Linguistic Gap Analysis
Linguistic Gap Analysis: Ezekiel (English → Arabic)
Purpose and Method
This document performs the PRD Phase 1 Step 8 linguistic gap analysis for the Ezekiel curriculum. It extends — and does not contradict — the Romans baseline translation_memory.json and bible_term_registry.json, and builds directly on the Ezekiel-specific extensions already logged in assets/bible_term_registry.json and analysis/08_core_glossary.md. Its purpose is narrower and more diagnostic than the glossary: for each of the seven curriculum doctrines it asks (1) what Arabic vocabulary is available, (2) where that vocabulary is structurally weak for the doctrine it must carry, (3) where vocabulary is simply missing and must be built by compounding or paraphrase, (4) where vocabulary exists but sits in a crowded semantic neighborhood shared with Islamic theology and must be fenced off, and (5) where a transliteration vs. paraphrase decision must be made explicitly rather than left to translator instinct.
The core passage, Ezekiel 37:1-14 (the Valley of Dry Bones), is the theological anchor of this analysis, but — per the full-book-coverage mandate — every chapter of Ezekiel has been reviewed for load-bearing vocabulary gaps; chapters contributing no new gap beyond what is already logged are noted as reviewed in the Full-Book Coverage Log (§6).
1. Doctrine Vocabulary Matrix
| Doctrine | Available Arabic Terms | Structural Weaknesses | Recommended Strategy |
|---|---|---|---|
| The Glory and Sovereignty of God | المجد (kavod), الرب / السيد الرب (YHWH / Adonai YHWH), يد الرب (hand of the LORD) | المجد is a single static noun; Arabic has no morphological way to mark that this occurrence of “glory” is part of a single continuous narrative arc (departure ch.8–11 → return ch.43) rather than an isolated attribute-mention each time. السيد الرب, while an established Van Dyck compound, can visually blur with generic “master/lord” honorifics used of human rulers in narrative Arabic if not typeset/glossed consistently. | Reuse المجد and السيد الرب exactly per registry; supplement with a running cross-reference apparatus (marginal note keyed to ch.1/8-11/43) rather than inventing new vocabulary — the gap is structural/discourse-level, not lexical, and is best solved editorially, not terminologically. |
| Individual Responsibility for Sin (Ezekiel 18) | الخطية (sin), بار/شرير (righteous/wicked person), يتوب/يرجع (turn/repent), إثم/معصية (iniquity/transgression) | Arabic بار and شرير are common everyday adjectives-as-nouns and carry no built-in signal that Ezekiel 18 is running a formal forensic case-law argument (apodictic “if…then” legal style); a reader may flatten the chapter into generic moral exhortation rather than recognizing its courtroom register. | Reuse existing terms exactly; rely on structural/register cues in the surrounding study material (headers marking “legal case,” repeated formulaic verses) rather than seeking specialized legal-Arabic vocabulary, which would sound archaic/juridical in a devotional register. |
| The New Heart and New Spirit | قلب جديد / روح جديدة (new heart/spirit), قلب من حجر / قلب من لحم (heart of stone/flesh), روح (ruach — polysemous) | No single Arabic root cleanly spans “wind / breath / Spirit” the way Hebrew ruach does; روح alone, isolated from context, defaults toward the specifically religious-spiritual sense in contemporary Arabic (colored by الروح القدس), understating the physical breath/wind sense active in 37:9-10. Conversely, ريح (wind) alone loses the personal, life-giving divine agency. | Context-dependent triad already logged in the registry (روح / ريح / نَفَس) must be applied verse-by-verse within 37:1-14 itself (see §4 below) rather than harmonized to one term; this is a designed feature of the passage, not an error to be smoothed over. |
