Linguistic Gap Analysis
Linguistic Gap Analysis: Exodus
English → Bengali | Doctrine Vocabulary Matrix, Semantic Neighborhood Fencing, Transliteration Decisions, Ranked Ambiguities
Curriculum: Exodus (full book, chapters 1–40)
Core passage: Exodus 12:1–14
Destination language: Bengali (বাংলা)
Governing authority: This analysis extends the Romans baseline (translation_memory.json, bible_term_registry.json, doctrine_risk_registry.json) and is the analytical rationale layer underneath 08_core_glossary.md and the Exodus bible_term_registry.json. Nothing here contradicts those artifacts; where this document identifies a term not yet formally registered, it is flagged as a forward gap for the next glossary revision, not a live contradiction.
0. Method and Scope
Exodus poses a linguistically distinct challenge from Romans: Romans is doctrinal argument in Koine Greek addressed to an already-Christian congregation; Exodus is Hebrew narrative-and-law addressed first to Israel, later read by the church as the primary Old Testament type of Christ’s redemptive work. Two consequences follow for Bengali:
- Bengali has no native Old Testament translation tradition independent of the Bible Society editions. Every rendering choice here either draws on that existing Bengali OT tradition (সদাপ্রভু, নিস্তারপর্ব, ব্যবস্থা) or must coin/extend vocabulary using the same registers those editions use.
- Exodus’s subject matter — blood sacrifice, a name-revealing God, a portable sanctuary, priestly mediation — sits directly on top of the most heavily populated semantic real estate in Bengali religious vocabulary, because Bengali religious Sanskrit (tatsama) vocabulary was shaped first and most densely by Hindu Vedic and Puranic sacrificial, temple, and avatar theology, and second by Islamic legal and devotional vocabulary. Exodus therefore generates more crowded-neighborhood risk per chapter than Romans did.
This document is organized in four sections per the Phase 1 Step 8 mandate: (1) doctrine vocabulary matrix, (2) missing vocabulary vs. crowded neighborhoods needing fencing, (3) transliteration vs. paraphrase decisions, (4) ranked highest-risk ambiguities. A full-book chapter coverage table closes the document.
1. Doctrine Vocabulary Matrix
For each of the eight curriculum doctrines: the Bengali vocabulary currently available, its specific weaknesses, and the recommended strategy already encoded (or to be encoded) in the glossary/registry.
1.1 The Exodus as Redemption
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| মুক্তি (liberation) | Doubly compromised: (a) baseline’s existing Hindu-moksha concern; (b) unique to this curriculum, মুক্তি is the operative word in মুক্তিযুদ্ধ (“Liberation War,” 1971) and মুক্তিযোদ্ধা (“freedom fighter”) — Bangladesh’s foundational national-political narrative. Using it for God’s redemption of Israel would cause the text to be heard as political liberation theology or nationalist allegory before it is heard as a theological claim. | Forbid outright. Use উদ্ধার / উদ্ধার করা throughout. Where the ransom/price-paid nuance of גָּאַל must be made explicit (e.g., 6:6, 15:13), extend to মূল্য দিয়ে উদ্ধার করা (“redeemed by paying a price”). |
| উদ্ধার (rescue/deliverance) | Slightly under-specifies the kinsman-redeemer and price-paid nuance carried by גָּאַל versus generic rescue (הִצִּיל). | Retain as primary term; supply the price-paid compound only where the Hebrew root itself is גָּאַל, not generic rescue verbs. |
| পরিত্রাণ (salvation) | Already fixed Critical-risk baseline term; correctly reserved for the theological category “salvation,” not the historical-narrative verb “redeem/deliver.” | Use পরিত্রাণ only at 14:13 and comparable summary statements, not as a substitute verb for উদ্ধার করা throughout the plague/Exodus narrative. |
Gap: Bengali lacks a single word carrying both “rescue” and “kinsman with a legal claim/price to pay” simultaneously, the double sense of גָּאַל. This is a genuine lexical gap, not merely a collision risk; the compound phrase is the only adequate fix, and should be flagged for consistent use at every גָּאַל occurrence (6:6; 15:13; also anticipatorily for later curricula at Isaiah 43–44, Ruth).
