Zephaniah — Linguistic Gap Analysis (Hmong Language Package)

Status: Phase 1, Step 8 deliverable. This document extends, and never contradicts, the Romans Hmong baseline (assets/translation_memory.json, assets/bible_term_registry.json, assets/doctrine_risk_registry.json, 12_ai_translation_requirements.md) and the Zephaniah-curriculum documents already produced (analysis/08_core_glossary.md, assets/bible_term_registry.json for Zephaniah). Every term ID referenced below (TM-# for Romans Translation Memory keys, Z-# for the Zephaniah registry id field) is reused exactly as previously fixed; no alternate gloss is proposed for any Critical/High term without explicit escalation.

Core passage anchor: Zephaniah 3:14-17 (the joy-summons and its climax, “he will rejoice over you with singing”). This passage is the theological center of gravity for this analysis but not its scope boundary — all three chapters are covered per the full-book-coverage mandate.

---

0. Method and Scope

This analysis works from the two prior Zephaniah artifacts (08_core_glossary.md, the Zephaniah bible_term_registry.json) and asks four questions the PRD requires at Step 8:

1. Doctrine vocabulary matrix — for each of the four assigned doctrines, what Hmong vocabulary is available, where does it underperform, and what is the recommended mitigation strategy for curriculum writers (not translators of the base text — this package assumes the base Hmong Bible text already exists; the gap analysis governs curriculum language built around it).
2. Missing vocabulary vs. crowded semantic neighborhoods — where Hmong has no ready term (a gap) versus where Hmong has one overloaded term serving multiple biblical senses that must be fenced apart (a collision risk).
3. Transliteration vs. paraphrase decisions — for proper names and untranslatable concepts, is a loanword/transliteration or a descriptive paraphrase the safer choice, and why.
4. Ranked ambiguity list — the highest-risk items in Zephaniah specifically, ordered by combined doctrinal weight × collision severity, to direct reviewer attention.

---

1. Doctrine Vocabulary Matrix

1.1 Doctrine: The Day of the Lord as Judgment and Restoration

Available Hmong term(s)	Weakness	Recommended strategy
Hnub ntawm tus Tswv (Z-4, NEW fixed collocation)	Hmong has no pre-existing fixed idiom for a single eschatological “Day” bearing two opposite theological valences (terror and joy) within one book. The literal Hmong reading of “hnub” (day/sun) carries no inherent judgment-or-restoration connotation on its own — all doctrinal weight is carried by the surrounding teaching apparatus, not the noun phrase itself.	Treat Hnub ntawm tus Tswv as a single, non-negotiable fixed collocation used identically in ch.1 (terror movement) and ch.3 (joy movement). Every lesson introducing either movement must explicitly state “this is the same Day,” using a bridging sentence (e.g., “Tib hnub uas ntxhov siab hauv tshooj 1, yog tib hnub uas muaj kev zoo siab hauv tshooj 3”) rather than allowing the two movements to read as two unrelated topics.
nyob rau hnub ntawd (Z-7, “on that day”)	Functions only as a continuity marker pointing back to Z-4; on its own it is doctrinally empty and could be mistaken for an unrelated future reference if the ch.1↔ch.3 link is not made explicit.	Never introduce Z-7 without first re-establishing Z-4 in the same lesson unit. Pair every occurrence with an explicit backward-reference footnote.
tej kev txiav txim rau koj (Z-5, judgments against you)	Directional ambiguity: the identical phrase describes both judgment announced (ch.1) and judgment revoked (3:15). Hmong has no morphological marker (tense/aspect prefix) that would disambiguate “sentence pending” from “sentence lifted” the way, e.g., English can lean on “removed” vs. “coming.”	Mandatory: every occurrence must carry an explicit verb of direction — tabtom los rau (coming upon) for the ch.1 sense, raug rho tawm lawm (has been removed) for the 3:15 sense. Never leave the bare noun phrase to carry the directional load alone.
kev npau taws ntawm Vajtswv (TM-inherited, Z-6)	Inherited High risk from Romans baseline: capriciousness collision with ancestor/spirit anger.	Reuse Romans baseline’s mandatory contrast-teaching verbatim; no new mitigation required, but every Zephaniah ch.1 lesson touching wrath must re-surface it (reading-level re-gloss rule, §6 of 12_ai_translation_requirements.md).
kev cawm dim / cawm (TM-inherited, Z-38)	Inherited High risk: “dim” (escape) parallels shamanic soul-rescue (ua neeb).	Anchor explicitly to tus Tswv in the same clause (3:17’s own grammar already does this — exploit it, do not weaken it in paraphrase).

