Linguistic Gap Analysis
Linguistic Gap Analysis — Mark 1–16 (English → Assamese)
Purpose
This document performs the linguistic gap analysis mandated by PRD Phase 1 Step 8 for the Gospel of Mark, destination language Assamese. It (a) maps each of the curriculum’s eight core doctrines against the available Assamese lexical resources, naming specific weaknesses in each candidate term; (b) separates true vocabulary gaps (concepts with no ready Assamese word at all) from crowded semantic neighborhoods (concepts where multiple existing Assamese words compete and collide, requiring deliberate lexical fencing); (c) records transliteration-vs-paraphrase decisions with reasoning; and (d) ranks the curriculum’s highest-risk ambiguities for prioritized Phase 2 review routing.
This analysis does not overturn any baseline (translation_memory.json, Romans) or working (08_core_glossary.md, bible_term_registry.json) rendering. It explains why those renderings were necessary by exposing the underlying lexical terrain, and it surfaces a small number of open decisions that Phase 2 reviewers must watch closely, especially around the core passage, Mark 10:35–45.
1. Method
For each doctrine, three questions were asked of the Assamese lexical field:
- What words already exist in Assamese (general register, Vaishnavite/Ekasarana devotional register, and existing Assamese Bible Society usage) that a translator might reach for?
- What theological content does each candidate word carry into the text that the source term does not intend — or fail to carry that the source term requires?
- Given (1) and (2), should the working glossary reuse a baseline term, coin/extend a compound, or transliterate?
The eight curriculum doctrines are treated as the organizing frame; term-level detail defers to 08_core_glossary.md and bible_term_registry.json, which this document explains and stress-tests rather than duplicates.
2. Doctrine Vocabulary Matrix
| Doctrine | Available Assamese Term(s) | Weaknesses of Each | Recommended Strategy |
|---|---|---|---|
| Jesus as Suffering Servant and Son of God | দুখভোগী সেৱক (suffering servant, descriptive phrase); ঈশ্বৰৰ পুত্ৰ (baseline, reused) | No single Assamese devotional-register word combines suffering with chosen honor, because in both mainstream Puranic and Ekasarana devotional narrative, a deity’s descent (avatar) is glorious and triumphant, not degraded and executed; suffering-as-exaltation is a genuinely foreign narrative shape. ঈশ্বৰৰ পুত্ৰ alone (without Isaiah-53 framing) risks being heard as ordinary royal-messianic triumph language, flattening the Servant motif. | Reuse ঈশ্বৰৰ পুত্ৰ (Critical, baseline) for Sonship; render “Suffering Servant” descriptively as a phrase (দুখভোগী সেৱক) rather than seeking a single-word equivalent, and mandate a translator note at first occurrence (10:45; 15:39 context) tying suffering explicitly to Isaiah 53 and to voluntary self-giving, not defeat or karmic misfortune. |
| The Kingdom of God Breaking In | ঈশ্বৰৰ ৰাজ্য (baseline, reused); “ওচৰ চাপি আহিছে” (has drawn near, 1:15) | ৰাজ্য (kingdom) alone pulls toward territorial/political kingdom associations, sensitized by Assam’s Ahom-kingdom history (per baseline note); the inaugurated-yet-future tension of 1:15 (ἤγγικεν) has no ready single Assamese verb — most natural options (“আহিব,” “হ’ব”) default to purely future sense. | Reuse ঈশ্বৰৰ ৰাজ্য; render ἤγγικεν with ওচৰ চাপি আহিছে (“has drawn near/arrived close”), a present-perfect construction that permits an “already-but-not-yet” reading; mandatory translator note at 1:15 stating the reign is breaking in now, not merely predicted for later. |