1.2 Passover and Substitutionary Atonement
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| প্রায়শ্চিত্ত (atonement) | This is the standard Bengali Sanskrit-derived word for self-performed penance/expiation to remove sin or bad karma (present in both Hindu and lay-Bengali usage). Using it risks the reader supplying “atonement = something I do” precisely where the text teaches “atonement = something God provides via a substitute.” | Retain প্রায়শ্চিত্ত (no better word exists) but attach a mandatory boundary note at every occurrence distinguishing divinely-provided substitutionary atonement from self-achieved expiation — structurally identical to the baseline’s existing অনুগ্রহ/পুণ্য boundary. |
| বলি (sacrifice/offering) | Generic; by itself does not distinguish substitutionary killing-in-place-of from a general offering/gift presented to a deity (common across Hindu পূজা vocabulary). | Use compounds: বলি করা (verb, “to sacrifice”) for שָׁחַט; পাপার্থক বলি, সম্পূর্ণ দগ্ধ বলি for specific offering types; avoid bare বলি when the substitutionary logic is the specific point being taught. |
| জবাই করা (ritual slaughter) | Strong halal/culinary-butchery register in contemporary Bengali (both Hindu qurbani-adjacent and Muslim everyday usage); flattens the cultic-substitution sense into food preparation. | Reject for doctrinal contexts; reserve বলি করা. |
| প্রতিমা / সোনার বাছুর (idol/golden calf) | প্রতিমা is the everyday word for a Hindu deity-image, most vivid during Durga Puja season — an unavoidable, live collision. | Do not soften. The text’s own anti-idolatry polemic is sharpened, not dulled, by naming the golden calf with the same word Bengali readers use for image-worship they may practice or witness; retain with a contextual note rather than seeking a neutral substitute that does not exist. |
Gap: There is no existing Bengali term that transparently conveys “an innocent substitute dies in the place of the guilty” without importing either the karma-penance frame (প্রায়শ্চিত্ত) or the Vedic fire-ritual frame (হোম, যজ্ঞ — both correctly avoided per the registry). The descriptive-compound strategy (সম্পূর্ণ দগ্ধ বলি, বলিদানের বেদি) is a paraphrase-by-necessity, not a stylistic preference.
1.3 The Character and Name of God (I AM)
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| সদাপ্রভু (YHWH) | No inherent weakness — this is the single strongest asset in the entire Exodus vocabulary set, an established, theologically load-bearing Bengali OT term with no competing religious claimant. | Use without exception for YHWH; never conflate with ঈশ্বর (Elohim) or প্রভু (Adonai). |
| আমি যে আছি সেই আমি (I AM WHO I AM) | Heavy risk of being heard as a philosophical statement about Being-in-general, which both Vedantic (Brahman as sat) and Islamic (Al-Qayyūm) frameworks in Bengal are fully equipped to absorb and reinterpret on their own terms, quietly erasing the personal, historical, covenant-specific self-revelation the text intends. | Mandatory translator note anchoring the phrase in God’s continuing, active, covenant-keeping relationship with Abraham/Isaac/Jacob and now Israel — never presented as free-standing ontology. |
| সর্বশক্তিমান ঈশ্বর (El Shaddai) | Contains the শক্তি root the baseline forbids in Romans 1:16 contexts, but functions here as a frozen liturgical compound already naturalized in Bengali Christian usage (creed, Lord’s Prayer tradition). | Retain; flag for reviewer awareness only, not as a forbidden-substitution violation — an important distinction to record explicitly so Phase 2 automated validators do not false-flag it. |
| ঈর্ষাপরায়ণ (jealous) | ঈর্ষা is the ordinary Bengali word for petty envy/resentment between people (including marital jealousy), a strongly negative connotation in everyday register. | Retain for lack of a better single word, but require a translator note every occurrence distinguishing covenant-exclusive zeal from human envy. |
| করুণাময় ও কৃপাময় (compassionate and gracious, 34:6) | Low risk given the baseline’s own carve-out permitting কৃপা in non-soteriological contexts; risk is only that a careless reader collapses কৃপাময় here with অনুগ্রহ (grace) elsewhere. | Reuse as registered; boundary note required distinguishing this general-mercy usage from soteriological grace. |
Gap (forward-looking, not yet in registry): Exodus contains additional compound revelatory names of God beyond El Shaddai that this curriculum’s core-doctrine list does not explicitly name but which appear in the text and will surface in teaching: “the LORD who heals you” (יהוה רֹפְאֶךָ, Exodus 15:26) and “The LORD is my Banner” (יהוה נִסִּי, Exodus 17:15, the name Moses gives the altar after defeating Amalek). Recommendation: extend the registry with সদাপ্রভু-compound renderings (সদাপ্রভু, আমার আরোগ্যকারী / সদাপ্রভু, আমার বিজয়-পতাকা) rather than leaving these to be improvised ad hoc in Phase 2, since they belong to the same YHWH-name theological cluster as I AM and El Shaddai and carry comparable Critical-adjacent risk if mistranslated as generic titles.