| Judgment on Jerusalem and the Nations | دينونة (judgment), غضب الرب (wrath of the LORD), أصنام (idols), رجس/رجاسة (abomination), جوج وماجوج (Gog and Magog) | غضب الرب risks being read through a generically retributive, non-covenantal lens if not anchored to the specific covenant-violation context each time. جوج وماجوج is the single highest-collision item in the whole book (see §3, §5). | Reuse existing terms; mandatory theologian-reviewed disambiguation note at first occurrence of جوج وماجوج (already flagged Critical in the registry); reinforce covenant-lawsuit framing for غضب الرب via surrounding study-note apparatus rather than new vocabulary. |
| Restoration and the Valley of Dry Bones | سأُصعدكم من قبوركم / أُقيمكم (I will raise you from your graves), عظام (bones), جيش عظيم جدًا جدًا (exceedingly great army), عهد سلام / عهد أبدي (covenant of peace / everlasting covenant) | Arabic has one obvious noun for “resurrection” (القيامة) already fixed by the Romans baseline for Christ’s unique historical resurrection; Ezekiel 37 requires the same underlying hope-shape without borrowing that exact noun, forcing a verbal paraphrase strategy that must be handled with discipline across every occurrence in the chapter. | Verb-only strategy (أُقيمكم / سأُصعدكم), never القيامة (noun), per registry; supply an explicit typological note connecting national restoration to bodily resurrection hope without collapsing the two referents. |
| God’s Presence Departing and Returning | المجد (glory, shared with sovereignty doctrine), باب مغلق (shut gate), موضع أخمصَي قدميّ (place of the soles of my feet), الرب هناك (YHWH shammah) | The strongly embodied idiom “the place of the soles of my feet” has no natural Arabic devotional equivalent that isn’t either (a) overly literal/crude in register or (b) so abstracted it loses the deliberate intimacy of the image. | Render literally as logged (موضع أخمصَي قدميّ) rather than softening to a generic “dwelling place” phrase; pair with an explicit interpretive note (analogical, not literal-embodiment, language) modeled on Ezekiel’s own visionary caution in ch.1 (“likeness of,” “appearance of”). |
| The Vision of the New Temple | المقدس / الهيكل (sanctuary/temple), قدس الأقداس (holy of holies), رئيس (prince), نهر الحياة (river of life) | رئيس alone is a generic modern Arabic word for “president/head/chief” (e.g., رئيس الجمهورية) and carries no built-in signal of the deliberate theological demotion from מֶלֶךְ (king) that Ezekiel intends; a reader could miss that “prince” is a pointed choice rather than a synonym for “king.” | Retain رئيس/أمير per registry (already avoiding خليفة-adjacent vocabulary); supplement every occurrence with an explicit study note contrasting “prince” with “king” (ملك) to surface the deliberate subordination the Hebrew nasi encodes. |
2. Missing Vocabulary (True Lexical Gaps)
These are cases where Arabic simply has no existing word or fixed compound carrying the required sense, and one must be built (as already done provisionally in the Step-7 glossary) rather than merely selected from competing options.
| Concept | Gap Description | Constructed Arabic Solution | Notes |
|---|---|---|---|
| God’s self-vindicating holiness before onlooking nations (niqdashti) | Arabic تقديس/مقدس is oriented toward objects/people being sanctified by God, not God publicly vindicating his own reputation as the grammatical subject-and-object of the same verb. | أتقدّس / أُظهر قداستي (constructed reflexive paraphrase) | No pre-existing fixed compound in Arabic Christian tradition; this is a genuine coinage-by-paraphrase, flagged High risk, requiring theologian sign-off before entering standard use. |
| ”Dung-pellet” contemptuous nuance of gillulim | Standard أصنام is doctrinally accurate but semantically neutral; Hebrew’s insult-nuance has no Arabic lexical equivalent, since أصنام itself is already the dignified/standard term. | No lexical solution; explanatory footnote strategy only | Accept the gap; do not attempt an insulting neologism, which would read as crude rather than theologically pointed in formal register. |
| The Hebrew ruach’s simultaneous wind/breath/Spirit polysemy inside a single 14-verse unit (ch.37) | No single Arabic root spans all three senses without translator selection collapsing the wordplay. | Context-governed triad: روح / ريح / نَفَس, selected verse-by-verse (see §4) | This is the single most consequential missing-vocabulary case in the entire curriculum because it sits inside the core passage itself. |