| Faith and Discipleship amid Fear | বিশ্বাস (baseline, reused); ভয় নকৰিব (working term); অবিশ্বাস (working term) | বিশ্বাস risks being read through the Ekasarana bhakti lens (devotional surrender/trust in a chosen deity) rather than as reasoned trust anchored in a specific historical person’s demonstrated authority; “do not fear” risks being read as a call to suppress emotion (a stoic/yogic ideal) rather than displacement of fear by trust in Christ’s presence. | Reuse বিশ্বাस, always keeping Jesus explicit as its grammatical object in context (per baseline instruction); render “do not fear” as ভয় নকৰিব with translator notes at 4:40, 5:36, 6:50 clarifying that the command is answered by Christ’s presence/identity, not by internal calm-cultivation technique. |
| The Necessity of the Cross | নিশ্চয় হ’ব লাগিব (δεῖ, working term); ক্ৰুচ (transliteration, working term) | The most natural Assamese fatalism-vocabulary — ভাগ্যত থাকে (“it is in one’s fate”), or কৰ্মফল-adjacent phrasing — is doctrinally forbidden (collides with karma/fate); ক্ৰুচ (transliterated) carries no native semantic content at all in Assamese and must be entirely taught rather than translated. | Use নিশ্চয় হ’ব লাগিব for δεῖ (never fate-language); retain ক্ৰুচ as an established Assamese Christian loanword (already current from existing Assamese liturgical usage) rather than coining a native paraphrase, since a paraphrase (“দুখৰ কাঠ,” “wood of suffering”) would either sound invented or drift toward generic-suffering metaphor, losing the once-for-all historical execution-instrument sense. |
| Servanthood versus Worldly Greatness | সেৱক (diakonos, working term); দাস (doulos, working term); প্ৰভুত্ব দেখুৱাই অত্যাচাৰ কৰা (domineering rule, working term) | Assamese social memory (historical bonded labor, caste-linked service roles) attaches real stigma and status-loss to দাস; conversely সেৱক in modern Assamese civic usage has been flattened to mean any paid employee/functionary, losing the passage’s rhetorical escalation from “servant” to “slave of all.” | Keep সেৱক and দাস lexically distinct per the working registry, escalating from one to the other exactly where Mark 10:43–44 escalates; mandatory translator note at 10:44 (দাস) clarifying voluntary Christlike self-giving, never a caste- or labor-status assignment. |
| The Messianic Secret | ৰাজ্যৰ ৰহস্য (mystery of the kingdom, working term); silencing commands (1:34, 1:44, 8:30) rendered with standard verbs of command | Assamese “ৰহস্য” (rahasya) is the exact loanword used across Vedantic, tantric, and Ekasarana esoteric literature for a permanently guru-gated secret teaching reserved for initiates — the opposite theological shape from Mark’s temporary, progressively-disclosed, ultimately-proclaimed secret. | Retain ৰাজ্যৰ ৰহস্য (no better native alternative exists) but mandate a standing translator note, applied consistently at every anchor passage (4:11, 1:34/1:44, 8:29-30, 9:9), stating explicitly that this is time-bound concealment moving toward open proclamation (4:22), not a permanent closed gnosis. |
| Jesus’ Authority over Sickness, Sin, and Nature | অধিকাৰ (ἐξουσία, working term); ঈশ্বৰৰ শক্তি (δύναμις, baseline reused); অশুচি আত্মা/দুষ্ট আত্মা (unclean spirit, working term) | ক্ষমতা and অধিকাৰ are near-synonyms in everyday Assamese, risking their collapse into one word for “power” generally, erasing Mark’s careful distinction between delegated right (ἐξουσία) and raw power (δύναμις); ভূত (the natural word reached for by ordinary speakers for “unclean spirit”) is a folk-ghost term implying a wandering ancestral or nature-spirit, not an organized agent of a personal Satan. | Keep অধিকাৰ (right to act) and ঈশ্বৰৰ শক্তি (raw power) lexically distinct wherever both concepts appear near each other (e.g., 1:22 vs. 1:27); never render “unclean spirit/demon” as ভূত — use অশুচি আত্মা/দুষ্ট আত্মা consistently, per the working registry. |