Doctrine-level gap finding: There is no missing vocabulary here — every needed lexical item exists in Hmong or has already been fixed by TM inheritance or the Z-4/Z-5/Z-7 new collocations. The risk is entirely structural/pedagogical: Hmong’s lack of built-in tense-marking on nominal judgment-vocabulary means curriculum prose, not lexical choice, must carry the disambiguating weight. Risk ceiling: Critical (matches doctrine risk registry expectation below).

---

1.2 Doctrine: God’s Judgment on Idolatry and Complacency

Available Hmong term(s)	Weakness	Recommended strategy
dab mlom (TM-inherited, Z-8)	None new — correctly reuses the Romans-fixed, doctrinally-accurate term for venerated spirit-images.	No change; reuse verbatim.
Npa-as / Minkoos / cov pov thawj dab (Z-9, Z-10)	Proper names have zero native Hmong lexical equivalent — pure gap, not a collision. cov pov thawj dab risks being read as a general category term (any idol-priest, anywhere) rather than the specific historical Ammonite/Canaanite priesthood Zephaniah names, if the proper-name anchor (Npa-as/Minkoos) is dropped in paraphrase.	Transliterate the proper names and always keep the proper name adjacent to the priest/idol-category noun in the same clause on first mention per lesson (never let cov pov thawj dab stand alone without its named referent close by).
cov hnub qub saum nruab ntug (raug pe hawm) (Z-11)	Comprehension gap only (no Hmong star-cult analogue) — genuinely missing vocabulary domain, not a crowded one.	Use the descriptive paraphrase as-is; no fixed collocation needed since there is no competing traditional referent to fence against. Low investment item.
kev tua khoom fij / dawb huv nws cov qhua (Z-12)	Critical — flagged [TRANSLATOR REVIEW NEEDED] already in the Zephaniah registry. This is a crowded-neighborhood problem: Hmong fij/dawb huv vocabulary is already load-bearing for TWO Romans-fixed senses (Christ’s atonement; ethical living-sacrifice) and Zephaniah introduces a third sense — God “sacrificially consecrating” hostile pagan armies as instruments of judgment.	Do not resolve this gap unilaterally. This document formally logs it as unresolved pending human theologian review (see §4 escalation list, item R-1). Curriculum content MUST NOT publish material using this sense until reviewed.
cov neeg uas nyob tsis quav ntsej, xav tias Vajtswv yuav tsis ua dabtsi (Z-15, complacency)	Direct collision risk with Hmong fatalism (hmoov), a vocabulary family the Romans baseline permanently forbids for describing divine action. The literal Hmong idiom for apathy (“not caring/not paying attention”) does not, on its own, distinguish “I don’t care because nothing changes anything” (fatalism) from “I don’t care because I’ve wrongly assumed God is passive” (Zephaniah’s actual target).	Mandatory explicit contrast clause in every lesson: state directly that Zephaniah condemns apathy toward an active God, and is the theological opposite of fatalism, never its equivalent. Never allow this phrase to stand without the contrast sentence attached — this is a stronger requirement than a mere footnote; it must be structurally present in the same paragraph.
Vajtswv txoj kev khib siab rau nws haiv neeg (Z-16, fire of jealousy/zeal)	khib siab in ordinary usage = petty envy/possessiveness among people (e.g., spousal jealousy, sibling rivalry over inheritance). Using it of God without a qualifying clause risks a hearer imagining God as insecure or petty.	Never use bare khib siab for God’s zeal; always pair with an explanatory relative clause distinguishing covenant-protective zeal from human possessiveness (e.g., “…tsis zoo li kev khib siab ntawm tibneeg, tiamsis yog nws txoj kev tsom kwm nws haiv neeg kom lawv tsis mus pehawm luag tej dab” — “not like human jealousy, but his determination to guard his people from worshiping others’ idols”).
lus dawb huv hauv lawv qhov ncauj (Z-14, pure lips/speech)	Reuses the dawb huv root already flagged for the sanctification-vs-ritual-purity collision (Romans TM-46) and now for a THIRD context (worship-speech). This is a genuinely crowded neighborhood: dawb huv must now serve moral-transformation (Romans ch.6), and worship-speech-purity (Zephaniah 3:9), while never sliding into folk ritual-cleansing.	Apply the same disambiguation strategy used for Romans sanctification: teach explicitly that this is progressive, eschatological transformation of what is said in worship (calling on the LORD truthfully, together, replacing idol-invocation), not a magical speech-cleansing rite.

