Linguistic Gap Analysis
Linguistic Gap Analysis — Colossians (English → Fulfulde)
0. Scope and Method
This analysis executes PRD Phase 1 Step 8 for the Colossians curriculum. It covers the entire letter, chapters 1–4, with load-bearing material concentrated in the core passage (1:15–20) and in chapter 2’s polemical material — but no chapter is skipped. Where a chapter contributes no new terminology beyond what is already resolved elsewhere, that is stated explicitly rather than silently omitted (see §5, chapter coverage notes).
This document draws on, and must not contradict:
- The baseline Romans
translation_memory.json(all REUSE terms below are locked exactly as recorded there). 08_core_glossary.md(Colossians-specific NEW term proposals).- The Colossians extension of
bible_term_registry.json.
The purpose of this Step is diagnostic: to name where Fulfulde vocabulary is missing entirely, where it is crowded by competing Islamic/Pulaaku/folk-religious semantic fields and needs explicit doctrinal “fencing,” and to decide, term by term, between transliteration and descriptive paraphrase — then to rank the letter’s highest-risk ambiguities so Phase 2 review effort is allocated correctly.
1. Doctrine Vocabulary Matrix
Organized by the eight doctrines named in the curriculum parameters. “Available terms” lists the current Fulfulde candidate(s) (REUSE = locked baseline term; NEW = provisional Colossians-specific term from 08_core_glossary.md).
1.1 The Supremacy and Sufficiency of Christ over Creation (1:15–20, 2:9–10)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| image of God | mbaadi (NEW, Critical) | Reads as visual/idol likeness to an aniconist Tawhid-formed ear | Paraphrase-with-gloss: pair mbaadi with an explicit teaching gloss (“the one who makes the unseen God fully known in a person,” not “a picture of God”) at every occurrence, not just first use |
| firstborn | raɓɓiɗɗo (NEW, Critical) | Reads chronologically (“first thing made”) | Never occur without “before all things” (1:17) in the same teaching unit; treat the two verses as one inseparable doctrinal statement |
| creation/creator distinction | tagu (NEW, High) | On its own, low risk; risk is placement, not the word | Syntax check at Phase 2: Christ must always be grammatical agent of tagu-verbs, never grammatical object/member of the created set |
| preeminence | adiiɗo e fof (NEW, High) | Idiomatic softening temptation (“very important”) | Enforce literal totality (“first/foremost in absolutely everything”); forbid comparative-only renderings |
| fullness/deity bodily | timmal + Alanaagal + e neɗɗaagal ɓalndu makko (NEW, all Critical) | No crystallized precedent anywhere in the thinner Fulfulde Bible-translation tradition; three separate compounds must cohere as one claim | Treat 1:19 and 2:9 as a single fixed unit requiring theologian-confirmed identical wording at both occurrences (see §4 below) |
| hold together / sustaining | dentinaa (NEW, High) | Risks an impersonal “laws of nature” reading | Explicit personal-agent framing required in surrounding teaching text |
1.2 Christ as Head of the Church (1:18, 2:19)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| head | hoore (NEW, High) | Collides with clan/political headship register (parallel to baseline’s Laamiiɗo caution) | Consistent modifying phrase distinguishing organic/life-giving headship from ruler-headship; never pair with laamu vocabulary in the same clause |
| church (as body’s referent) | kawtal (REUSE, Medium) | Risk of narrowing to a single local congregation | Explicit universal-body framing note required at 1:18, 1:24, 4:15-16 |
| body (organic) | ɓalndu (NEW, Medium) | Must not merge with kawtal | Maintain as two distinct, non-interchangeable lexical items throughout all four chapters |
1.3 Reconciliation through the Cross (1:20–22, 2:13–15)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| reconciliation | artirgol e Alla (NEW, Critical) | Zero existing precedent — Islamic soteriology has no mediated-atonement/restored-relationship category at all | Treat as a doctrine requiring full explanatory unpacking every time it occurs, not a word that carries its own meaning; theologian-authored explanatory footnote mandatory |