| The Ransom for Many | উদ্ধাৰ-মূল্য (ransom, working term); …ৰ সলনি (ἀντὶ πολλῶν, working term); নিয়মৰ তেজ (blood of the covenant, working term) | This is the curriculum’s single greatest lexical danger: the most natural pre-existing Bengali-Assamese Christian coinage for “ransom” is built on the মুক্তি root (মুক্তিমূল্য), which is the exact forbidden root the baseline salvation entry rules out because মুক্তি names liberation from the সংসাৰ rebirth cycle; “in place of” can be flattened by a general-benefit preposition (…ৰ কাৰণে/বাবে) that loses the exchange-logic central to substitutionary atonement. | Mandatory: উদ্ধাৰ-মূল্য (never মুক্তিমূল্য) built on the baseline-sanctioned উদ্ধাৰ/নিস্তাৰ family; mandatory: …ৰ সলনি (never …ৰ কাৰণে/বাবে) for ἀντὶ πολλῶν. Every occurrence of Mark 10:45 and its immediate discourse context must route to human theologian review without exception. |
3. Missing Vocabulary vs. Crowded Semantic Neighborhoods
3.1 True Vocabulary Gaps (no ready Assamese word exists at all)
These are concepts for which Assamese, even drawing on Sanskrit tatsama resources, has no existing single word — the source concept is genuinely foreign to the receptor culture’s conceptual inventory, requiring either a compound coinage or a transliteration:
| Concept | Gap Description | Resolution |
|---|---|---|
| Ransom (λύτρον) as substitutionary release-price | Assamese has words for “price” (দাম, মূল্য) and for “rescue” (উদ্ধাৰ, নিস্তাৰ) but no single existing word fusing “price paid to free a captive/slave” with the theological weight of a voluntary, once-for-all substitutionary death. | Coined compound: উদ্ধাৰ-মূল্য (working term, Critical). |
| Crucifixion-as-execution-instrument (σταυρός) | No native Assamese word for a Roman execution stake exists, nor should one be invented from native metaphor (risk of drifting into generic “suffering” imagery). | Transliteration: ক্ৰুচ (established loanword). |
| Gehenna as a specific, final, non-cyclical judgment-place | Assamese inherits a rich vocabulary for hell(s) (নৰক, and its many named subdivisions in Puranic cosmology) but every existing option denotes a temporary, karma-proportioned punishment from which a soul is eventually reborn out. | Reuse নৰক (working term, High) with a mandatory distinguishing note at every occurrence rather than coin a new word, since a coinage would be unrecognizable to readers and νόμος-level precision is achievable through the note instead. |
| ”I am” (ἐγώ εἰμι) carrying Exodus 3:14 LXX self-naming resonance | Assamese মই আছোঁ is a perfectly ordinary self-identification formula with zero built-in echo of a divine self-naming tradition; there is no elevated register of Assamese self-reference that would carry this automatically. | Plain rendering মই আছোঁ plus mandatory translator note at every occurrence (6:50; 14:62), since no lexical solution alone can supply the missing resonance. |
| Corban (a Hebrew vow-formula with binding legal-religious force) | No Assamese equivalent vow-category exists with this specific structure (dedicating property to God in a way that legally overrides filial duty). | Transliteration কৰ্বান with explanatory gloss (Low-Medium risk; comprehension gap only, not doctrinal collision). |
3.2 Crowded Semantic Neighborhoods (too many competing words, requiring fencing)
These are concepts where the risk is the opposite of a gap: Assamese offers several plausible words clustered around one idea, each importing a different — and sometimes doctrinally incompatible — frame. Each neighborhood below requires an explicit fence: a rule stating which word is used for which sense, and which words are permanently excluded.
Neighborhood 1 — Liberation/Deliverance cluster: মুক্তি, মোক্ষ, নিস্তাৰ, উদ্ধাৰ, পৰিত্ৰাণ
- Fence: পৰিত্ৰাণ = eternal salvation only (Critical, baseline). নিস্তাৰ = historical/physical deliverance narrative only (e.g., Passover background, নিস্তাৰ-পৰ্ব). উদ্ধাৰ = rescue-action generally, and as the root of উদ্ধাৰ-মূল্য (ransom). মুক্তি and মোক্ষ = permanently forbidden across the entire curriculum, in every sense, because both denote escape from the সংসাৰ rebirth cycle.