Doctrine-level gap finding: This doctrine has the curriculum’s single largest crowded-neighborhood problem (fij/dawb huv, now carrying four distinct senses across Romans + Zephaniah) and its clearest fatalism-collision risk (hmoov vs. complacency). Risk ceiling: Critical.

---

1.3 Doctrine: The Remnant of Israel

Available Hmong term(s)	Weakness	Recommended strategy
cov seem (TM-inherited, Z-21)	None — stable, low risk.	Reuse verbatim.
kev txo hwjchim / cov txo hwjchim (Z-23, humility)	Cross-curriculum trap, not a traditional-religion collision: this idiom uses hwjchim in its personal-status/authority sense, a completely different semantic field from the Romans-flagged hwjchim = supernatural/shamanic power (dynamis), which mandatorily pairs with “Vajtswv.” A curriculum writer trained on the Romans package’s hwjchim-qualifier rule could wrongly force “Vajtswv hwjchim” onto this idiom, producing a nonsensical or theologically confused phrase (“Vajtswv’s power-lowering” instead of “lowering one’s own status”).	Insert an explicit disambiguating footnote at first occurrence in any Zephaniah lesson: “This hwjchim refers to personal social status, NOT the divine/supernatural power sense flagged elsewhere in this curriculum family; the ‘Vajtswv hwjchim’ qualifier rule does NOT apply here.” This is a documentation gap, not a lexical gap — the word choice itself is fine.
ncaj ncees bare, no reckoning-qualifier (Z-20, ethical “seek righteousness” sense)	Cross-curriculum trap in the opposite direction from Z-19/TM-25: Romans’ ncaj ncees is Critical and requires the forensic qualifier raug suav tias ncaj ncees at every occurrence. Zephaniah 2:3 uses the same Hmong adjective in its OT ethical-behavioral sense (practicing justice), where importing the Romans forensic qualifier would be a doctrinal category error.	Document explicitly, in every lesson using Z-20, that the Romans qualifier rule is suspended for this specific verse/sense — this must be a conspicuous exception flag, not a silent omission, since automated qualifier-enforcement tooling built for the Romans package would otherwise incorrectly “auto-correct” this verse.
nrhiav tus Tswv (Z-22, seek the LORD) / tos kuv (Z-24, wait for me) / tso siab rau tus Tswv lub npe (Z-25, trust in the name)	No collision; mild redundancy risk (three near-synonymous phrases for faith-posture in close proximity within ch.2-3) could blur together for learners if not distinguished.	Teach as a deliberate triad (seeking, waiting, trusting) rather than three interchangeable synonyms — curriculum should name the triad explicitly so redundancy becomes reinforcement rather than confusion.
pub kom rov qab muaj huv si dua (Z-27) / kuv yuav sau cov raug ntiab tawm... (Z-28) / kuv yuav coj koj rov qab los tsev... (Z-29)	No collision — genuinely positive cultural bridge via Hmong refugee/diaspora displacement-and-return experience.	No mitigation needed; flag as a pastoral opportunity for curriculum writers (illustrative testimonies of literal homecoming may be used to teach this vocabulary without doctrinal risk).

Doctrine-level gap finding: No missing vocabulary; the risk here is entirely documentation/tooling risk — two separate places where Romans-trained automated qualifier-enforcement (per 12_ai_translation_requirements.md §2, rule 2) would incorrectly fire if applied mechanically to Zephaniah text without the exceptions in this document being consulted first. Risk ceiling: High (elevated from the glossary’s own “High” ceiling specifically because of this qualifier-collision risk, which is a new category this analysis is the first to name).