| blood of the cross | ɓiiɓi kuruusi (NEW, Critical) | Compounds crucifixion-denial (Quran 4:157) with polemical freight around “cross” vocabulary specifically | Dialect-confirm kuruusi before deployment; never omit the historical-event framing |
| record of debt | winndaandi ɓaggeeli (NEW, High) | No ready equivalent; legal-financial metaphor unfamiliar in this exact form | Full descriptive teaching required; do not abbreviate to a bare noun once introduced |
| making peace | waɗugol jam (NEW, Medium, built on REUSE jam) | jam’s greeting-register dominance can flatten the costly, cross-purchased sense | Explicit “not the everyday jam of greeting, but peace bought at the cost of blood” contrast note |
1.4 Warning against False Teaching and Syncretism (2:4, 2:8, 2:16–23)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| elemental spirits | korɗe aduna (NEW, Critical) | Directly parallels barke/gris-gris/amulet practice, the single most active syncretism risk in the region | Frame as Paul speaking directly and pastorally into a live contemporary situation, not an ancient curiosity |
| worship of angels | malaykaaji + diine (NEW, Critical) | malaykaaji is shared, prestige-bearing Islamic vocabulary (Jibril, angel of death) | Must warn against displacing Christ with ANY intermediary devotion (explicitly including, by application, marabout/saint veneration) without caricaturing ordinary Islamic angelology as such |
| philosophy / human tradition | filosofi / anndal aduna; aada yimɓe (NEW, Medium/High) | Risk of reading as a blanket attack on Islamic kalam scholarship or fiqh/Sharia tradition generally | Target language must repeatedly specify “this teaching,” “this human tradition,” never generalize to “philosophy” or “tradition” as categories |
| self-made religion / false humility | diine mo neɗɗo suɓii hoore-makko; surtaare/hoyeende (NEW, High) | surtaare/hoyeende used both pejoratively (2:18, 2:23) and approvingly (3:12) in the same letter | Mandatory context-flagging at both senses; never let Phase 2 default to one sense across the whole book |
| shadow vs. substance | sowdu (NEW, Medium, unconfirmed) | No confidently attested Fulfulde term identified at all | Hold as open item pending native-speaker confirmation; do not finalize before Phase 2 |
1.5 Union with Christ — Died and Raised With Him (2:12, 2:20, 3:1–4)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| raised with Christ | immoraa wondude e Kiristu (NEW, High, built on REUSE immorde e maayde) | Risk of collapsing into, or replacing, the historical bodily resurrection claim | Always co-occur with an explicit statement that Christ’s own historical resurrection remains a separate, prior, non-negotiable fact |
| baptism | batisim (NEW, High) | Surface-formal resemblance to wuduu/ghusl (ritual ablution) | Explicit once-for-all/identification-with-Christ’s-death framing at every occurrence, contrasted with repeated ritual washing |
| circumcision (figurative) | kuɓɓal (NEW, High) | Literal rite is routine and uncontroversial; figurative sense risks collapsing back into the literal | Explicit “not the physical rite but…” framing at 2:11 |
1.6 Putting Off the Old Self and Putting On the New (3:5–17)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| put off old self / put on new self | yaltinde neɗɗo on ɓooyɗo; naatnude neɗɗo kesun (NEW, High) | Risk of assimilation to Islamic tawba (moral reform within a continuous self) rather than identity replacement | Explicit teaching contrast between “reforming the same self” and “a decisively new identity” required at 3:9-10 |
| idolatry (covetousness as) | sirku (NEW, Critical) | Double-edged: strengthens the warning but invites the reader to redirect the same accusation at Trinitarian/deity-of-Christ doctrine | Mandatory two-directional framing note wherever sirku is used — see §4, ranked risk #2 |
| renewed in knowledge, image of Creator | anndal timmuɗo + mbaadi (NEW, High) | Deliberately reconnects to 1:15’s Christology; risk is losing that structural link in translation | Phase 2 style guide should cross-reference 1:15 explicitly whenever 3:10 is rendered |