| ”Covenant of peace” as comprehensive restoration (not a cessation-of-hostilities treaty) | العهد + السلام compound (عهد سلام) exists as a literal compound but Modern Standard Arabic readers default to the modern political sense of “peace treaty” (e.g., اتفاقية سلام) given contemporary regional usage. | عهد سلام retained, but requires an explicit disambiguating gloss distinguishing it from a modern political peace accord | Political-register collision risk is contextual/contemporary rather than inherent to the words themselves; flag for native speaker review given regional sensitivity. |
| ”Prince” (nasi) as deliberately non-royal subordinate ruler | Arabic أمير historically denotes a regional sovereign ruler (e.g., Umayyad/Abbasid amirs) and in some modern contexts a monarch-adjacent title (e.g., Gulf amirates), which can accidentally re-import the very royal connotation the Hebrew is avoiding. | رئيس preferred over أمير as the primary rendering; أمير retained only as a secondary gloss | Reverses the intuitive first choice; رئيس’s blandness is a feature, not a bug, here. |
| Ezekiel’s forensic “case-law” register in ch.18 and 33 | No dedicated Arabic religious-legal register distinct from إفتاء (Islamic legal-opinion process) exists that would not risk sounding like a fatwa-style argument. | No lexical solution; rely on discourse markers/structure, not specialized vocabulary | Deliberately avoid borrowing Islamic fiqh-register vocabulary (e.g., حكم شرعي) which would misleadingly cast the chapter as a legal ruling rather than prophetic argumentation. |
3. Crowded Semantic Neighborhoods Requiring Fencing
These are cases where Arabic vocabulary is not missing — quite the opposite, a ready-made word exists — but that word’s dominant contemporary or classical association belongs to a different, competing Islamic theological framework, requiring active fencing (explicit disambiguating notes, footnotes, or paired qualifiers) rather than passive reliance on context.
| Arabic Term | Competing Islamic Association | Fencing Strategy |
|---|---|---|
| جوج وماجوج (Gog and Magog) | Near-homophonic with يأجوج ومأجوج (Ya’juj wa-Ma’juj), Quran 18:94-98, 21:96-97 — a related but narratively distinct apocalyptic tradition (Dhul-Qarnayn’s barrier, unresolved eschatological sign vs. Ezekiel’s decisively-defeated invading power). | Mandatory theologian-reviewed footnote at first occurrence (ch.38): same name, distinct narrative traditions; do not harmonize. Never gloss Ezekiel’s Gog using Quranic Ya’juj-Ma’juj apocalyptic details (no wall, no Dhul-Qarnayn figure, different geography/timing). |
| القيامة (resurrection, reserved) vs. سأُصعدكم / أُقيمكم (Ezekiel 37’s raising language) | القيامة is both the fixed Romans-baseline term for Christ’s unique historical bodily resurrection AND the general Islamic Yawm al-Qiyamah (Day of Resurrection) doctrine shared by all people. | Absolute rule: never use القيامة (noun) for Ezekiel 37; use verb forms only. Explicit note distinguishing (a) 6th-century-BC national restoration from exile, (b) Christ’s unique historical resurrection, (c) the general Yawm al-Qiyamah — three related but non-identical referents. |
| الروح القدس (Holy Spirit) vs. bare روح (ruach) occurrences | Quran 2:87, 2:253 uses the identical phrase الروح القدس/روح القدس for a created being (commonly identified in mainstream tafsir with the angel Jibril), not a co-equal divine Person. | Reserve الروح القدس only for Ezekiel’s most explicit personal-divine-indwelling instances (36:27, 37:14), each requiring theologian-approved footnoting bridging to Romans 8/Acts 2; use روح / ريح / نَفَس elsewhere per context. Never let the mere presence of the word روح trigger automatic Trinitarian reading in the OT base text. |