---

1.4 Doctrine: God’s Joyful Love for His People

Available Hmong term(s)	Weakness	Recommended strategy
kev hlub (TM-inherited, Z-40)	None — stable.	Reuse verbatim.
nws yuav zoo siab kawg li rau koj / nws yuav hu nkauj qw zoo siab rau koj (Z-41, God rejoicing over his people with singing)	No lexical gap, but a significant conceptual gap: Hmong has ready vocabulary for joy/singing, but the theological category of a deity joyfully delighting in worshipers — apart from ritual performance or appeasement outcomes — has no traditional-religion analogue to borrow from. Hmong ancestor/household-spirit relationships are transactional (offerings secure protection/luck; anger is placated by correct ritual); no traditional spirit is depicted as spontaneously, unconditionally delighting in the person it relates to.	This absence is a teaching opportunity, not a lexical problem — but it must be surfaced explicitly, not left implicit, since silence would let hearers unconsciously map this verse onto the nearest available schema (transactional spirit-favor) by default. Every lesson on 3:17 must include an explicit contrast statement: God’s joy here is not a reward for ritual correctness, and is not contingent on the worshiper’s continued appeasement — it flows from his own covenant love (kev hlub, Z-40) already declared its cause in the same verse.
nyob hauv nruab nrab ntawm koj (Z-36, “in your midst”)	Must be kept lexically distinct from the Romans NT fixed formula nyob hauv Khetos — related theological reality (divine presence), different covenant-administration-specific formula. Risk is a curriculum writer collapsing the two into one collocation for stylistic reasons (“since they mean basically the same thing”).	Explicit rule, not just a note: never substitute or merge nyob hauv nruab nrab ntawm koj and nyob hauv Khetos. A bridging teaching sentence connecting the two (“God who dwelt among his people now dwells within them through Christ”) is permitted and encouraged, but the two fixed phrases themselves must never be interchanged.
tus muaj zog loj (Z-37, mighty one/warrior)	Correctly and deliberately not mapped onto the Romans-flagged hwjchim. No weakness identified — this is a case of successful gap-avoidance already executed in the prior registry.	No mitigation needed; flag here only to confirm the decision is documented and stable for future curriculum reuse.
nws yuav ua rau koj nyob twjywm ... vim nws txoj kev hlub (Z-39, “will quiet you”)	Genuine Hebrew text-critical ambiguity (MT yacharish, “he will be silent,” vs. widely-followed emendation “he will renew/quiet you”), not a Hmong-language gap.	Footnote the Hebrew ambiguity explicitly; do not present the Hmong rendering as settled beyond dispute the underlying Hebrew itself does not settle.

Doctrine-level gap finding: This doctrine has no lexical gaps and no direct traditional-religion collision term to fence, but it carries the curriculum’s most significant conceptual/worldview gap: Hmong traditional religion supplies no schema for a deity’s unconditional relational delight. This is flagged as a distinct risk category from lexical collision (§4 below ranks it accordingly). Risk ceiling: Low-Medium lexically, but High pastorally/catechetically — the doctrine risk registry should treat this doctrine’s teaching requirement as seriously as the lexical-collision doctrines even though no single word is at fault.

---

2. Missing Vocabulary vs. Crowded Semantic Neighborhoods

2.1 Genuinely Missing Vocabulary (true gaps — no native term exists, comprehension-level only)

These require descriptive paraphrase, not fencing, because there is no competing traditional-religion referent to collide with:

Item	Registry ID	Nature of gap
Host of heaven / astral worship	Z-11	No Hmong star-cult practice to map onto or collide with.
Foreign national proper names (Philistia, Moab, Ammon, Cush, Assyria, Nineveh)	Z-45	Pure onomastic gap; transliteration is the only option.
Baal, Milcom (as specific named deities)	Z-9, Z-10	Proper names with no Hmong equivalent; gap requires transliteration + gloss, not paraphrase alone.
God’s unconditional relational delight (theological category underlying Z-41)	Z-41	Conceptual gap in the surrounding worldview, not a missing word — Hmong has the words for joy/singing, but no schema for a non-transactional divine-human relationship of this kind.
”Silent/quiet in love” idiom (Z-39)	Z-39	Exegetical gap in the source text itself, not a Hmong deficiency.

2.2 Crowded Semantic Neighborhoods (one Hmong term/root serving multiple, doctrinally distinct senses — require active fencing)

This is the single most important structural finding of this analysis. Three Hmong lexical roots are now each carrying three or more distinct senses across the combined Romans + Zephaniah curriculum family, and each requires an explicit fencing strategy so curriculum writers do not collapse senses into one another:

(a) The fij / sacrificial-offering root

Sense	Curriculum	Hmong rendering	Status
1. Christ’s unique, unrepeatable atonement	Romans 3:25	qhov theej txhoj kev txhaum	Fixed, Critical, TM-permanent
2. Ethical/devotional “living sacrifice”	Romans 12:1	khoom fij uas muaj txoj sia nyob	Fixed, High, TM-permanent
3. God “consecrating” hostile nations as judgment-instruments	Zephaniah 1:7-8	kev tua khoom fij / dawb huv nws cov qhua	UNRESOLVED — escalated, see §4 R-1

Fencing requirement: all three senses must remain visually/lexically distinguishable in any single curriculum module that touches more than one (e.g., a combined Romans-and-Zephaniah devotional series). Sense 3 must not be published pending review.