Neighborhood 2 — Power/Authority cluster: শক্তি, অধিকাৰ, ক্ষমতা, বল
- Fence: ঈশ্বৰৰ শক্তি = δύναμις, God’s own raw power (baseline). অধিকাৰ = ἐξουσία, delegated right to act (working term). ক্ষমতা is permitted only as an occasional secondary gloss for অধিকাৰ, never as the primary glossary rendering, to avoid the two Greek concepts collapsing into one Assamese word when they appear in the same pericope (e.g., Mark 1:22 teaching-with-authority vs. 1:27 astonishment-at-power).
Neighborhood 3 — Lordship/domineering-rule cluster: প্ৰভু, প্ৰভুত্ব, কৰ্তৃত্ব
- Fence: প্ৰভু = exclusive title for Christ’s Lordship only (Critical, baseline) — never applied to any human ruler, Gentile or otherwise. The condemned domineering-rule verbs of 10:42 (κατακυριεύω/κατεξουσιάζω) must be rendered with a phrase (প্ৰভুত্ব দেখুৱাই অত্যাচাৰ কৰা) that keeps visible distance from প্ৰভু as a bare title, precisely because both share the κύριος root in Greek — a literal cognate-for-cognate rendering in Assamese would be the single easiest way to accidentally besmirch Christ’s own title in translation.
Neighborhood 4 — Servitude cluster: সেৱক, দাস, ভৃত্য, চাকৰ
- Fence: সেৱক = diakonos, dignified active service (working term). দাস = doulos, total self-giving bondservant status, reserved for the rhetorical climax of 10:44 (working term, Critical, with mandatory note against caste/bonded-labor misreading). ভৃত্য and চাকৰ (ordinary hired-servant/domestic-help words) are excluded from the glossary entirely — too weak for either Greek term’s force.
Neighborhood 5 — Spirit-being cluster: আত্মা, ভূত, দুষ্ট আত্মা, অশুচি আত্মা, প্ৰেত
- Fence: পবিত্ৰ আত্মা = the Holy Spirit only, never elsewhere (Critical, baseline). অশুচি আত্মা / দুষ্ট আত্মা = unclean spirits/demons (working term). ভূত and প্ৰেত (folk ghost/wandering-spirit vocabulary) are excluded entirely from theological use in this curriculum — they would trivialize organized, personal, Satan-directed hostile spiritual agency into culturally familiar ancestor-ghost or nature-spirit folklore.
Neighborhood 6 — Compassion/mercy cluster: দয়া, কৰুণা, কৃপা, অনুগ্ৰহ
- Fence: অনুগ্ৰহ remains reserved for grace (Critical, baseline — though “grace” is not a headline doctrine in Mark, it appears in background discipleship contexts and must not drift). কৃপা may be used only for ordinary non-salvific compassion per baseline note. Jesus’ σπλαγχνίζομαι moments (1:41; 6:34; 8:2) use কৰুণাবোধ কৰা (working term) rather than bare দয়া, to convey active visceral movement-to-act rather than passive pity.
Neighborhood 7 — Purity/cleansing cluster: পবিত্ৰ, শুচি, নিৰ্মল, অশুচি
- Fence: পবিত্ৰ = moral-relational holiness/set-apartness (baseline). শুচি (and শুচি কৰা) = ritual/physical cleanliness, reserved for the leprosy-cleansing narrative (1:40-45) and generally for ceremonial-purity contexts (7:1-23 hand-washing controversy) — this ritual-register word is deliberately retained here (unlike some baseline contexts) because Mark’s own narrative logic depends on the reader recognizing the ritual category before Jesus subverts it by relocating impurity to the heart (κοινόω, 7:15-23).