| unity list (no distinction) | leƴƴi goo + list, incl. maccuɓe/rimɓe axis (Critical) | Engages both Sahelian inter-ethnic tension and the live Fulani rimɓe/maccuɓe caste axis simultaneously | Full pastoral seriousness required; must not be treated as ancient-history filler |
1.7 Household Codes (3:18–4:1)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| wives submit / husbands love | jebbilaade; giggol (NEW, High/Medium) | Intersects both Islamic family-law norms and Pulaaku patriarchal structure | Never render submission instruction without the co-located sacrificial-love instruction in the same teaching unit |
| slaves / masters | maccuɓe; jom-galle/giimaajo (NEW, both Critical) | maccuɓe names a real, in places still socially live Fulani caste category — a risk axis absent from the Romans baseline entirely | Mandatory theologian AND native-speaker review with pastoral framing notes; must always cross-reference 3:11 and 4:9 (Onesimus) |
| Lord vs. master disambiguation | Joom (divine only) vs. jom-galle/giimaajo (human) | Same Greek κύριος word for both referents across 3:22–4:1 — a disambiguation hazard with no Romans precedent | Segment-by-segment manual check of every κύριος occurrence in 3:18–4:1; Joom is NEVER permitted for the human referent |
1.8 Fullness of Deity in Christ Bodily (1:19, 2:9)
| Element | Available Fulfulde Term(s) | Weakness | Recommended Strategy |
|---|---|---|---|
| fullness | timmal (NEW, Critical) | Must not suggest partial/shared divine fullness | Identical rendering required at both 1:19 and 2:9 |
| deity/Godhood | Alanaagal (NEW, Critical) | Builds on baseline’s Alanaagal Almasiihu (deity_of_christ doctrine) but names the abstract essence for the first time as a standalone noun | Confirm this noun functions distinctly from Alla and Ɓiɗɗo Alla in the target dialect before deployment |
| dwell bodily | ñiiɓde + e neɗɗaagal ɓalndu makko (NEW, Critical) | Single most direct shirk-flashpoint phrase in the letter | Requires the most extensive theologian-authored explanatory apparatus of any phrase in this curriculum |
2. Missing Vocabulary vs. Crowded Semantic Neighborhoods
2.1 Missing Vocabulary (no crystallized Fulfulde term exists at all; requires new descriptive coinage)
These are genuine gaps — the underlying concept has no ready-made word in any dialect’s existing Christian or Islamic religious vocabulary, requiring the pipeline to construct new descriptive compounds from scratch (following the baseline’s established pattern for justification, incarnation, imputed_righteousness, etc.):
| Concept | Status | Note |
|---|---|---|
| reconciliation (artirgol e Alla) | Missing entirely | No Fulfulde precedent anywhere; Islamic soteriology has no equivalent doctrinal slot. Highest-priority new coinage in the whole curriculum. |
| record of debt (winndaandi ɓaggeeli) | Missing entirely | Legal-financial metaphor with no cultural analogue offered ready-made. |
| filling up what is lacking in Christ’s afflictions (1:24) | Missing; no fixed term proposed | Descriptive-only; must be handled per-instance with theologian guidance to avoid contradicting Sufficiency-of-Christ doctrine. |
| shadow vs. substance (sowdu) | Missing/unconfirmed | No confidently attested term identified by any consulted source; open item. |
| self-made religion (diine mo neɗɗo suɓii hoore-makko) | Missing; long descriptive compound only | No single-word candidate exists or is safe to invent. |
| fullness / deity as abstract nouns (timmal, Alanaagal) | Missing as fixed theological technical terms | Built from existing roots but not previously crystallized in this specific Christological sense. |
| disarmed / triumph over powers (2:15) | Missing; compound only | No existing Fulfulde victory-over-spirits idiom identified that carries the right theological force without also implying folk-religious power contests. |
Implication: every item in this table must be flagged, per the baseline’s own rule for provisional compounds, for MANDATORY dual theologian + native-speaker review before Phase 2 treats it as final. This is a larger missing-vocabulary set than Romans had proportionally, because Colossians’ polemical and cosmic-victory material (ch. 1–2) pushes further into technical territory than most of Romans does.