(b) The dawb huv (pure/holy) root

Sense	Curriculum	Hmong rendering	Status
1. Progressive Spirit-wrought moral sanctification	Romans 6	kev raug ua kom dawb huv	Fixed, High, TM-permanent
2. Folk ritual purity from spirit-contamination (the FORBIDDEN reading, not a valid sense but the constant collision risk)	(traditional religion)	—	Permanently guarded against
3. Purified worship-speech	Zephaniah 3:9	lus dawb huv hauv lawv qhov ncauj	Fixed this curriculum, High
4. Consecration-for-judgment (overlaps with fij sense 3 above)	Zephaniah 1:7	(part of Z-12)	Unresolved, tied to R-1

Fencing requirement: every occurrence of dawb huv in any lesson must specify which sense is active in its immediate co-text; the word must never appear without a sense-anchoring clause, mirroring the Romans baseline’s own sanctification-disambiguation rule but now applied across two senses instead of one.

(c) The hwjchim (power/authority/status) root

Sense	Curriculum	Hmong rendering	Status
1. Divine/supernatural power (shamanic-power collision)	Romans 1:16, 15:19	Vajtswv hwjchim (mandatory qualifier)	Fixed, High, TM-permanent
2. Delegated civil authority	Romans 13	cov muaj hwjchim kav teb chaws	Fixed, High, TM-permanent
3. Personal social status/rank (unrelated meaning)	Zephaniah 2:3, 3:12	kev txo hwjchim (humility)	Fixed this curriculum, Medium

Fencing requirement: sense 3 is genuinely a different word-sense (homonym-like divergence within one lexical item) and does not require the “Vajtswv” qualifier; the risk is tooling/writer confusion, not audience confusion — the explicit disclaimer in §1.3 above is the fencing mechanism.

2.3 Summary Table: Gap Type Classification

Category	Count of items	Representative examples
True missing vocabulary (paraphrase-only, no collision)	5	Z-11, Z-45, Z-9/Z-10 (names), Z-41 (concept), Z-39 (exegetical)
Crowded neighborhood requiring active fencing	3 root families (11 distinct senses total across Romans+Zephaniah)	fij, dawb huv, hwjchim
Cross-curriculum qualifier-rule traps (documentation gap, not lexical gap)	2	Z-20 (ncaj ncees without qualifier), Z-23 (hwjchim = status, not power)

---

3. Transliteration vs. Paraphrase Decisions

Item	Decision	Rationale
YHWH → tus Tswv	Paraphrase (title), not transliteration	Follows Romans baseline exactly (kyrios → tus Tswv); explicitly forbids NWT-style divine-name transliteration (“Yehauvas”) per inherited Romans forbidden-substitution rule. Applies identically whether the Hebrew source says YHWH or Adonai.
Baal, Milcom	Transliteration + mandatory first-use gloss (Npa-as, Minkoos)	These are named rival deities, not generic categories; a paraphrase (“an idol”) would erase the specific historical indictment that Judah worshiped this particular, named, foreign god. Transliteration preserves specificity; the gloss supplies comprehension.
Foreign nations (Philistia, Moab, Ammon, Cush, Assyria, Nineveh)	Transliteration only, no paraphrase needed	Pure geographic/ethnic proper nouns with no doctrinal content of their own; paraphrase would be both unnecessary and impossible (there is no descriptive substitute for a place name).
Zion, Jerusalem (as “Daughter of…”)	Transliteration of place name (Xi-oo, Yeluxalees) + literal-personification retained, not paraphrased away	The personification device (“Daughter of Zion”) is theologically and poetically significant (a city addressed with filial tenderness, anticipating 3:17’s parental joy-imagery); collapsing it into a paraphrase like “the people of Zion” would flatten the affective force the doctrine of God’s Joyful Love depends on. A comprehension-level footnote on the personification device is preferred over paraphrase-substitution.
Yom YHWH → Hnub ntawm tus Tswv	Paraphrase (compositional phrase), not transliteration	A transliterated “Yom” would be opaque and untaught; the compositional Hmong phrase using already-fixed tus Tswv correctly inherits stability from the Romans baseline while still requiring the fixed-collocation discipline documented in §1.1.
Abba-style Aramaic retention pattern (cf. Romans TM-51)	Not applicable in Zephaniah — no Aramaic loanwords occur in this Hebrew-source book	Logged here only to confirm the pattern was checked and found not to recur, avoiding a silent omission.