1.4 The Mosaic Covenant and Law
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| ব্যবস্থা (Law/Torah) | Fixed baseline term, well-insulated from ধর্ম and শরিয়ত. Weakness is not lexical but scope: Exodus uses ব্যবস্থা for the whole Sinai legislation, while distinguishing sub-categories (commandments, ordinances) that Bengali must also keep visually and lexically separate. | Maintain three-tier distinction: ব্যবস্থা (whole Law) / আজ্ঞা (individual commandment) / বিধি (case-law ordinance, mishpatim). Teaching material should state this three-tier structure explicitly at least once, since Bengali readers accustomed to a single word for “religious law” (whether ধর্ম or শরিয়ত) will not default to a three-way distinction unprompted. |
| নিয়ম (covenant) | Fixed baseline term; strong asset (already stabilized via পুরাতন নিয়ম, “Old Testament”). Weakness: নিয়ম also means ordinary “rule” in everyday Bengali, so relational-covenant weight can be lost without reinforcement. | Where covenant’s relational, promissory character is the specific teaching point (chs. 19, 24, 34), pair with descriptive language (“সদাপ্রভুর সাথে ইস্রায়েলের নিয়ম-সম্পর্ক”) rather than relying on নিয়ম alone to carry that weight. |
| পবিত্র জাতি / যাজকীয় রাজ্য (holy nation / kingdom of priests) | জাতি (nation) risks bleeding into the baseline’s পরজাতি (gentiles/other nations) if not visually distinguished; রাজ্য (kingdom) risks a political misreading. | Maintain exactly as registered; explicit teaching note that Israel-as-জাতি (this nation, chosen) is the deliberate mirror-opposite of পরজাতি (other nations), not an accidental overlap. |
Gap: No single Bengali word covers the semantic range of the English “law” that spans “Torah as covenant-constitution,” “individual commandment,” and “case-law statute” — but this is a manageable three-term fencing solution already captured in the registry, not an unfilled gap.
1.5 The Tabernacle and God’s Presence
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| সমাগম তাম্বু (tabernacle/tent of meeting) | No existing single-word Bengali Bible precedent as strong as সদাপ্রভু; this is a constructed compound (“tent of meeting/assembly”), transparent but not yet as deeply naturalized. | Retain and use with total consistency chs. 25–40; never substitute মন্দির (temple) even in casual paraphrase, since মন্দির is specifically the fixed, permanent Hindu temple-shrine concept the mobile tent structure is designed to contrast with. |
| নিয়ম-সিন্দুক (ark of the covenant) | Established, low risk. | Retain. |
| ঈশ্বরের মুখ / ঈশ্বরের উপস্থিতি (face/presence of God) | The devotional concept nearest at hand for a Bengali Hindu reader is দর্শন (darshan) — viewing a deity’s image, understood as auspicious and safe. Exodus 33:20 teaches the opposite: unmediated sight of God’s face is fatal. | Mandatory translator note at 33:18–23 explicitly contrasting দর্শন-as-safe-viewing with the text’s own claim that God’s unmediated presence cannot be safely viewed apart from mediation — this is a substantive theological collision point, not a minor stylistic one. |
| যাজক / মহাযাজক (priest/high priest) | যাজক is reasonably neutral, but readers may default-map it onto either a Hindu পুরোহিত (temple ritual officiant, mantra-based) or, more distantly, an Islamic ইমাম (prayer-leader, non-sacrificial). Neither role carries the specific blood-sacrifice mediatorial function যাজক must convey here. | Teaching material (not the term itself) must carry the functional definition; the term is adequate but underdetermined without it. |
| অভিষেক করা (anoint) | Also the standard word for Hindu deity-consecration and royal coronation. | Retain (no viable neutral alternative exists, and the baseline already accepts this class of retained-despite-origin term, e.g. ঈশ্বর itself); flag the productive positive link to খ্রীষ্ট (“the Anointed One”) as a teaching opportunity rather than only a risk. |
Gap: Bengali has no word for “a holy, mobile, tent-based meeting place distinct from both a house and a permanent temple” that does not require construction from scratch. সমাগম তাম্বু is a necessary coinage-by-compound, and its comparative novelty (versus সদাপ্রভু’s centuries of established use) means Phase 2 must apply it with unusual internal consistency, since there is no existing reader-familiarity to fall back on if usage drifts.
1.6 Deliverance from Bondage to Sin
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| দাসত্ব (bondage/slavery) | Literal sense (Israel’s slavery under Pharaoh, chs. 1–14) is unambiguous. Risk is entirely at the metaphorical extension: Bengali readers, especially in Bangladesh, may default to reading the whole book through a social/political liberation lens (given দাসত্ব’s contemporary resonance with historical indentured and bonded-labor discourse) and never make the move to “bondage to sin” the curriculum doctrine requires. | The literal-to-spiritual metaphorical extension must be made explicit in teaching notes at every occurrence where the curriculum intends the doctrinal (not merely narrative) sense; the term itself cannot carry this shift unaided. |
| পাপ (sin) | Fixed baseline term, stable. | No new weakness in Exodus; golden calf narrative (ch. 32) is the load-bearing occurrence. |
Gap: There is no Bengali idiom equivalent to the English theological shorthand “slave to sin” that would carry the metaphor automatically; দাসত্ব পাপের (“slavery of/to sin”) is a workable compound but is a paraphrase construction requiring explicit doctrinal framing every time, not a ready-made idiom.