2.2 Crowded Semantic Neighborhoods (vocabulary exists, but is already occupied by a competing Islamic, Sufi, or Pulaaku concept and requires explicit doctrinal “fencing”)
Unlike the missing-vocabulary set, these terms have a ready Fulfulde word — the danger is the word’s existing meaning smuggling in unwanted content. Each requires a specific fencing strategy (a teaching-note convention that must travel with the term every time it is used):
| Term | Crowding Neighbor | Fencing Strategy |
|---|---|---|
| mbaadi (image) | Aniconist Tawhid (“nothing is like unto Him”) | Fence with: “not a picture/idol, but God’s own nature made knowable in a person” |
| raɓɓiɗɗo (firstborn) | Islamic Christology’s Jesus-as-created-prophet | Fence by mandatory co-occurrence with ado kala huunde (“before all things”) |
| annoora (light) | Sufi Nūr Muhammad (pre-existent “Light of Muhammad”) devotional theology | Fence with explicit note: shared word, unrelated doctrine — Christ’s light is not analogous to Nūr Muhammad |
| jipporɗe laamu / ardiiɓe / kattanɗe (cosmic powers) | jinn, ancestral/territorial spirits, barke-mediating powers | Fence as really-defeated-by-Christ, never as confirmation of independently potent rival forces requiring separate management |
| malaykaaji (angels, in 2:18) | Islamic angelology (Jibril, recording angels) | Fence as a warning against displacing Christ with intermediaries, not a critique of angels’ existence or ordinary Islamic angelology as such |
| korɗe aduna (elemental spirits) | barke / gris-gris / amulet practice | Fence by naming the real contemporary parallel directly rather than treating this as ancient-only |
| sirku (idolatry, 3:5) | shirk — the gravest sin in Islamic theology, routinely used polemically against Trinitarian doctrine | Fence bidirectionally: strengthens Paul’s point AND must proactively address “then is the Trinity/deity of Christ also shirk?” — see ranked risk #2 below |
| Arjunde (grace) | barke (marabout-mediated reciprocal blessing) | Fence per baseline: always pair with explicit “apart from works” statement |
| hakkille (wisdom) | Pulaaku’s own innate-shrewdness value | Fence as Spirit-given revealed wisdom in Christ, not native cultural cleverness |
| anndal timmuɗo (full knowledge) | Sufi ma’rifa (esoteric, initiatory mystical knowledge via a spiritual order) | Fence as open, revealed truth for every believer, never elite/initiatory |
| gundo Alla feeñi (mystery) | sirru (Sufi esoteric secret knowledge transmitted progressively by a marabout) | Fence by NEVER using sirru at all; mystērion is a secret NOW REVEALED to all, the structural opposite of sirru |
| jom-galle/giimaajo vs. Joom | The very real risk of scribal/oral drift back toward Joom for the human “master” referent given their semantic closeness in English “lord/master” | Fence with mandatory segment-level manual check across 3:18-4:1 (no exceptions) |
| maccuɓe (slaves) | The live, in-places-still-operative rimɓe/maccuɓe Fulani caste structure | Fence by mandatory cross-reference to 3:11 and 4:9 at every occurrence; never let the term stand without its reversal |
| surtaare/hoyeende (humility) | Same word field used pejoratively (2:18/2:23, ascetic performance) and approvingly (3:12, genuine virtue) resonant with Pulaaku semteende | Fence by explicit sense-marking at each occurrence, following the baseline’s precedent for “called” |
| filosofi / aada yimɓe (philosophy/tradition) | Prestige of Islamic kalam and fiqh scholarly tradition | Fence by specifying “this teaching” rather than the category itself |
3. Transliteration vs. Paraphrase Decisions
Following the baseline’s stated preference (“plain descriptive compound over inventing an opaque neologism where no crystallized term exists”), each load-bearing Colossians term is assigned one of three strategies: Transliterate (adopt an existing loanword, typically Arabic-derived and already current), Paraphrase (native-rooted descriptive compound), or Hybrid (transliterated core with an obligatory native gloss).