General transliteration policy for this curriculum: transliterate proper names (deities, peoples, places) that carry no independent doctrinal content beyond their referential specificity; paraphrase (with fixed-collocation discipline where doctrinally load-bearing) everything else, exactly mirroring the Romans baseline’s own practice (Yexus Khetos, Povlauj transliterated; charis, dikaiosynē paraphrased).

---

4. Ranked List of Highest-Risk Ambiguities

Ranked by combined doctrinal weight (how central to the four assigned doctrines) × collision severity (how easily the ambiguity inverts or neutralizes the doctrine if mishandled):

1. R-1 — fij/dawb huv third sense (God consecrating armies as judgment-instruments, Z-12, Zephaniah 1:7-8). Highest risk in the entire curriculum: it is the least-resolved item (formally flagged [TRANSLATOR REVIEW NEEDED] in the prior registry and NOT resolved here), it sits in the already-most-crowded lexical neighborhood in the whole Romans+Zephaniah family, and a mishandling would blend Judgment-on-Idolatry doctrine with either Christ’s atonement or ethical-sacrifice doctrine from Romans — a cross-book doctrinal contamination risk unique to this item. Action: escalate to human theologian review before any publication using this verse-range; do not resolve via this document.

2. R-2 — Complacency vs. fatalism collision (Z-15, Zephaniah 1:12). Second-highest: directly engages the Romans baseline’s permanently forbidden hmoov vocabulary family, and the doctrinal content is a precise inversion risk (Zephaniah condemns apathy toward an active God; a mishandled rendering would sound like an endorsement of fatalism, exactly backward). Structural mitigation exists (§1.2) but requires disciplined, non-optional application in every single occurrence, making it high residual risk if curriculum writers are inattentive.

3. R-3 — Hnub ntawm tus Tswv directional/continuity fracture (Z-4, Z-5, Z-7, across ch.1 and ch.3). Third-highest: the doctrine of Day of the Lord as Judgment AND Restoration is architecturally dependent on Hmong prose (not lexical choice) correctly signaling that ch.1’s terror and ch.3’s joy are one Day, not two. Since Hmong lacks grammatical tense-marking to carry this automatically, the entire doctrinal burden rests on curriculum-writing discipline — a single lesson written in isolation (e.g., a stand-alone ch.3 lesson without the ch.1 back-reference) could silently sever the doctrine’s unity without anyone perceiving an error, since each piece is individually well-formed.

4. R-4 — Divine jealousy/zeal vs. human khib siab (Z-16). Fourth: a real, specific collision (God sounding petty/insecure) with moderate frequency (ch.1 and ch.3) and a clear, documented mitigation (mandatory contrast clause) — ranked below R-1/R-2/R-3 because the mitigation is simpler and more mechanically enforceable (a single relative clause) than the structural, multi-lesson discipline R-2 and R-3 require.

5. R-5 — ncaj ncees qualifier-suspension trap (Z-20, Zephaniah 2:3) vs. Z-19 (3:5, which DOES require the qualifier). Fifth: risk is narrow in scope (one verse) but high in automation danger — any Romans-trained enforcement tooling or reviewer habit would, by default, incorrectly demand the forensic qualifier here, actively introducing an error where none previously existed. Ranked below R-1–R-4 because the doctrinal content itself (OT ethical exhortation) is less central to this curriculum’s four assigned doctrines than judgment/idolatry/day-of-the-lord material, but ranked above lower items because the failure mode is a positive intervention that creates an error rather than a passive omission.

6. R-6 — hwjchim sense-3 (personal humility/status, Z-23) mistaken for sense-1 (divine power, requiring “Vajtswv” qualifier). Sixth: same class of automation/tooling risk as R-5, but lower doctrinal stakes (humility exhortation in the Remnant doctrine is important but not as load-bearing as the Law/idolatry material), and the failure mode (an awkward, obviously wrong phrase like “Vajtswv-power-lowering”) is more likely to be self-evidently strange and caught in review than R-5’s more plausible-sounding error.