1.7 Moses as Mediator and Deliverer
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| উদ্ধারকর্তা (deliverer) | Primary risk is conflation upward: a careless rendering could let Moses absorb language properly reserved for Christ. | Never use ত্রাণকর্তা (reserved Christ-title) for Moses; flag every occurrence for theologian review per registry. |
| মধ্যস্থ (mediator) | Built on the same root as baseline’s মধ্যস্থতা (intercession) — an asset, since it creates a visible lexical bridge to Christ’s mediatorial role (1 Timothy 2:5, Hebrews 8:6) without collapsing Moses into Christ. | Retain; teaching material should name the type/antitype relationship explicitly rather than leaving the bridge implicit. |
| আহুত/আহ্বান (called/calling) | Fixed baseline term; Moses’ call narrative (ch. 3) supplies a strong, concrete illustration of the term’s “sovereign, resistible-in-feeling-but-not-in-fact summons” sense that Romans’ more abstract usage lacks. | Use Exodus 3–4 as the anchor illustration when the term is introduced pedagogically across the curriculum. |
| বিচারক (judge) | Avoids কাজী (Islamic-associated judicial-religious title), a real and correctly-flagged collision. | Retain. |
Gap: No gap; this doctrine is well-served by existing/registered vocabulary, its risk being conflation (Moses/Christ) rather than lexical absence.
1.8 Holiness and the Fear of the Lord
| Available Bengali terms | Weakness | Recommended strategy |
|---|---|---|
| পবিত্র (holy) | Fixed baseline term; Exodus 3:5 (holy ground) is the first OT anchor for a term the New Testament will use pervasively — an asset for cross-curriculum continuity. | Retain without variation. |
| সদাপ্রভুর ভয় (fear of the LORD) | Two live collisions: (a) generic terror, indistinguishable in ordinary Bengali from the Israelites’ fear of Pharaoh in ch. 1; (b) folk-religious appeasement-dread of an unpredictable deity, a register strongly available in Bengali Shakta devotional culture (fear of Kali’s wrath, propitiated through ritual); a third, more distant risk (c) is confusion with the Islamic devotional category তাকওয়া (taqwa), which this curriculum does not import but which some Bengali Muslim-background readers may supply unprompted. | Mandatory translator note at every load-bearing occurrence (1:17, 21; 14:31; 19:16; 20:20) distinguishing reverent, relational, obedience-producing awe from all three neighboring senses. |
| ঈশ্বরের ক্রোধ (wrath of God) | Same Shakta-appeasement risk as fear-of-the-LORD: Bengali readers may default to a folk-religious frame in which an angry deity must be ritually placated to avert unpredictable harm. | Note required distinguishing righteous, judicial, covenant-triggered anger (Exodus 32, tied to a specific covenant violation) from capricious or appeasement-requiring divine temper. |
Gap: No lexical gap — সদাপ্রভুর ভয় and ঈশ্বরের ক্রোধ are adequate word choices — but both sit in an unusually crowded semantic neighborhood (see §2 below) that requires active fencing via translator notes rather than passive reliance on the words alone.
2. Missing Vocabulary vs. Crowded Semantic Neighborhoods
2.1 Genuinely Missing Vocabulary (requires coinage/compound, not selection among existing options)
| Concept | Why no existing Bengali word suffices | Resolution adopted |
|---|---|---|
| Kinsman-redeemer sense of גָּאַל (redeem, with price-paid nuance) | No Bengali word combines “rescue” + “one with a prior legal/family claim who pays a price to effect the rescue.” | Compound: মূল্য দিয়ে উদ্ধার করা, used alongside plain উদ্ধার করা where the price-paid dimension is not the specific point. |
| Tabernacle as mobile holy meeting-tent distinct from a fixed temple | Bengali religious vocabulary (মন্দির, ধর্মধাম, গির্জা) is built around fixed structures; no inherited word for a holy, portable, disassemble-and-carry sanctuary. | Constructed compound সমাগম তাম্বু, applied with strict internal consistency since it lacks pre-existing reader familiarity. |
| ”Pass over” (protective sparing, פָּסַח) vs. “pass through” (judgment, עָבַר) as a paired, contrastive theological hinge | Bengali has words for both underlying physical actions but no existing pair of words that carries the same built-in contrast the Hebrew text constructs on purpose within a single verse-unit (Exodus 12:12–13). | Two deliberately distinct renderings — রক্ষা করে চলে যাওয়া (protective passing-over) vs. ভেতর দিয়ে যাওয়া (judgment passing-through) — functioning as a matched pair, always introduced together with an explanatory note the first time they co-occur. |