| النفس التي تخطئ هي تموت (the soul who sins shall die, 18:4,20) | Genuinely and unusually convergent with Quran 6:164/17:15/35:18 (“no soul shall bear the burden of another”). | Fence against over-convergence: theologian note must state both truths together — individual accountability (Ezekiel 18) AND federal/Adamic solidarity in sin (Romans 5) — without letting Ezekiel 18 be used to dismiss or “resolve” the Romans argument in Islam’s favor. |
| يتوب / يرجع (turn/repent, shuv) | Conceptually adjacent to Islamic tawbah, a return to right standing through sincere repentance. | Fence against full equivalence: tawbah operates within a deeds-weighing (mizan) framework with no assured outcome; Ezekiel’s shuv, read canonically with Romans, issues toward grace-secured restoration. State this distinction explicitly wherever shuv appears in a doctrinally weighted passage (18:21-32; 33:11-19). |
| رجس / رجاسة (abomination, to’evah) | Positively overlaps with Quranic “rijs” (5:90) but that overlap is only partial. | Fence by scope: explicitly note to’evah’s broader range (cultic idolatry AND social injustice) versus the narrower ritual-impurity range typical of rijs usage, so the term is not read as exclusively ritual. |
| من أجل اسمي / اسمي القدوس (for my name’s/holy name’s sake) | Superficially resonant with Islamic tawhid’s emphasis on Allah’s self-sufficient transcendent honor (jalal, kabir). | Fence against a merit-implying misreading: the motive clause explicitly excludes Israel’s merit (20:9; 36:22,32) as the ground of restoration; the resonance with divine self-sufficiency is a legitimate bridge only if paired with this explicit denial of merit. |
| عهد سلام (covenant of peace) | Modern Arabic political register (اتفاقية سلام, “peace agreement/accord,” e.g. in regional news media) risks reading Ezekiel’s promise as a mere diplomatic settlement. | Fence via register: gloss explicitly as comprehensive relational restoration (health, security, divine presence), not a negotiated cessation of hostilities between two human parties. |
| موضع أخمصَي قدميّ (place of the soles of my feet, 43:7) | Islamic tanzih strongly resists any physically-embodied predication of God; an unqualified literal rendering could be read as a crude anthropomorphism (or dismissed as such). | Fence with the same analogical caution Ezekiel’s own visionary language models (ch.1’s repeated “likeness of,” “appearance of” hedges); explicit note that this is covenantal-presence idiom, not embodiment doctrine. |
| رئيس / أمير (prince, nasi) | أمير in particular can evoke a regional sovereign ruler or modern Gulf monarchic title; both terms sit near خليفة/إمام in the broader Islamic political-religious leadership semantic field. | Fence by consistently avoiding خليفة and its cognates entirely (per baseline kingdom_mission caution) and by explicit contrast-with-”king” notes (see §1 table above). |
| الشركة (fellowship, reused from baseline) — relevant background for “one nation, one king” and covenant unity themes | Shares the ش-ر-ك root with shirk, Islam’s cardinal sin (associating partners with God) — flagged already at the Romans baseline level. | No new fencing required beyond the baseline’s existing caution; noted here only because Ezekiel’s reunification/covenant-community language (37:15-28) intersects with this root-resonance risk. |
4. Transliteration vs. Paraphrase Decisions
| Item | Decision | Rationale |
|---|---|---|
| Gog and Magog | Transliterate: جوج وماجوج (not the Islamic-form يأجوج ومأجوج, not a descriptive paraphrase like “الأمة الشمالية المعادية”) | Transliteration preserves the proper-noun identity of Ezekiel’s specific oracle while remaining visually/phonetically distinct enough from the Quranic form to support (rather than erase) the required disambiguating footnote. Paraphrase would obscure the very collision that must be actively managed, not hidden. |
| Ben-adam (“son of man” address to Ezekiel) | Fixed established rendering: ابن آدم (literal), never ابن الإنسان | Paraphrase risks losing the address-formula’s repeated, formulaic force (~93 occurrences); transliteration is unnecessary since Arabic has a natural literal equivalent — but that literal equivalent must be actively fenced off from the distinct NT title ابن الإنسان (Daniel 7 / Christ’s self-designation), per the registry. |
