Joshua — Linguistic Gap Analysis (Hmong Language Package)

0. Purpose and Method

This document performs the linguistic gap analysis mandated by PRD Phase 1 Step 8: for each of the seven assigned doctrines, it inventories the Hmong vocabulary currently available (inherited Romans TM entries plus the new Joshua-specific candidates proposed in 08_core_glossary.md / assets/bible_term_registry.json), names the specific weakness of each option, and recommends a concrete disambiguation strategy. It then separates the book’s terminology problems into two structurally different categories — missing vocabulary (concepts Hmong has no ready word for at all) versus crowded semantic neighborhoods (concepts where Hmong has a ready word, but that word is already occupied by a competing traditional-religion referent and therefore requires fencing). It records the transliteration-vs-paraphrase decision for every proper noun and technical term newly introduced by Joshua. Finally it produces a ranked list of the book’s highest-risk ambiguities to drive Phase 1 Step 9 (doctrine risk registry) and downstream Phase 2 enforcement.

Nothing in this document overrides assets/translation_memory.json. Every REUSED term retains its Romans-baseline rendering unchanged; this analysis only evaluates NEW Joshua terms and the interaction between NEW and REUSED terms.

---

1. Doctrine Vocabulary Matrix

1.1 God’s Faithfulness to the Land Promise

Available term(s)	Source	Weakness	Recommended strategy
lus cog tseg (promise)	REUSED, Romans TM promise	None on its own — stable, unilateral-pledge sense already established for Rom 4, 9, 15.	Retain unmodified; anchor every occurrence in Joshua to the patriarchal covenant (Gen 12, 15) explicitly, not merely to “a promise” in the abstract.
thaj av uas Vajtswv cog lus tseg (Promised Land)	NEW, registry #21	Built on lus cog tseg, but the noun av (“land/earth/soil”) independently carries deep Hmong ancestral-land-veneration resonance — land as the dwelling-place of ancestor spirits and the site of ritual land-blessing (hu plig av, fiv yeem rau thaj av). Left unanchored, “the land God promised” risks being heard as “ancestral territory requiring ongoing ritual maintenance” rather than a historical, redemptive gift.	Never use bare thaj av for the Promised Land; always retain the full compound with the cog lus tseg qualifier, and pair first occurrences with an explicit teaching note distinguishing God’s unilateral covenantal gift from ancestral land-spirit custodianship.
tsis tau tso tawm/ntes tag (land not yet fully possessed)	NEW, registry #54	Comprehension gap only — no traditional-religion collision, but a theological flattening risk: without this term, faithfulness could be over-read as instant, total, triumphalist fulfillment (contradicted by 13:1,13; 15:63; 23:4-5,12-13).	Must be taught alongside every “faithfulness to the land promise” lesson to preserve Scripture’s own already/not-yet honesty; omitting it risks producing a doctrine of promise more triumphalist than the text itself supports.
qub txeeg qub teg (inheritance/portion)	REUSED root, Romans TM heir; registry #22	Positive bridge, minimal weakness — Hmong lineage-inheritance custom maps well onto tribal land-allotment. Minor risk: Hmong inheritance customs are patrilineal and could be read as implying the land is earned/deserved by blood-descent merit rather than granted by grace-based promise.	Pair with explicit teaching that the land-inheritance is granted on the basis of God’s oath to Abraham, not Israel’s merit (cf. Deut 9:4-6, echoed implicitly throughout Joshua).

Doctrinal note: No Critical-tier lexical collision exists here (unlike, e.g., grace/koob hmoov in Romans). The risk is entirely in the land-veneration resonance of av and the triumphalism risk of under-teaching #54. Recommend High, not Critical, risk classification for this doctrine’s core terms — the danger is theological flattening and syncretistic land-reverence drift, manageable with consistent qualifier discipline, not an unavoidable lexical collision.

