Phase 2, equal in enforcement priority to the Romans baseline’s forbidden-substitution list — not left to discretionary review.
2. Route Titus 2:11–3:8 as a single, unhurried, theologian-reviewed unit rather than ordinary distributed batch segments.
3. Attach mandatory, non-optional fencing notes to every occurrence of: 복종하다 in 2:5 (against 삼종지도), 선한 일 throughout (against 공덕), 자비 in 3:4 (against Buddhist karuna/metta), 족보 in 3:9 (against positive-valence over-reading), and 이단에 속한 사람 in 3:10 (against modern institutional-label over-reading).
4. Carry forward all 27 Romans-baseline terms relevant to Titus without modification; do not re-derive or re-negotiate any baseline rendering.
5. Finalize and publish the updated translation_memory.json and doctrine_risk_registry.json (both delivered in this step) as the enforced Phase 2 databases, superseding no baseline entry but extending it in full.