Executive Summary — Hebrews (English → Assamese)

Why it matters

Hebrews is the highest-density doctrinal-risk book this pipeline has processed for Assamese to date. Its central argument — Christ as the unique, permanent Great High Priest whose once-for-all sacrifice ends the entire Levitical sacrificial system — runs directly against Assam’s own living religious landscape: hereditary Brahmin ritual priesthood (পুৰোহিত), the active Shakta tradition of votive blood sacrifice (বলি) centered at Kamakhya, and Sankardev’s Ekasarana Dharma theology of Krishna’s cyclical avatar-descent. Unlike Romans, where syncretism risk is real but diffuse across many doctrines, Hebrews concentrates its highest risk into a single seventeen-verse core passage (9:11–28) that is also this curriculum’s theological anchor. Getting this book’s terminology and doctrinal notes right — or getting them wrong — has outsized consequences for whether Assamese readers correctly understand Christ’s priesthood as unique and finished rather than as one more instance of a familiar sacrificial, priestly, or avatar-descent pattern.

Key findings

• 32 distinct doctrines identified, fully mapped chapter-by-chapter across all 13 chapters of Hebrews (11_doctrine_analysis.md), extending the Romans doctrine framework without contradiction.
• 75 new theological terms introduced by this book beyond the Romans baseline, now seeded into assets/translation_memory.json alongside all 43 inherited Romans-baseline terms, for a combined enforced vocabulary of 118 terms.
• Of the 75 new terms: 11 are Critical risk and 30 are High risk — meaning 41 of 75 new terms (55%) require mandatory human theologian review at every occurrence, a substantially higher proportion than the Romans baseline.
• Combined with the baseline’s own 12 Critical and 20 High inherited terms, the full Hebrews-curriculum vocabulary carries 23 Critical-risk terms and 50 High-risk terms — 73 terms in total requiring theologian-tier enforcement discipline.
• At the doctrine level, the doctrine_risk_registry.json (Hebrews) records 11 Critical-risk and 14 High-risk doctrines — 25 of 32 doctrines (78%) are routed to mandatory human theologian review, versus 6 to native-speaker review and 1 to automated review only.
• The single highest-concentration risk zone in the entire book is the core passage itself, Hebrews 9:11–28, where sacrifice (বলি), atonement (প্ৰায়শ্চিত্ত), and Christ’s “appearing” language (দেখা দিয়া) all cluster and require repeated, not merely first-occurrence, translator notes.

Risks

• বলি (sacrifice) / প্ৰায়শ্চিত্ত (atonement) — highest risk in the book. No substitute Assamese word exists for either; both collide with a living, still-practiced Assamese religious tradition (Kamakhya-adjacent Shakta blood sacrifice; self-performed Hindu penance), and both run in the theologically opposite direction from Hebrews’ argument. Managed only through mandatory, repeated translator notes — a first-occurrence-only note is insufficient and has been explicitly rejected as a strategy.
• Christ’s “appearing” / second-coming language (Hebrews 9:24–28) — highest avatar-collapse risk. The passage’s own once/twice structure superficially mirrors Ekasarana Dharma’s cyclical Krishna-avatar theology. Requires দেখা দিয়া (never a প্ৰকাশ-root) with mandatory notes at every one of three appearings within five verses.
• মহাযাজক/যাজক vs. পুৰোহিত. A single lapse into পুৰোহিত anywhere in the curriculum would collapse Christ’s unique, non-hereditary, non-successive priesthood into a caste-based ritual-specialist role — the exact error the book’s central argument is built to prevent.
• Apostasy/warning passages (6:4–8, 10:26–31, 12:15–17). Circulating Common Language paraphrase literature has been observed softening “impossible” toward “difficult” — a bias this curriculum must not echo. Every occurrence is Critical-risk and theologian-routed regardless of apparent simplicity.
• সম্পূর্ণ কৰা vs. সিদ্ধ, শাসন vs. শাস্তি, যুগ vs. linear history — three further Critical/High collision points where the more literarily natural Assamese word is the wrong one (tantric-adept attainment, karmic retribution, and cyclical cosmic ages, respectively).

Opportunities

• Hebrews contains genuine positive teaching bridges not present in Romans: vicarious substitutionary sin-bearing (9:28) has no parallel in strictly individual, non-transferable karma doctrine, and the book’s radical claim of unmediated access to God (4:16, 10:19–22) is a countercultural claim relative to both caste-purity gatekeeping and Satradhikar guru-mediated devotional structures. Both should be flagged in accompanying teaching commentary as distinctive strengths of the gospel message for this audience, not softened for comfort.
• Ekasarana Dharma’s own historical reform against Brahmanical animal sacrifice offers a partial, teachable point of cultural resonance with Hebrews’ movement away from repeated sacrifice — carefully distinguished (Hebrews replaces sacrifice with one finished sacrifice; Ekasarana replaces it with non-sacrificial devotional practice) but pastorally useful as a bridge.
• The consolidated 08_core_glossary.md and this package’s translation_memory.json now give Phase 2 a single, internally consistent, doctrine-tested vocabulary for the entire book — no chapter was found theologically inert, and full-book coverage is confirmed end-to-end.

Recommended actions

1. Lock assets/translation_memory.json (118 terms) and 12_ai_translation_requirements.md before Phase 2 begins. No segment translation should proceed without both loaded.
2. Route all 25 theologian-tier doctrines and all Critical/High terms through mandatory human review, with special priority queuing for Hebrews 9:11–28 (the core passage), 6:4–8, 10:26–31, and 12:15–17 (warning passages), and 9:24–28 (appearing/second-coming cluster).
3. Enforce the recurring-note requirement (not just first-occurrence) for বলি, প্ৰায়শ্চিত্ত, and দেখা দিয়া within the core passage specifically — validated automatically per the checklist in 12_ai_translation_requirements.md.
4. Confirm the YouVersion Assamese version ID against the live catalog before generating any Phase 2 hyperlinks, per 05_translation_landscape.md; do not hardcode a guessed ID.
5. Carry forward every forbidden substitution (পুৰোহিত, যজ্ঞ, প্ৰতিমূৰ্তি, সিদ্ধ, শাস্তি, পিতৃ-লোক, any প্ৰকাশ-root “appearing,” and all Romans-inherited forbidden terms) into automated validation gates for every Phase 2 output file.

Cross-reference: assets/translation_memory.json, analysis/12_ai_translation_requirements.md, analysis/11_doctrine_analysis.md, assets/doctrine_risk_registry.json, assets/bible_term_registry.json for full supporting detail.