1.2 Holy War and Divine Judgment

Available term(s)	Source	Weakness	Recommended strategy
khoom/tej uas raug muab cais rau Vajtswv kom puas tsuaj tag nrho (cherem)	NEW, registry #6	The single highest-risk term in the book. No existing Hmong lexeme covers “irrevocably devoted to destruction as a category of judgment,” so a periphrastic construction was built — but its individual components (muab cais, “set apart,” and puas tsuaj, “destroy utterly”) independently echo two competing domains: (a) ritual curse-transfer, where cursed/contaminated objects are ritually set apart and destroyed to remove misfortune from a household, and (b) shamanic spirit-appeasement destruction-offerings.	Mandatory theologian review at every occurrence (already flagged Critical in the registry). Every occurrence must carry a co-text note establishing cherem as (i) a unique, temporally-bounded act of divine judicial judgment at a specific redemptive-historical moment, (ii) categorically distinct from ongoing ethical norms, spirit-appeasement, or ritual curse-removal, and (iii) never a template for extending violence.
Vajtswv tau ua rog pab cov Yixalayees (the LORD fought for Israel)	NEW, registry #5	Shares surface grammar with the pattern of invoking a spirit-general’s aid in shamanic ritual combat (ua neeb tuag rog) — hearers may assimilate “God fought for Israel” into “Israel’s shaman successfully recruited a stronger spirit-ally.”	Must always retain Israel as recipient, never initiator, of the action; avoid any phrasing that suggests Israel ritually induced or negotiated divine intervention. Teaching note required distinguishing sovereign divine initiative from invoked/negotiated spirit-power.
kev npau taws ntawm Vajtswv (wrath of God)	REUSED, Romans TM wrath_of_god	Already documented Romans-baseline High risk (capricious ancestor/spirit-anger collision); Joshua’s judgment narratives (Jericho, Ai, the Canaanite cities) are the most vivid narrative enactment of this wrath anywhere in either Testament, raising the stakes on the existing Romans caution rather than introducing a new one.	Reuse unmodified; but flag every Joshua narrative occurrence for the same “not appeasable by ritual, not capricious, judicial not arbitrary” contrast-teaching already mandated in the Romans package, now applied to concrete historical judgment rather than abstract doctrine.
txhaum / kev txhaum (sin, corporate)	REUSED, Romans TM sin_act_and_condition; registry #49	Achan’s episode (ch. 7) surfaces a genuinely positive extension: corporate/representative guilt-consequence, which reinforces (does not contradict) the Critical-risk federal-headship caution already flagged for Romans 5.	No new fencing needed; cross-reference explicitly to the Romans Original Sin doctrine as a canonical parallel, strengthening rather than complicating the existing caution.

Doctrinal note: This is the book’s most linguistically dangerous doctrine because it requires inventing vocabulary (cherem) rather than merely disambiguating existing vocabulary. Recommend Critical risk tier, consistent with the registry’s #6 flag, and recommend this doctrine receive the single largest teaching-note budget in the curriculum.

1.3 Courage through God’s Presence

Available term(s)	Source	Weakness	Recommended strategy
khov kho thiab tawv siab (be strong and courageous)	NEW, registry #42	Positive cultural bridge to Hmong warrior-courage values, but Hmong battlefield courage is traditionally reinforced by invoking ancestral protector-spirits or wearing protective charms (yeeb yaj, protective amulets/ritual objects) — courage-language alone, unanchored, risks being heard as self-generated bravado or (worse) as implicitly assuming ritual protection is in place.	Structural fencing already correctly identified in the glossary: this phrase must never appear without the immediately co-occurring presence-clause (below). Recommend enforcing this as an atomic qualifier pair analogous to the Romans TM’s hwjchim → Vajtswv hwjchim rule.
Vajtswv nrog nraim koj (the LORD is with you)	NEW, registry #43	Low weakness in isolation — clear, warm, relationally direct idiom. Risk is entirely structural: if separated from #42 by more than a sentence or paraphrased loosely in summary/quiz contexts, the courage-presence link (the doctrine’s entire theological point) is lost.	Enforce co-occurrence within the same sentence or immediately adjacent clause, mirroring the Romans TM’s kev cawm dim + Yexus Khetos adjacency rule (§2 of the Romans requirements).
ntshai thiab hwm Vajtswv (fear the LORD)	NEW, registry #33	Related but distinct concept — courage and reverent fear are not opposites in Joshua (1:9//4:24) but two sides of covenant relationship; risk is that hearers conflate “fear the LORD” with “fear of the enemy” (which courage is meant to overcome) or with ntshai dab (anxious placation of unpredictable spirits).	Keep visually/lexically distinct from any rendering of “fear of the enemy/battle” in the same lesson; always pair ntshai in this sense with hwm (honor/reverence) to mark reverential rather than anxious fear.

Doctrinal note: No lexical collision rises to Critical here — Hmong has perfectly serviceable vocabulary for both courage and divine presence. The entire risk is structural/relational (whether the mandatory pairing is preserved), not referential (whether the words themselves point to the wrong concept). Recommend Medium-High risk, driven by enforcement discipline rather than word choice.

