Executive Summary: Ephesians (English → French)

Why it matters

Ephesians extends the Romans-derived French Language Package into six new, doctrinally dense chapters built around one core passage — Ephesians 2:1-10, salvation by grace through faith, not works — that sits at the exact historic Trent/Reformation flashpoint already flagged Critical in the Romans baseline. Unlike Romans, Ephesians never mentions a church building, introduces the letter’s own technical vocabulary (mystery, headship, fullness, the armor of God), and repeatedly collides with distinctly French cultural fault lines: the Jansenist predestination controversy, the Code Noir/Taubira-law memory around slavery language, and live #MeToo-era public debate over “soumission.” Getting this Language Package right determines whether French learners receive Paul’s actual argument or a domesticated, sacramentally- or secularly-flattened substitute.

Key findings

• 37 doctrines analyzed across all 6 chapters, full-book coverage confirmed chapter-by-chapter (no section silently skipped) — 15 Critical + 11 High = 26 doctrines require mandatory human theologian review; 9 require native speaker review; only 2 are automated-only.
• 142 terms recorded in assets/translation_memory.json (39 inherited exactly from the Romans baseline, 103 new to Ephesians) — 21 Critical-risk and ~40 High-risk terms (≈61 total) require mandatory theologian oversight, more than 40% of the entire glossary.
• Ephesians reuses seven baseline-Critical terms (grâce, foi, justice, salut, Seigneur, Église, Esprit Saint, Dieu, élection) but intensifies four of them to Critical status in this curriculum (foi, salut, Dieu, Esprit Saint) because Ephesians 2:5-9 and 4:30/5:18 press these terms harder than Romans does.
• Five terms are unique high-severity French collision points with no equivalent in the Romans baseline: croix, esclaves/maîtres, se soumettre, un seul homme nouveau, mystère (dual-locus at 1:9/3:3-9/6:19 and again at 5:32).
• The letter’s own internal French-lexicon problem — “œuvres” (2:9, excluded) vs. “bonnes œuvres” (2:10, affirmed) vs. “ouvrage” (2:10, workmanship) — is lexically unavoidable and requires a mandatory teaching gloss or Paul’s central argument visually collapses in French.
• Segond 21 (S21) is recommended as the reference YouVersion version (book code EPH); TOB/Bible de Jérusalem base text is a safe register model but their confessional footnote apparatus must never be imported into curriculum material.

Risks

• Doctrinal collapse risk: the grace/works antithesis (2:9-10) and the Trent-vs-Reformation grace tension (2:5-8) are the letter’s headline thesis, not a side point — any softening here undermines the entire curriculum’s core passage.
• Cultural offense risk (unique to France): “se soumettre” (5:21-24) and “esclaves/maîtres” (6:5-9) intersect live, high-visibility French social controversies (feminist critique of submission language; the Code Noir and 2001 Taubira law’s national reckoning with slavery). Mishandling either risks the material being received — or rejected — as endorsing what it does not teach.
• Catholic-liturgical collision risk: “mystère” is saturated with Mass-liturgy and Rosary associations; “esclaves” and “tête” likewise interact with existing devotional/catechetical instincts. Unglossed use of any of these terms risks readers importing the wrong referent by default.
• Secular-reception risk: “l’armure de Dieu” and the hostile-powers vocabulary of 6:12 face a two-sided failure mode in laïque French culture — demythologizing dismissal on one side, sensationalized occultism on the other — with no ready-made corrective in any existing French Bible translation.
• Invisibility risk: “plénitude” (1:23, 3:19, 4:13) and the Greek perfect tense underlying “sauvés” (2:5,8) are the opposite problem — not collision but being skimmed past without doctrinal weight registering at all.

Opportunities

• Ephesians 4:4-6’s sevenfold unity confession is already a familiar, liturgical-sounding French text (echoed in Catholic Mass and ecumenical liturgy) — an asset for verbatim-consistent memorization across the curriculum, parallel to the baseline’s Romans 10:9-10 rule.
• French’s legal/commercial vocabulary (“arrhes,” “sceller”) maps cleanly onto Ephesians 1:13-14’s sealing/guarantee imagery without invention.
• The Catholic sacramental resonance of “mystère” is a productive, not only risky, overlap — both senses concern divine reality disclosed rather than reasoned out — and can be leveraged pastorally if every occurrence carries Paul’s specific contextual gloss.
• A single, well-built historical-cultural teaching note for 6:5-9 (distinguishing Greco-Roman bondservice from transatlantic chattel slavery, foregrounding v.9’s radical equality claim) converts the curriculum’s highest-sensitivity passage into a strong pastoral asset rather than a liability.

Recommended actions

1. Route all 26 theologian-review doctrines and ~61 Critical/High terms through mandatory human theologian sign-off before any Phase 2 segment is marked approved — no automated pass-through for this tier.
2. Attach the five mandatory teaching notes (works-triad fencing at 2:9-10; mystery contextual gloss at every occurrence; headship/fullness cross-occurrence consistency; submission’s mutual-submission framing; slavery’s historical-cultural note) as non-negotiable Phase 2 build gates, not translator discretion.
3. Adopt Segond 21 as the fixed YouVersion reference version and confirm the numeric version ID against the live catalog before publishing any hyperlink.
4. Load assets/translation_memory.json (Ephesians) together with — never instead of — the baseline Romans translation_memory.json for every Phase 2 session, per analysis/12_ai_translation_requirements.md.
5. Brief all theologian reviewers on the Ephesians-specific French cultural amplifiers (Jansenism, Code Noir/Taubira law, #MeToo-era submission debate, Catholic sacramental-mystery/marriage doctrine) before review begins, since these are not covered by the Romans baseline’s reviewer briefing.

---

This summary synthesizes analysis/07_semantic_analysis.md, analysis/08_core_glossary.md, analysis/11_doctrine_analysis.md, analysis/04_comparative_theology.md, analysis/05_translation_landscape.md, analysis/06_linguistic_gap_analysis.md, assets/bible_term_registry.json, assets/doctrine_risk_registry.json, and assets/translation_memory.json (Ephesians edition). All figures are drawn directly from those artifacts and must be updated together if any of them change.