| Compound self-revelatory names of YHWH beyond I AM and El Shaddai (Jehovah-Rapha, 15:26; Jehovah-Nissi, 17:15) | Not yet formally registered; Bengali has no pre-existing fixed rendering tradition for these particular compounds outside of transliteration-heavy older translations. | Forward gap — recommend registry extension: সদাপ্রভু, আমার আরোগ্যকারী (LORD my healer) and সদাপ্রভু, আমার বিজয়-পতাকা (LORD my banner), built on the fixed সদাপ্রভু base rather than a fresh transliteration, for cross-curriculum consistency with the I AM/YHWH cluster. |
| A single Bengali term for “consecrated for God’s exclusive service without becoming ritually inaccessible/withdrawn” (as applied to priests, the tabernacle, and Israel corporately) | Closest available terms (পবিত্র করা, উৎসর্গ করা) each carry partial overlap with either general holiness-language or with generic “dedicate/donate,” neither of which captures the specific “set apart yet still active in ordinary covenant life” sense the doctrine of sanctification requires (paralleling the baseline’s own note under Sanctification/separation_unto_gods_service). | Retain পবিত্র করা / পবিত্রীকরণ per baseline, but require the same “not sannyasa-style withdrawal” clarifying note the Romans baseline already applies to Romans 12:1–2, now extended explicitly to priestly/tabernacle consecration contexts (chs. 28–29, 40). |
2.2 Crowded Semantic Neighborhoods Requiring Active Fencing
These are cases where Bengali does supply an adequate word, but that word’s dominant everyday or religious usage sits close enough to a different tradition’s concept that fencing (via translator notes, paired-term discipline, or forbidden-substitution rules) is required to keep the intended sense from bleeding into a neighboring one.
| Bengali term | Neighboring concept(s) crowding it | Fencing mechanism |
|---|---|---|
| প্রায়শ্চিত্ত (atonement) | Hindu/general self-performed penance to remove sin or bad karma | Mandatory boundary note: God-provided substitutionary atonement, never self-achieved. |
| প্রতিমা (idol) | Everyday, neutral-to-positive word for a Hindu deity-image (Durga Puja) | Deliberately retained unsoftened; note explains the text’s own polemic rather than seeking a euphemism. |
| অভিষেক করা (anoint) | Hindu deity-consecration and royal coronation ritual | Retained per baseline precedent (cf. ঈশ্বর); positive typological link to খ্রীষ্ট noted rather than treated only as risk. |
| সর্বশক্তিমান ঈশ্বর (El Shaddai) | The baseline’s forbidden শক্তি (Shakta goddess-power) association | Flagged as an exception to the forbidden-substitution rule because it is a frozen liturgical compound, not an active soteriological collocation — must be explicitly whitelisted in Phase 2 validators to prevent false-flagging. |
| ঈর্ষাপরায়ণ (jealous, of God) | Ordinary human petty envy/marital jealousy | Mandatory note at every occurrence distinguishing covenant-exclusive zeal from insecure envy. |
| ঈশ্বরের ক্রোধ (wrath of God) / সদাপ্রভুর ভয় (fear of the LORD) | Shakta appeasement-dread of an unpredictable deity (Kali); secondarily, Islamic তাকওয়া | Mandatory note anchoring both in righteous, covenant-relational categories, not appeasement or generic terror. |
| ঈশ্বরের মুখ/উপস্থিতি (face/presence of God) | Hindu darshan (viewing a deity’s image, understood as safe/auspicious) | Mandatory note at 33:18–23 contrasting darshan-as-safe with the text’s own claim that unmediated divine presence is fatal apart from mediation. |
| যাজক (priest) | Hindu পুরোহিত (mantra-reciting temple officiant) / Islamic ইমাম (non-sacrificial prayer leader) | Functional definition supplied in teaching material; term itself retained as the least-bad option. |
| আমি যে আছি সেই আমি (I AM) | Vedantic sat (Being) and Islamic Al-Qayyūm (Self-Subsisting) as abstract ontological categories | Mandatory note anchoring the phrase in personal, historical, covenant-specific self-revelation rather than philosophical Being-talk. |
| সমাজ (congregation) | The baseline’s already-reserved মণ্ডলী (NT church) on one side, and জামাত (Bangladesh’s Jamaat-e-Islami political party) on the other | Both alternatives are forbidden; সমাজ is the fenced middle option, requiring no further note beyond registry documentation. |
| মুক্তি (avoided entirely for “redeem”) | Hindu moksha AND Bangladesh’s 1971 মুক্তিযুদ্ধ national-liberation narrative — a double, mutually reinforcing collision unique to Exodus among the two curricula built so far | Forbidden outright; উদ্ধার/উদ্ধার করা substituted throughout, consistent with baseline’s existing pattern of pre-emptive forbidden-term listing. |
3. Transliteration vs. Paraphrase Decisions
Every term below required an explicit choice between (a) transliterating the Hebrew/proper name, (b) using an established Bengali Bible descriptive term, or (c) constructing a new descriptive compound. The table records the decision and its rationale.