1.4 Rahab and the Inclusion of the Nations by Faith

Available term(s)	Source	Weakness	Recommended strategy
Vajtswv uas nyob saum ntuj thiab hauv av (God in heaven and on earth)	NEW, registry #10	No traditional-religion collision — the confession is unusually clean doctrinally. The weakness is purely rhetorical/pedagogical: without explicit framing, learners may hear Rahab’s confession as incidental praise from a minor character rather than recognize it as a theologically precise, salvation-historically decisive Gentile faith-confession preceding Cornelius/the Ethiopian eunuch/Ruth as a canonical type.	Teaching note must explicitly name this as evidence for Gentile inclusion by faith, cross-referenced forward to the Romans TM’s nations_gentiles entry and the “obedience of faith” doctrine, to prevent it from being read as merely local color.
kev ntseeg / ntseeg (faith/believe)	REUSED, Romans TM faith_believe	Already flagged Medium in Romans for generic-belief drift; Rahab’s case sharpens the risk because her “belief” is expressed through action (hiding the spies, giving the sign) rather than a verbal creed alone, and Hmong hearers may therefore separate “what she did” from “what she believed” as two different things rather than one integrated faith-response.	Anchor explicitly to her spoken confession (2:11) AND her action (2:4-6, 15-16) as a single unified faith-response, following the Romans TM’s mandatory object-anchoring rule; use this passage to reinforce (not complicate) the existing Romans faith/works integration teaching.
dab mlom (idol)	REUSED, Romans TM idol	Stable High-risk term reused correctly here in contrastive rather than descriptive function — Rahab’s own Canaanite background is implicitly dab mlom-adjacent, which is precisely the rhetorical point (faith transcends prior idol-worshiping background).	No new fencing needed beyond the existing Romans mandate; but recommend an explicit teaching note stating that Rahab is not praised despite her background but that her faith is credited regardless of it — reinforcing the grace-not-merit doctrine already Critical in Romans.
kev hlub ruaj khov (chesed)	NEW, registry #11	Genuine three-way crowding risk against two already-fixed Romans terms: txiaj ntsim (grace, unmerited favor to the guilty) and txoj kev khuv leej (mercy, compassion toward the miserable). Chesed is neither — it is oath-bound covenant loyalty, closer to faithfulness-in-relationship than either grace or mercy.	Do not substitute for either fixed Romans term under any circumstance; teach as a third, distinct category using the spies’ oath (2:12,14) and its fulfillment (6:22-25) as the illustrating narrative, explicitly naming the distinction from txiaj ntsim and txoj kev khuv leej in the same lesson to prevent flattening.
lus twv cog tseg (oath)	NEW, registry #12	Correctly distinguished from lus cog tseg (promise) by the added twv, but the distinction is subtle (one added syllable/word) and easily lost in oral delivery or careless paraphrase.	Recommend the distinction be reinforced with an explicit unilateral-vs-bilateral teaching note at first occurrence (2:12) rather than relying on the lexical marker alone to carry the theological weight.

Doctrinal note: This doctrine has the best available vocabulary of the seven — no forced or invented terms, no Critical lexical collisions. Its risk is entirely pedagogical: under-teaching the theological weight of a minor, ethnically-and-morally-marginal character’s confession. Recommend High (not Critical) risk, driven by teaching-emphasis need rather than translation difficulty.