| Adonai YHWH | Fixed established compound: السيد الرب (neither transliterated nor re-paraphrased) | Already an established Van Dyck-tradition rendering; introducing a transliteration (e.g., “أدوناي يهوه”) would break with two centuries of Arabic Bible translation convention and confuse readers moving between Ezekiel and other OT/NT materials. |
| Cherubim | Transliterate: الكروبيم (not paraphrased as “angelic guardians” or equated with generic ملائكة) | No precise Arabic/Islamic equivalent category exists; transliteration signals “this is a specific, technical class of being” and avoids the false-equivalence risk of collapsing cherubim into the general Quranic malā’ika category, which would erase Ezekiel’s specific throne-guardian imagery. |
| Chayot (“living creatures,” ch.1, 10) | Paraphrase, not the older transliteration/loan: كائنات حية (deliberately not حيوانات) | Older Van Dyck-era حيوانات has drifted in contemporary Modern Standard Arabic toward simply “animals,” an unintentionally demeaning register for beings attending God’s throne; a fresh paraphrase is required to preserve dignity, overriding historical precedent here. |
| Holy of holies | Fixed established compound: قدس الأقداس | Long-established Arabic Christian liturgical/translation compound; no decision needed, reuse as-is. |
| YHWH shammah (“the LORD is there”) | Paraphrase (translate the meaning), not transliterate the Hebrew phrase | This is a climactic statement, not a proper name to be preserved phonetically (contrast with Abba in Romans 8:15, which is preserved as an intimacy-marking interjection); translating as الرب هناك keeps the declarative, capstone force intact for a reading/teaching audience. |
| Kipper (“atone,” 16:63) | Paraphrase using existing Arabic Christian theological vocabulary: يكفّر / كفّارة (not a Hebrew transliteration) | Established Arabic Christian atonement vocabulary already exists and should be reused rather than transliterating an OT sacrificial term that has no independent currency in Arabic outside translated Scripture. |
| Nasi (“prince”) | Paraphrase, not transliterate: رئيس / أمير | A transliterated “nasi” would be meaningless to an Arabic reader with no Hebrew background; the semantic content (subordinate ruler, not king) must be conveyed through careful Arabic word choice, not phonetic preservation. |
| Abba-style intimacy markers | Not applicable in Ezekiel (no Aramaic intimacy term analogous to Romans 8:15 appears in this book) | Noted for completeness; no Ezekiel-specific decision required, but flagged so Phase 2 does not mistakenly import the Romans “retain untranslated” pattern where it does not apply. |
5. Ranked List of Highest-Risk Ambiguities
Ranked by combined severity (doctrinal weight × likelihood of misreading) and grounded in the specific collision or gap identified above. All Critical items require mandatory human theologian review at every occurrence per baseline routing conventions; High items require theologian review where doctrinally load-bearing and native-speaker review otherwise.
-
Ruach polysemy within Ezekiel 37:1-14 itself (Critical). The single core passage of the entire curriculum contains wind (v.9), breath (v.5,6,8,9,10), and Spirit (v.14) senses of the same Hebrew root within fourteen verses. A translator defaulting to one Arabic word throughout (especially a premature slide into الروح القدس) would either flatten the passage’s deliberate progression from physical to spiritual restoration, or, conversely, understate its climactic movement toward God’s personal indwelling Spirit in v.14. This is the single highest-stakes ambiguity in the whole book because it sits inside the anchor passage.
-
جوج وماجوج vs. يأجوج ومأجوج (Critical). Near-homophonic proper noun with a related but non-identical, textually distinct Islamic apocalyptic tradition. Highest collision risk of any single lexical item in the curriculum; failure to fence explicitly risks readers assuming the two traditions are the same narrative, or assuming Ezekiel’s passage is describing (or being fulfilled by) an unresolved Islamic end-times sign.