| Term | Decision | Rationale |
|---|---|---|
| Passover (פֶּסַח) | Established descriptive term — নিস্তারপর্ব | Long-standing Bengali Bible tradition already supplies a semantically transparent term (“the festival of deliverance/sparing”); transliterating (পেসাখ) would sacrifice transparency for no gain, since no existing Bengali reader-familiarity with a transliterated form exists. |
| Manna (מָן) | Transliteration — মান্না | No descriptive translation is possible or needed; “what is it?” is a wordplay specific to the Hebrew narrative moment (16:15) and is better explained via a footnote than forced into the noun itself. Established Bengali Bible practice already transliterates. |
| YHWH (יהוה) | Established substitute-term, not transliteration — সদাপ্রভু | Bengali Bible tradition follows the wider Christian convention (cf. English “LORD” in small caps) of using a reverential substitute rather than transliterating the Tetragrammaton; this is correct and must not be revisited toward a transliterated “যাহোয়া” or similar, which would break with the entire existing Bengali OT corpus. |
| I AM WHO I AM (אֶהְיֶה אֲשֶׁר אֶהְיֶה) | Paraphrase (full descriptive phrase) — আমি যে আছি সেই আমি | A transliteration would be meaningless to a Bengali reader with no independent access to Hebrew; the phrase must be paraphrased to preserve any content at all, at the cost of requiring a mandatory anchoring note (see §1.3, §2.1). |
| Abba-style preserved Aramaic terms (cf. baseline’s আব্বা) | N/A for Exodus — no Aramaic terms occur; noted here only to confirm the transliteration-preservation principle established in Romans does not apply to any Exodus vocabulary item. | Exodus is composed in Hebrew, and its key terms (YHWH, Passover, tabernacle) already have Bengali OT-tradition equivalents; there is no analog to Romans’ আব্বা that calls for preserved-transliteration treatment. |
| El Shaddai (אֵל שַׁדַּי) | Descriptive compound, not transliteration — সর্বশক্তিমান ঈশ্বর | A transliterated “এল শাদ্দাই” would communicate nothing to a Bengali reader on first encounter; the existing liturgical compound already carries recognized theological weight (creed usage) and is preferred. |
| Jehovah-Rapha / Jehovah-Nissi (compound divine names, 15:26; 17:15) | Recommend descriptive compound built on সদাপ্রভু, not transliteration (forward gap, see §2.1) | Consistency with the YHWH decision above; transliterating “যিহোবা-রাফা” would both violate the established সদাপ্রভু convention and introduce a second, competing substitute-name for YHWH within the same book. |
| Sabbath (שַׁבָּת) | Established descriptive/near-transliteration — বিশ্রামবার | বিশ্রামবার (“rest-day”) is a naturalized compound already standard in Bengali Bible usage; not a raw transliteration but not a free paraphrase either — it sits at the established midpoint and should not be altered. |
| Tabernacle (מִשְׁכָּן / אֹהֶל מוֹעֵד) | Constructed descriptive compound — সমাগম তাম্বু | No transliteration option carries any meaning; no single pre-existing Bengali word exists (see §2.1); a new but transparent compound is the only viable path. |
| Mercy seat (כַּפֹּרֶת) | Descriptive compound aligned to LXX theological weight — পাপাবরণ | Chosen specifically to preserve the lexical link to ἱλαστήριον (Romans 3:25) rather than to transliterate কাপ্পোরেৎ, which would sever that cross-curriculum typological connection entirely. |
| Amen, Hallelujah (general worship transliterations, established already in baseline) | Transliteration, per baseline precedent — আমেন, হাল্লিলূয়া | Confirmed as already settled by the Romans baseline; Exodus introduces no new instance requiring a fresh decision, but the principle (established liturgical transliterations are retained) applies unchanged. |
General decision rule derived from the above: transliteration is reserved for (a) proper names with no semantic content to translate (মান্না) and (b) established cross-tradition liturgical exclamations (আমেন); paraphrase/descriptive-compound is required wherever the term names a concept rather than a name, and especially wherever an existing Bengali word would import an unwanted competing theology (মন্দির, মুক্তি, প্রায়শ্চিত্ত-as-self-effort).
4. Ranked List of Highest-Risk Ambiguities
Ranked by combined severity (doctrinal centrality × likelihood of misreading given Bengali’s specific religious-vocabulary landscape).
