01 — Executive Summary: Joshua Language Package (English → Mandarin)

Why it matters

Joshua is the first curriculum built on top of the Romans baseline Language Package, and it stress-tests the baseline in a new genre: narrative conquest-and-covenant history rather than doctrinal epistle. The book’s seven core doctrines — God’s Faithfulness to the Land Promise, Holy War and Divine Judgment, Courage through God’s Presence, Rahab and the Inclusion of the Nations by Faith, Covenant Renewal at Shechem, Obedience and the Conquest, and Rest in the Promised Land as a Type of Salvation Rest — surface a dense new set of collisions with Chinese folk religion, Confucian virtue ethics, and Buddhist karma/cessation concepts that the Romans package did not need to address. Getting this right matters because the core passage, Joshua 24:14-15 (“choose you this day whom you will serve”), is the most quotable, most-taught verse pair in the curriculum and sits directly on top of the single highest-probability translation error identified in this analysis.

Key findings

• Term-level risk: the extended translation_memory.json now carries 63 enforced terms for Joshua (15 inherited from Romans + 48 new). Of these, 7 are Critical risk and 19 are High risk — 26 terms (41% of the glossary) requiring mandatory human theologian review on every occurrence.
• Doctrine-level risk: the doctrine_risk_registry.json (Joshua) documents 19 doctrines across all 24 chapters. 1 is Critical (Holy War and Divine Judgment) and 12 are High — 13 doctrines total requiring human theologian review, plus 5 Medium-risk doctrines (native speaker review) and 1 Low-risk doctrine (automated review only).
• Chinese Traditions are the dominant collision risk for every one of the seven core doctrines (comparative theology, Step 04) — a stronger and more consistent pattern than in the Romans baseline, where collisions were more evenly spread across Buddhist and Confucian categories.
• Full-book coverage confirmed: all 24 chapters were reviewed; chapters 12, 15, 16, and 17 contribute no new doctrine or term and are explicitly logged as “reviewed, not omitted,” satisfying the PRD Phase 1 mandate.
• No true vocabulary gaps requiring new transliteration were found; every proper noun already has a stable CUV form. The dominant risk profile is crowding — good existing Mandarin terms sitting next to strong, live competing associations (folk pantheon worship, temple divination, ancestor veneration, karmic retribution, feng shui, martial-deity devotion).

Risks

• Highest risk: 拣选 (God’s election of persons) being mistakenly substituted for 选择 (human choice) in Joshua 24:15 — the single most-quoted line in the curriculum — inverting the direction of agency at its rhetorical climax.
• Critical risk: 当毁灭的物 / 尽行毁灭 (herem/“utterly destroy,” chs. 6-11) being read as impersonal karmic retribution (报应) or generalized license for religious violence, rather than a bounded, historical act of divine judicial judgment.
• High risk: 安息 (“rest”) defaulting to the standard Chinese Christian/funerary phrase 安息主怀 (death repose) or to Buddhist cessation concepts (涅槃/寂灭) before its this-worldly, covenantal sense is ever established.
• High risk: 产业 (“inheritance/portion”) collapsing into its modern homonym “industry/business assets,” severing the doctrine’s grace-not-merit force.
• High risk: 别神/外邦的神 (“foreign gods”) and 忌邪的神 (“jealous God”) being read henotheistically or as folk-deity status competition, softening the core passage’s exclusivity claim.

Opportunities

• The Romans baseline’s Critical-tier grace/merit distinction (恩典 vs. 功德) transfers directly and reinforces the land-inheritance-as-gift doctrine (chs. 13-21) — a ready-made teaching bridge requiring no new vocabulary.
• Joshua 24’s suzerain-vassal covenant treaty form is a widely recognized literary structure that can be taught cross-confessionally, giving native-speaker reviewers a stable non-theological entry point before the exclusivity claims are introduced.
• Rahab’s narrative offers a strong, deliberately provocative teaching opportunity against Confucian status/face hierarchies — a collision to be leaned into pastorally, not smoothed over.
• The Joshua/Jesus name typology (both Ἰησοῦς in the LXX) is low-risk textually (CUV’s 约书亚 and 耶稣 are fully distinct) and can be taught as a pedagogically rich, low-risk bridge into Christology once properly fenced.

Recommended actions

1. Add the 选择/拣选 forbidden-substitution rule to the Phase 2 automated validator at Critical tier, with a standing three-verse flag radius around Joshua 24:15 (done in 12_ai_translation_requirements.md v1.28).
2. Route all segments in Joshua 6, 7, 8, 10, 11, and 24 to human theologian review by default, given their concentration of Critical/High-risk terms — do not rely on per-term flagging alone.
3. Enforce the rest-sequencing rule (this-worldly land-rest before eschatological typology) as a hard lesson-ordering constraint in Phase 3 curriculum design, not merely a translation note.
4. Carry forward the 7 Critical and 19 High risk terms into Phase 2 reviewer staffing plans now, so theologian review capacity is budgeted before translation begins rather than discovered mid-pipeline.
5. Brief native-speaker reviewers explicitly on the Rahab occupation non-euphemism rule and the jealous-God/folk-deity-competition distinction, since both require active resistance to a “smoothing” instinct that would otherwise feel like normal politeness in Mandarin register.

See assets/translation_memory.json (v2) for the full enforced term database, assets/doctrine_risk_registry.json (Joshua) for the doctrine matrix, and analysis/11_doctrine_analysis.md for full chapter-by-chapter traceability.