-
القيامة (reserved) vs. Ezekiel 37’s national-restoration raising language (Critical). Three distinct but easily conflated referents (Ezekiel’s exilic national restoration; Christ’s unique bodily resurrection; general Yawm al-Qiyamah) share overlapping vocabulary space. Verb-only discipline (أُقيمكم/سأُصعدكم) must be maintained without exception across every occurrence in ch.37, since this is precisely the passage most likely to be quoted devotionally and thus most exposed to drift.
-
النفس التي تخطئ هي تموت — the individual-accountability/federal-solidarity tension (Critical). Genuine convergence with Quranic no-inherited-guilt teaching creates a real risk that Ezekiel 18 will be deployed (intentionally or not) to settle the Romans 5 federal-headship question in Islam’s favor. Requires deliberate both/and framing at every teaching use, not just accurate translation.
-
قلب جديد / روح جديدة — divine unilateral gift vs. human imperative (Critical). The doctrinal tension between 36:26 (“I will give you”) and 18:31 (“make yourselves a new heart”) must not be silently resolved in translation toward either the grace side (erasing the imperative) or the self-effort side (inviting an Islamic dhikr/taqwa-style self-cultivation reading of the New Heart doctrine). This is a translation and curriculum-design risk simultaneously.
-
الروح القدس reservation discipline (Critical, downstream of #1). Every occurrence where a translator is tempted to render bare روح as الروح القدس outside the two theologian-approved climactic instances (36:27, 37:14) risks importing full NT Trinitarian conclusions into the OT base text prematurely, while under-using it at 37:14 risks losing the passage’s climactic force.
-
السيد الرب / الرب typographic and contextual distinguishability from Christological الرب (High). Because Ezekiel uses the identical Arabic word الرب (rendering YHWH) roughly 430 times, and Romans applies the same word Christologically to Jesus (10:9), cross-curriculum learners must be given explicit contextual signals (not a different word) to track which referent is active — a subtle but high-frequency risk given the sheer volume of occurrences in this book.
-
موضع أخمصَي قدميّ — embodiment idiom (High). Ezekiel’s own vision-language caution (ch.1’s repeated “likeness of,” “appearance of”) must be extended analogically to this later, more intimate idiom; omitting the analogical framing risks either offense (misread as crude embodiment claim) or accidental doctrinal overcorrection (softened into pure metaphor that loses the passage’s emphasis on God’s genuinely settled, permanent presence).
-
عهد سلام — contemporary political register collision (High). Given the ongoing regional weight of “peace agreements” in contemporary Arabic news discourse, and the pre-existing “israel” political-sensitivity caution already logged in the baseline and registry, this compound is unusually exposed to being read as commentary on present-day geopolitical peace processes rather than Ezekiel’s comprehensive covenantal restoration promise.
-
رئيس vs. أمير — subordinate-ruler vocabulary drifting toward khilafah-adjacent connotation (High). Especially relevant across ch.34, 44-46, where the “prince” motif recurs extensively; أمير’s contemporary monarchic associations in some Gulf contexts make it the riskier of the two options and it should be treated strictly as a secondary gloss, not a primary rendering.
-
أصنام — loss of gillulim’s contemptuous nuance (High). While doctrinally safe (both traditions treat idolatry as unambiguously wrong), the flattening of Ezekiel’s deliberately insulting wordplay is a genuine communicative loss across roughly 39 occurrences, cumulatively dulling one of the book’s sharpest rhetorical features.
-
يتوب / يرجع — shuv vs. tawbah equivalence drift (High). Repeated across the individual-responsibility chapters (18, 33); the risk compounds with frequency, since each individual occurrence is low-stakes but the cumulative teaching impression, if left unfenced, could implicitly assimilate Ezekiel’s repentance theology into a deeds-weighing (mizan) framework foreign to its grace-oriented canonical context.
-
مياه خارجة / نهر الحياة — river-of-life bridge vs. al-Kawthar syncretism (Medium-High). A genuinely positive bridge point (47:1-12) that nonetheless risks collapsing into generalized “paradise river” equivalence if the river’s specific christological/temple-presence source is not made explicit at first introduction.