7. R-7 — dawb huv sense-3, purified worship-speech (Z-14, Zephaniah 3:9), sliding toward folk ritual-cleansing. Seventh: real collision risk but lower severity than R-1 because this sense does not additionally overlap with the atonement/sacrifice domain (it is a two-sense collision — sanctification vs. folk purity — not the three-to-four-way collision of R-1), and the existing Romans-derived disambiguation strategy transfers directly with minimal adaptation.

8. R-8 — Absence of a relational-delight schema for God’s joy (Z-41, Zephaniah 3:17, the core-passage climax). Eighth: ranked lower than the above not because it is doctrinally unimportant — it is arguably the single most theologically climactic verse in the book — but because it is a conceptual/catechetical gap requiring explicit teaching, not a lexical collision risk requiring term-fencing; the mitigation (an explicit contrast-with-transactional-spirit-relationships statement) is well-understood and low-ambiguity to execute, even though skipping it would be a serious pastoral loss.

9. R-9 — be-qirbekh (“in your midst,” Z-36) vs. Romans’ nyob hauv Khetos merger risk. Ninth: real but narrow risk (a stylistic collapsing of two fixed phrases), affecting register/consistency more than core doctrinal content, since the underlying theological reality (God’s presence with/in his people) is compatible across both Testaments’ formulas even if the phrases must not be lexically merged.

10. R-10 — Proper-name transliteration drift (Z-9, Z-10, Z-45) if glosses are dropped in abbreviated curriculum formats (e.g., quiz answer options, headers). Tenth (lowest ranked among logged risks): comprehension-level only, easily corrected, no doctrinal inversion possible even in worst-case mishandling — included in the ranking only to ensure full-book coverage per the PRD mandate, not because it approaches the severity of R-1–R-9.

---

5. Chapter Coverage Confirmation

All three chapters of Zephaniah have been reviewed for this linguistic gap analysis, consistent with the full-book-coverage mandate:

• Chapter 1 — primary source of Judgment-on-Idolatry vocabulary (idols, Baal, Milcom, host of heaven, complacency, fire of jealousy) and the terror-movement half of the Day-of-the-LORD doctrine. Fully analyzed above (§1.1, §1.2).
• Chapter 2 — primary source of Remnant-of-Israel vocabulary (seek the LORD, humility, righteousness-ethical-sense, restore fortunes) and the universal-scope nations material (proper-name transliteration decisions). Fully analyzed above (§1.3, §3).
• Chapter 3 — contains the core passage (3:14-17) and is the primary source of God’s-Joyful-Love vocabulary, the restoration-movement half of the Day-of-the-LORD doctrine, and the bulk of the crowded-neighborhood dawb huv/fij items originating structurally in ch.1 but resolved/echoed in ch.3 (e.g., pure lips, 3:9). Fully analyzed above (§1.1 restoration column, §1.4, §2.2).

No chapter or sub-section is silently omitted; verses contributing no additional load-bearing vocabulary beyond what is cataloged in 08_core_glossary.md and the Zephaniah term registry (e.g., extended geographic detail in 2:4-6, the superscription in 1:1) were reviewed and confirmed to introduce no further gap-analysis items.

---

6. Escalation Summary (feeds forward to doctrine risk registry / AI requirements documents)

ID	Item	Required action	Routing
R-1	fij/dawb huv third sense (judgment-consecration, Z-12)	Do not publish; escalate for formal TM-revision review	Human theologian
R-2	Complacency/fatalism contrast (Z-15)	Mandatory structural contrast clause at every occurrence; no automated bypass	Human theologian (per Critical-tier routing)
R-3	Day-of-the-LORD continuity discipline (Z-4/5/7)	Mandatory cross-referencing sentence in every lesson touching either movement	Human theologian (per Critical-tier routing)
R-5	ncaj ncees qualifier-suspension at Zephaniah 2:3	Document as standing exception to Romans automated qualifier rule	Native speaker + theologian sign-off
R-6	hwjchim sense-3 vs. sense-1 disambiguation	Insert disclaimer footnote at first occurrence in curriculum	Native speaker review sufficient

This table should be treated as the authoritative hand-off to the next PRD step (doctrine risk registry construction and AI translation requirements drafting for Zephaniah), which must incorporate R-1 through R-6 without silently resolving any of them beyond what is documented here.