| Term | Strategy | Form | Rationale |
|---|---|---|---|
| Christ (title) | Transliterate | Kiristu | Established proper-name convention; must stay visually/aurally distinct from Almasiihu and Yeesu (per registry note) |
| Messiah | Transliterate | Almasiihu | Shared Arabic root with Quranic al-Masih is unavoidable and, per baseline, usable as a point of contact provided content is taught explicitly |
| baptism | Transliterate | batisim | Existing loanword; paraphrase would risk inventing a term that reads as a generic washing rite, losing the once-for-all identification-with-Christ sense entirely |
| philosophy | Hybrid | filosofi / anndal aduna | Either a loanword (filosofi) or a native compound depending on dialect register; both must carry an explicit “this specific false teaching” gloss |
| cross | Transliterate (pending confirmation) | kuruusi | No native Fulfulde term for the Roman execution device exists; transliteration is unavoidable, but the exact form must be theologian-confirmed per target dialect before use |
| circumcision (literal) | Paraphrase (native) | kuɓɓal | Fully native, uncontroversial, established rite-term; no transliteration needed |
| deity/Godhood | Paraphrase | Alanaagal | Native abstract-noun derivation from the same root as the baseline’s Alanaagal Almasiihu; avoids importing an unfamiliar Arabic theological-technical term with its own (denial-laden) Islamic content |
| fullness | Paraphrase | timmal | Native root (“completeness”); no suitable loanword exists and none is needed |
| reconciliation | Paraphrase | artirgol e Alla | Must be paraphrase, not transliteration — there is no Arabic-Islamic loanword to transliterate, since the underlying doctrine has no Islamic theological analogue at all |
| mystery | Paraphrase (explicit anti-transliteration decision) | gundo Alla feeñi | The obvious transliteration candidate (sirru) is explicitly REJECTED because it is already occupied by Sufi esoteric-secret theology; this is a case where transliteration is available but doctrinally forbidden |
| elemental spirits | Paraphrase | korɗe aduna | Avoids importing any Arabic-Islamic spirit-power vocabulary (e.g., jinn-adjacent terms) that would confirm rather than subordinate these powers |
| record of debt | Paraphrase | winndaandi ɓaggeeli | No loanword tradition for this legal-financial metaphor exists in any consulted Fulfulde source |
| worship of angels | Hybrid | malaykaaji + diine | malaykaaji is an unavoidable, already-current Arabic loanword (angels); diine (“religion/worship-practice”) supplies the native paraphrase component that frames the warning correctly |
| slaves/masters (household code) | Paraphrase (native) | maccuɓe / jom-galle / giimaajo | Fully native social-structure vocabulary; no transliteration is relevant or available |
| self-made religion | Paraphrase (long compound, no shortening) | diine mo neɗɗo suɓii hoore-makko | Deliberately verbose to avoid inventing a misleadingly compact neologism that could be mistaken for an established technical term |
General rule derived from this analysis: Colossians’ new vocabulary skews more heavily toward paraphrase than Romans’ did, because chapters 1–2 introduce concentrated, technical Christological and cosmic-victory content with essentially no existing Fulfulde precedent (contrast with Romans, where several key terms — arjunde, bakkaatu, sariya — already had Arabic-Islamic-loanword currency to build on, even if imperfectly). Where a loanword IS available (mystery→sirru; philosophy→a kalam-adjacent term), the analysis frequently recommends AGAINST using it, because the loanword’s existing Islamic theological content is a worse risk than an unfamiliar-but-clean native compound.
4. Ranked List of Highest-Risk Ambiguities in Colossians
Ranked by combined severity (doctrinal centrality × likelihood of default misreading × absence of existing safeguard), specific to this letter (not a repeat of the Romans ranking):
-
“Fullness of Deity dwelling bodily” (1:19, 2:9) — timmal / Alanaagal / e neɗɗaagal ɓalndu makko. The single most concentrated Incarnation-and-full-deity statement in the New Testament, landing directly on the sharpest edge of Tawhid/shirk objection with a triple-compound Fulfulde rendering that has no historical precedent at all. If any one of the three compound elements is softened, the entire curriculum’s Christology is compromised. Mandatory theologian pre-confirmation before Phase 2.
-
“Idolatry” as sirku applied to covetousness (3:5) — the double-edged shirk-vocabulary problem. This is Colossians’ unique highest-risk ambiguity not present anywhere in Romans: the very term that gives Paul’s warning force among Islamically-formed readers is also the term routinely used in regional Islamic polemic against Trinitarian/deity-of-Christ doctrine. Teaching materials that use sirku here without proactively addressing “then isn’t calling Jesus God itself shirk?” will hand readers exactly this counter-argument unaddressed.