-
“Redeem” (גָּאַל) vs. মুক্তি’s double collision (Hindu moksha AND 1971 Liberation War). Highest rank because the wrong word would not merely blur doctrine but actively re-narrate the entire book as national-political allegory for a Bangladeshi audience — a distinctly Exodus-curriculum risk with no Romans precedent. Chapters 6, 15.
-
প্রায়শ্চিত্ত (atonement) collapsing into self-performed penance. This is the term underlying the entire Passover/substitutionary-atonement doctrine cluster (the curriculum’s core passage); if the boundary note is dropped anywhere downstream, the whole doctrine of a God-provided substitute silently reverts to a merit-based framework the baseline’s own grace/পুণ্য distinction was built to prevent. Chapters 12, 29, 30, 32.
-
পাসাচ (“pass over,” protective) vs. আভার (“pass through,” judgment) — the twin-verb theological hinge of Exodus 12:12–13. If Bengali renders both with a single verb (a natural drift, since both describe physical movement through/past a location), the passage’s central logic — universal judgment on the same night as particular, blood-marked rescue — collapses into an undifferentiated single action. This is the sharpest lexical fork in the entire core passage. Chapter 12.
-
আমি যে আছি সেই আমি (I AM) read as Vedantic/Islamic abstract Being-language rather than personal covenant self-revelation. High doctrinal centrality (Character and Name of God doctrine) combined with genuinely strong neighboring frameworks (both Hindu and Muslim readers in the dual audience have ready-made categories for “self-existent Being” that are not the same claim). Chapter 3.
-
সদাপ্রভুর মুখ/উপস্থিতি (presence/face of God) read through the lens of darshan (safe deity-viewing). Directly inverts the text’s own claim (33:20) that unmediated divine presence kills; the collision is not merely stylistic but produces the opposite theological conclusion if unfenced. Chapter 33, with echoes in 19, 24, 40.
-
দাসত্ব (bondage) failing to make the literal-to-spiritual jump the “Deliverance from Bondage to Sin” doctrine requires. Because Bengali readers have strong, vivid contemporary associations with দাসত্ব as historical/social bonded labor, the text may be read exclusively as social-liberation narrative, with the spiritual-bondage application never registering without explicit teaching intervention. Chapters 1–6, 13, 20 (metaphorical extension throughout curriculum).
-
উদ্ধারকর্তা (Moses as deliverer) drifting into ত্রাণকর্তা’s reserved Christ-title territory. A Critical-risk conflation risk: Moses-as-type could be read, especially in summary/devotional material, as functionally equivalent to Christ-as-antitype, eroding the uniqueness of Christ’s mediatorial/saving work that the whole New Testament assumes as background. Chapters 3, 14, 32–34 (intercession scenes).
-
ঈর্ষাপরায়ণ (jealous God) read as petty human-style envy. Risk is amplified because Exodus 20 and 34 place this description directly alongside covenant law, where a misreading could suggest an insecure, temperamental deity rather than an exclusively covenant-loyal one — a serious characterological distortion, though narrower in doctrinal scope than items 1–5. Chapters 20, 34.
-
ঈশ্বরের ক্রোধ / সদাপ্রভুর ভয় (wrath / fear of the LORD) read through Shakta appeasement-dread. Real risk given Bengal’s specific religious geography (Kali/Durga worship centers), but somewhat mitigated by the fact that the baseline has already established related fencing language for শক্তি elsewhere, giving translators and reviewers a template to extend. Chapters 1, 14, 19, 20, 32.
-
প্রতিমা (idol/golden calf) — accepted collision, low residual risk after fencing. Ranked lowest among Critical/High items because, unlike the above, this collision is deliberately retained rather than fenced away — the text’s own polemic is well served by the word’s vividness, and the risk is not misreading but merely requiring the note to explain why the “obvious” softer alternative was not chosen. Chapter 32.
-
অভিষেক করা (anoint) — accepted collision, positive typological upside. Similarly low residual risk: the term’s Hindu consecration/coronation association is real but is outweighed by (a) precedent (ঈশ্বর itself is retained despite Vedantic roots) and (b) the productive link to খ্রীষ্ট (“Anointed One”) that actually aids catechesis rather than undermining it. Chapters 28–30, 40.
-
সমাজ (congregation) vs. জামাত/মণ্ডলী — well-fenced, low residual risk. Ranked lowest because both dangerous alternatives are already excluded by clear, non-overlapping reasons (political-party association; baseline’s NT reservation), leaving সমাজ as an essentially safe residual choice requiring no further translator intervention beyond registry documentation. Chapters 12, 16, 35.