Neighborhood 8 — Fear/Trust cluster: ভয়, বিশ্বাস, শ্ৰদ্ধা, ভক্তি
- Fence: বিশ্বাস remains the sole glossary term for πίστις (baseline; never শ্ৰদ্ধা/ভক্তি). ভয় নকৰিব names the fear-displacement command; no lexical overlap is permitted between “fear” vocabulary and “faith” vocabulary even though the two are structurally paired throughout Mark’s healing/nature-miracle narratives (4:40; 5:36; 6:50).
4. Transliteration vs. Paraphrase Decisions
| Term | Decision | Reasoning |
|---|---|---|
| ক্ৰুচ (cross) | Transliterate | Established Assamese Christian loanword; no native metaphor exists that would not either sound invented or generalize the term into ordinary “suffering,” losing its historical specificity as a Roman execution-instrument. |
| আব্বা (Abba) | Transliterate (baseline) | Preserves Aramaic filial-intimacy register that formal পিতা alone cannot convey; retained per baseline instruction. |
| হোচেয়ন্না (Hosanna) | Transliterate + gloss | A liturgical acclamation with no native Assamese equivalent; a paraphrase (“ৰক্ষা কৰক,” “save now”) would lose the crowd’s scriptural (Psalm 118) intertextual echo, so transliteration with a short explanatory gloss at first occurrence (11:9) is preferred. |
| কৰ্বান (Corban) | Transliterate + gloss | A specific Hebrew legal-vow category with no Assamese analog; low doctrinal risk, purely a comprehension gap, resolved by explanatory gloss rather than paraphrase. |
| মচীহ / খ্ৰীষ্ট (Messiah/Christ) | Transliterate, dual-form (baseline) | মচীহ preserves OT-messianic-expectation discourse; খ্ৰীষ্ট functions as Jesus’ established proper name-title. Both are transliterations from source-language proper nouns/titles, not native paraphrase, by design — this is the safest strategy given the doctrine’s Critical syncretism risk with avatar theology. |
| উদ্ধাৰ-মূল্য (ransom) | Paraphrase/compound coinage, NOT transliteration | A transliterated Greek λύτρον would be theologically opaque to Assamese readers; a native compound built on baseline-sanctioned roots is required so the concept is intelligible while avoiding the forbidden মুক্তি root. |
| …ৰ সলনি (ἀντὶ πολλῶν) | Paraphrase | No transliteration is possible for a grammatical preposition-phrase; the paraphrase must be chosen specifically to preserve exchange-logic (see §2 and §3.1 above). |
| মানুহৰ পুত্ৰ (Son of Man) | Paraphrase (established Bible-tradition calque) | A literal calque of the Semitic idiom is already the Assamese Bible Society’s established rendering; no transliteration is applicable, and the calque is transparent enough with a translator note to carry the Daniel 7 background. |
| নিয়মৰ তেজ (blood of the covenant) | Paraphrase (compound of two reused/working terms) | Built from নিয়ম (baseline covenant) + তেজ (blood, plain vocabulary); no transliteration applies; the compound must be kept intact rather than split so both sacrificial and covenant-making senses travel together. |
General principle applied throughout: transliteration is reserved for (a) proper names/titles already established in Assamese Bible tradition, (b) liturgical exclamations with no native equivalent, and (c) culturally specific single-referent legal/ritual terms (Corban). Paraphrase/compound coinage is reserved for theological concepts that must remain intelligible to a first-time reader while resisting collapse into a doctrinally incompatible existing Assamese word (ransom, substitution, Son of Man).
5. Ranked List of Highest-Risk Ambiguities (Mark 1–16)
Ranked by combined severity of syncretism risk × frequency/centrality × difficulty of lexical fencing. Ranks 1–5 require mandatory human theologian review at every occurrence; ranks 6–10 require native speaker review with theologian escalation on ambiguity.