-
κύριος disambiguation across 3:18–4:1 — Joom (divine) vs. jom-galle/giimaajo (human master), same Greek word, same short passage. No comparable hazard exists in Romans. A single missed occurrence collapses Christ’s exclusive Lordship into ordinary household authority, or vice versa elevates a human slave-owner to divine status.
-
“Firstborn” (1:15) without “before all things” (1:17) co-occurrence — the single highest-probability default misreading in the whole letter, because it converges exactly with mainstream mainstream Sunni Christology (Jesus as an exalted created being, not eternal God) and requires the two verses to function as one inseparable unit in translation and teaching, not two independent statements.
-
“Reconciliation” (1:20-22) taught as if it were merely intensified “forgiveness” — because no Fulfulde precedent exists for this doctrine at all, there is a real risk that translators/reviewers unconsciously flatten artirgol e Alla into a synonym for the already-established yaafuyee bakkaatuuji (forgiveness of sins), losing the restored-relationship/removed-hostility dimension entirely.
-
“Elemental spirits” / “disarmed the rulers and authorities” (2:8, 2:15) read as confirming rather than defeating rival spiritual powers — pastorally the most consequential ambiguity in the letter, because a mistranslation here does not just create doctrinal confusion but could actively reinforce the very barke/amulet-seeking behavior the passage is written to dismantle.
-
“Circumcision of Christ” (2:11) collapsing back into the literal rite — because the literal practice is routine and uncontroversial in Fulani communities (unlike first-century Judaism’s controversy), readers may simply not register that a figurative, spiritual sense is intended at all, rather than actively misreading a contested term.
-
maccuɓe (slaves) in the household code (3:22-25, 4:1) landing on a live, present-tense social caste category — a risk axis with no Romans precedent whatsoever; mistranslation or careless framing here risks real present-day social harm, not merely doctrinal confusion.
-
“Image of God” (1:15) read as a claim about visual representation — high risk in isolation, ranked below the fullness/deity items above because 1:15’s aniconism collision, while severe, is more readily defused by a single well-placed teaching gloss than the compounding triple-risk of item #1.
-
“Light” (1:12-13) resonating with Sufi Nūr Muhammad devotional theology — a subtler collision than the above (most readers will not consciously invoke Sufi Nūr theology on first hearing), but likely to surface in later devotional/hymnic use of this material where the connection becomes explicit and unaddressed.
-
“Mystery” (1:26-27, 2:2, 4:3) rendered with, or informally drifting toward, sirru — ranked lower than items above because the correct alternative (gundo Alla feeñi) is already specified and relatively easy to enforce mechanically, but the risk of oral/informal drift back toward the far more familiar sirru remains real given the thinner base of established Christian devotional vocabulary.
-
“Wisdom” and “full knowledge” (1:9, 2:2-3, 3:10) read through a Sufi ma’rifa initiatory-knowledge lens — moderate risk; the corrective framing (open, revealed, for every believer) is straightforward to state but easy to omit under space pressure in shorter teaching materials.
5. Chapter-by-Chapter Coverage Confirmation
- Chapter 1: Primary source of the letter’s highest-risk items (§4 ranks 1, 4, 5, 9, 10, 11, 12 all originate here). Fully analyzed above.
- Chapter 2: Primary source of the syncretism/false-teaching crowded-neighborhood items and ranks 2 (partially, via 2:9), 3, 6, 7, 11 (partially). Fully analyzed above.
- Chapter 3: Source of ranks 2, 8 (partially, via 3:11), and the union-with-Christ/old-new-self vocabulary gaps. Fully analyzed above.
- Chapter 4: Contributes rank 3 (4:1) and rank 8 (4:1, 4:9 cross-reference); also contributes the fellow-servant/maccuɗo dual-usage tension and open-door/mission vocabulary, both addressed in §1.7 and the missing-vocabulary table. Reviewed; no additional new load-bearing terminology beyond what is captured above — chapter 4’s remaining content (personal greetings, closing instructions) is Low risk and requires no further gap analysis.
This analysis must be read together with 08_core_glossary.md and the Colossians extension of bible_term_registry.json. All items marked NEW/provisional in both documents are formally unresolved until the theologian and native-speaker confirmation process specified in 12_ai_translation_requirements.md is completed; this document does not itself finalize any rendering.