5. Full-Book Chapter Coverage Confirmation
Every chapter of Exodus has been reviewed for load-bearing vocabulary relevant to the eight curriculum doctrines. Chapters not separately tabled above contributed no vocabulary beyond terms already covered in the sections cited.
| Chapters | Coverage status |
|---|---|
| 1 | Reviewed. Fear-of-Pharaoh vs. fear-of-God contrast (§1.8); bondage/দাসত্ব introduced (§1.6). No terms beyond those tables. |
| 2 | Reviewed. Narrative background to Moses; no new load-bearing doctrinal vocabulary beyond প্রথমজাত/covenant-continuity themes already covered under Redemption and Mediator clusters. |
| 3 | Reviewed. Core chapter for I AM, calling (আহুত), holy ground (পবিত্র) — fully covered §1.3, §1.7, §1.8, ranked item 4. |
| 4 | Reviewed. Sign (চিহ্ন), hardened heart introduced — covered in glossary Section B.2/B.3; no new gap items. |
| 5 | Reviewed. Intensified bondage (দাসত্ব); no new vocabulary beyond §1.6. |
| 6 | Reviewed. Redeem (גָּאַל), covenant name revelation — covered §1.1, ranked item 1. |
| 7–10 | Reviewed. Plague (আঘাত/মহামারী), sign, hardened heart — covered in glossary; no new gap items beyond those already registered. |
| 11 | Reviewed. Firstborn (প্রথমজাত) death announced — feeds directly into core passage, §1.2. |
| 12 | Core passage. Fully covered: Passover, lamb, blood, unleavened bread, pass-over/pass-through distinction — ranked items 2–3, §1.2. |
| 13 | Reviewed. Unleavened bread, firstborn consecration, pillar of cloud/fire — covered §1.2, §1.5. |
| 14 | Reviewed. Salvation/yeshuah (পরিত্রাণ), fear of the LORD (14:31) — covered §1.1, §1.8, ranked item 6. |
| 15 | Reviewed. Song of Moses; “the LORD shall reign” (kingdom echo); Jehovah-Rapha (15:26) — flagged as forward gap, §2.1. |
| 16 | Reviewed. Manna (transliteration decision, §3), Sabbath introduced, glory of the LORD appears (§1.5 cross-reference). |
| 17 | Reviewed. Testing (পরীক্ষা করা) at Massah/Meribah; Jehovah-Nissi (17:15) — flagged as forward gap, §2.1. |
| 18 | Reviewed. Judge/মধ্যস্থ cluster — covered §1.7. |
| 19 | Reviewed. Holy nation, kingdom of priests, fear/trembling theophany — covered §1.4, §1.8. |
| 20 | Reviewed. Ten Commandments, jealous God, steadfast love, altar — covered §1.3, §1.4, ranked item 8. |
| 21–23 | Reviewed. Ordinances/বিধি (case law), sojourner/প্রবাসী, Sabbath, feasts — covered §1.4, §2.2. |
| 24 | Reviewed. Blood of the covenant (নিয়মের রক্ত), Book of the Covenant — covered §1.4; direct link to NT Last Supper language flagged for cross-curriculum consistency. |
| 25–27 | Reviewed. Ark, mercy seat, altar, tabernacle construction instructions — covered §1.5, ranked item 3 (mercy seat), transliteration table (§3). |
| 28–29 | Reviewed. Priestly garments, priest/high priest, anointing, sin/burnt offerings — covered §1.5, §1.2, ranked item 11. |
| 30–31 | Reviewed. Incense altar, anointing oil, Sabbath as covenant sign, Spirit of God enabling Bezalel/Oholiab — covered §1.3 (Spirit of God note), §1.5. |
| 32 | Reviewed. Golden calf/idol, wrath of God, intercession, hardened heart — covered §1.2, §1.3, §1.6, ranked items 8, 10. |
| 33 | Reviewed. Presence/face of God — covered §1.5, ranked item 5 (highest-severity presence-doctrine risk). |
| 34 | Reviewed. Covenant renewal, compassionate and gracious, jealous, steadfast love, veil — covered §1.3, ranked item 8; veil flagged for future 2 Corinthians 3 cross-reference. |
| 35–39 | Reviewed. Execution of chs. 25–31 tabernacle instructions; vocabulary is a direct repetition of already-tabled terms (ark, mercy seat, priestly garments, anointing, Spirit of God/craftsmanship). No new load-bearing terms; confirmed reviewed with no additional gap items. |
| 40 | Reviewed. Tabernacle erected; glory of the LORD fills the tabernacle; pillar of cloud — covered §1.3 (glory), §1.5 (tabernacle/presence), closing the Tabernacle and Presence doctrine arc. |
This document must be loaded alongside 08_core_glossary.md, bible_term_registry.json, and doctrine_risk_registry.json before any further Phase 1 doctrine-risk registry extension or Phase 2 translation-memory build for the Exodus curriculum. Forward-gap items identified in §2.1 (Jehovah-Rapha, Jehovah-Nissi compound names) should be incorporated into the next version-incremented revision of the term registry before Phase 2 processing reaches Exodus 15 and 17.