- Ransom (λύτρον) / উদ্ধাৰ-মূল্য — Mark 10:45. Highest risk in the entire book: the most natural pre-existing Bengali-Assamese Christian coinage collides with the forbidden মুক্তি (mokṣa) root. This is the core passage’s defining term; any lapse here corrupts the curriculum’s theological center.
- Incarnation-adjacent language around Jesus’ seed-of-David descent and bodily reality (দেহধাৰণ background, cf. baseline) vs. অৱতাৰ. Though Mark does not use a direct incarnation formula as explicitly as Romans/John, every passage asserting Jesus’ full humanity alongside divine authority (1:1, 9:7, 15:39) is vulnerable to being read through the Ekasarana avatar-descent frame unless notes are consistently applied.
- Son of Man vs. Son of God (মানুহৰ পুত্ৰ vs. ঈশ্বৰৰ পুত্ৰ). Structurally paired but theologically distinct titles occurring in close proximity (e.g., 14:61-62); any bleed-through between them collapses Mark’s Christological argument.
- The Messianic Secret / ৰাজ্যৰ ৰহস্য. Direct terminological overlap with tantric/Vedantic “rahasya” (permanently guru-gated secret knowledge) risks converting Mark’s temporary, unfolding concealment into a closed-gnosis system — a subtle error likely to go unnoticed without the standing note.
- Domineering rule (κατακυριεύω/κατεξουσιάζω) vs. Lord (κύριος) — Mark 10:42 immediately preceding 10:45. Shared Greek root creates a live risk that a translator, working quickly through the core passage, reaches for a cognate-sounding Assamese phrase built on প্ৰভু and inadvertently associates Christ’s own exclusive title with the condemned Gentile-ruler behavior it is meant to contrast.
- Slave/doulos (দাস) — Mark 10:44. Historical bonded-labor and caste-servitude associations in Assamese social memory could cause readers to hear this as a caste-status assignment rather than voluntary Christlike self-giving, especially without the mandatory note.
- “I am” (ἐγώ εἰμι) — Mark 6:50; 14:62. The Exodus 3:14 LXX resonance is entirely dependent on the translator note; the bare Assamese phrase মই আছোঁ carries none of it on its own and could be read as merely calming Jesus’ frightened disciples with ordinary self-identification.
- Unclean spirit/demon (অশুচি আত্মা/দুষ্ট আত্মা) vs. folk ভূত. High everyday-speech pressure toward ভূত in casual retelling or oral paraphrase of healing narratives; must be actively guarded against in Phase 2 quality review, not just glossary listing.
- Gehenna/Hell (নৰক) — Mark 9:43-48. The word is unavoidable (no better native option exists) but carries strong built-in cyclical/temporary-punishment assumptions from Hindu cosmology that directly contradict the passage’s once-for-all judgment claim.
- Blood of the covenant (নিয়মৰ তেজ) — Mark 14:24, read alongside 10:45’s ransom. Risk is not mistranslation of either component word but failure to preserve their combined sacrificial-and-covenantal force together, especially if a Phase 2 segment-by-segment workflow processes 10:45 and 14:24 in separate batches without cross-referencing.
6. Summary Recommendations for Phase 2
- Load this analysis alongside
08_core_glossary.mdandbible_term_registry.jsonbefore any Mark segment is translated; the fences in §3.2 are binding, not advisory. - Every occurrence of Mark 10:35-45 (the core passage) must be treated as a single reviewed unit rather than split across independent segment batches, given the interlocking risks ranked #1, #5, and #10 above all converge there.
- Automated validation (per
12_ai_translation_requirements.mdconventions) should include a forbidden-substitution check for মুক্তিমূল্য (ransom) and ভূত (unclean spirit/demon) in addition to the existing baseline forbidden-term list. - Any newly encountered term not covered by this analysis or the working glossary must be run through the same three-question method in §1 before being added to translation memory.
This document extends but does not modify translation_memory.json (Romans baseline, v1). It presupposes and cross-references the working Mark glossary (08_core_glossary.md) and working term registry (bible_term_registry.json); all renderings cited here are identical to those artifacts and must not be altered independently in Phase 2.