-
الكروبيم / كائنات حية — dignity-of-register drift (Medium). Lower doctrinal stakes than the items above, but a real risk of unintentional demotion in reader perception (via the outdated حيوانات association) that would undercut the majestic tone Ezekiel’s throne-vision chapters require.
-
آية (sign-act, ot) vs. آية (Quranic verse) homograph (Medium). A structural ambiguity risk distinct from doctrinal collision — the identical Arabic word for “sign/miracle” and “a verse of the Quran” requires context to always make clear that ch.4, 12, and 24’s enacted prophetic dramas are sign-acts, not scripture-citations, particularly in any material that also quotes the Quran for comparative/apologetic purposes.
6. Full-Book Coverage Log
Every chapter of Ezekiel has been reviewed for load-bearing linguistic-gap content relevant to this analysis. Chapters not separately itemized above contributed no new gap or collision beyond what is already logged in §1–§5 or in the prior Step-7 glossary; they are recorded here as reviewed rather than silently omitted.
| Chapters | Status |
|---|---|
| 1 | Reviewed — living creatures/cherubim register gap (§4), embodiment-caution modeling referenced in §5 items 8, 9. |
| 2-3 | Reviewed — ben-adam address, watchman motif (contributes doctrine risk already logged in registry; no new gap beyond §1). |
| 4-7 | Reviewed — sign-act (ot) homograph risk (§5 item 15); no additional gaps. |
| 8-11 | Reviewed — glory departure narrative anchors §1’s Sovereignty/Presence entries; no new lexical gap beyond المجد’s discourse-level weakness already noted. |
| 12-14 | Reviewed — house of rebellion, divination vs. prophecy contrast (registry-logged); no new gap. |
| 15-17 | Reviewed — tender twig/sprig messianic imagery (registry-logged, background to seed_of_david); no new gap. |
| 18 | Reviewed in depth — individual responsibility vocabulary matrix (§1), ranked ambiguity item 4, 12. |
| 19 | Reviewed — lament genre marker; low risk, no new gap. |
| 20 | Reviewed — for my name’s sake motive clause (§3), law/statutes terminology (registry-logged). |
| 21-24 | Reviewed — sign-acts, vengeance, bloodguilt idiom; no new gap beyond registry entries. |
| 25-32 | Reviewed — oracles against the nations; Sheol/Pit vs. jahannam fencing (registry-logged); guardian cherub/seal of perfection disputed oracle (registry-logged); no new gap beyond existing entries. |
| 33 | Reviewed — watchman doctrine fully developed; shuv equivalence drift (§5 item 12). |
| 34 | Reviewed — shepherd/false shepherd imagery, prince vocabulary anticipation (§1, §4, §5 item 10). |
| 35-36 | Reviewed — vengeance, new heart/new spirit core statement (§1, §5 items 4, 5), for my name’s sake (§3). |
| 37 | Core passage — treated in full depth throughout §1-§5 (ranked ambiguity items 1, 3). |
| 38-39 | Reviewed in depth — Gog and Magog (§3, §5 item 2), highest-priority collision term in the book. |
| 40-42 | Reviewed — temple architecture vocabulary (sanctuary/temple, holy of holies); no new gap beyond §1’s rئيس-adjacent entries anticipated for ch.44-46. |
| 43 | Reviewed — glory’s return, place of the soles of my feet (§1, §3, §5 item 8). |
| 44-46 | Reviewed — prince vocabulary (§1, §4, §5 item 10), uncircumcised-in-heart bridge term (registry-logged). |
| 47 | Reviewed — river of life bridge point (§5 item 13). |
| 48 | Reviewed — YHWH shammah capstone statement (§1, §4). |
This analysis extends but does not alter the Romans baseline translation_memory.json/bible_term_registry.json. All Ezekiel-specific terms named here are cross-referenced to assets/bible_term_registry.json (prior output) and must be entered into the Phase 2 translation memory under the version-increment procedure specified in 12_ai_translation_requirements.md before any segment translation begins.