1.5 Covenant Renewal at Shechem

Available term(s)	Source	Weakness	Recommended strategy
kev cog lus (ntawm Vajtswv thiab nws Haiv Neeg) (covenant)	NEW, registry #16	No prior Romans TM entry exists for berith as a distinct concept from lus cog tseg (promise) — this is a genuine vocabulary gap, not a collision. The risk is that translators/curriculum-writers, lacking a fixed prior baseline, may casually substitute lus cog tseg for kev cog lus, erasing the unilateral/bilateral distinction the whole doctrine depends on.	Establish as a new permanent TM-candidate entry (pending review) with the qualifier structurally mandatory, exactly as Txoj Kevcai requires its qualifier; log for TM-revision review as Critical given its centrality to the core passage (Josh 24:14-15) and its role model for any future OT covenant-renewal curriculum.
Txoj Kevcai (ntawm Vajtswv / Mauxes) (Law of Moses)	REUSED, Romans TM law_mosaic_divine	Stable; already Critical in Romans for the kevcai/ancestral-custom collision. Joshua reuses it in a covenant-renewal ceremony context (24:25-26) that is structurally closer to the very “kevcai” (ritual custom) domain the qualifier is designed to fence against — the ceremonial setting (public assembly, oath, monument-stone) resembles traditional Hmong communal ritual observance more than Romans’ epistolary contexts do.	Reuse unmodified, but flag ch. 24’s ceremonial setting for heightened vigilance: the qualifier is even more essential here than in Romans because the surrounding narrative furniture (assembly, oath, stone) is genuinely ritual in form, raising the odds of an unqualified reading collapsing “the Law” into “a covenant ritual like our own.”
kevcai thiab kev txiav txim (statutes and ordinances)	NEW, registry #19	Deliberately echoes the Kevcai root for continuity, which is helpful for cohesion but marginally increases the ambient kevcai-collision density in ch. 24 (now three separate Kevcai-rooted terms co-occurring: Txoj Kevcai, Kevcai Hla Dhau [if cross-referenced from ch. 5], and this term).	Recommend explicit lesson-level signposting distinguishing all Kevcai-rooted terms present in a given ch. 24 lesson from each other and from generic kev cai dab qhuas, rather than assuming the individual qualifiers alone will prevent conflation when several appear together.
pob zeb ua timkhawv (witness stone)	NEW, registry #20; extends REUSED witness	Low weakness — but note the functional resonance with Hmong ancestral memorial-stone/spirit-marker practices at burial or ritual sites, which the term itself does not name but which some hearers may associate.	Low-priority teaching note sufficient (native-speaker review, not theologian-level): clarify the stone functions as legal/textual witness (a recording device against future covenant violation), not as a spirit-marker or site of ongoing veneration.

Doctrinal note: This doctrine’s central risk is the absence of a prior baseline distinguishing covenant from promise — a missing-vocabulary problem inherited structurally from Romans (which never needed to make this distinction, since Romans’ berith-equivalent contexts were handled generically). Recommend Critical risk tier and priority TM-revision review, since this gap will recur in every future OT curriculum this pipeline produces.

1.6 Obedience and the Conquest

Available term(s)	Source	Weakness	Recommended strategy
mloog lus (obey, base verb, cf. Romans TM obedience_of_faith root)	REUSED root, Romans TM obedience_of_faith	Romans’ fixed compound kev mloog lus ntawm kev ntseeg (obedience OF FAITH) is a specific genitive construction tied to gospel-proclamation contexts (Rom 1:5, 16:26); Joshua’s obedience-vocabulary is mostly command-compliance in a military/covenantal setting (1:7-8, 16-18; 22:2-3), not the same theological category.	Do NOT reuse the fixed Romans compound unmodified for Joshua’s obedience-to-command contexts; use the simpler mloog Vajtswv lus / mloog Vajtswv lus tag nrho lub siab (cf. registry #37, Caleb) and reserve the full Romans genitive compound only for contexts genuinely paralleling gospel-response obedience (arguably ch. 24’s covenant decision, which is the closest Joshua analogue).
mloog Vajtswv lus tag nrho lub siab (wholly followed the LORD)	NEW, registry #37	Positive exemplar term (Caleb); minimal weakness, but risks being heard as an achievable meritorious standard (i.e., “obedience earns land/inheritance,” cf. Caleb’s reward) rather than covenant faithfulness within an already-gracious framework (the land itself was promised, not earned).	Pair with explicit teaching distinguishing Caleb’s reward for faithfulness within the covenant from any suggestion that obedience is the ground of the land-gift itself (which is grace/promise-based per doctrine 1.1 above).
Vajtswv tau ua rog pab cov Yixalayees (the LORD fought for Israel)	NEW, registry #5; cross-referenced from 1.2	Same weakness as under Holy War: obedience-and-conquest narratives (esp. ch. 6, 10) risk collapsing into a false equation “if we obey the ritual instructions [march, shout, blow trumpets], God will act” — a formula uncomfortably close to shamanic ritual-technique efficacy.	Explicit teaching note required: Israel’s specific obedience (e.g., Jericho’s march, Ai’s ambush) was obedience to particular divine command in a particular redemptive-historical moment, not a discoverable ritual technique for producing divine intervention on demand.
khoj/tej uas raug muab cais... (cherem) cross-referenced from 1.2	NEW, registry #6	Same as 1.2 — obedience-and-conquest and holy-war-and-judgment doctrines share their highest-risk term.	Cross-reference rather than duplicate teaching notes; ensure both doctrines’ curriculum units point to the single canonical cherem teaching note rather than generating two independent (and potentially divergent) explanations.

Doctrinal note: The chief risk here is conflation with the ritual-technique pattern already flagged Critical for Holy War — obedience-and-conquest is functionally a subset of the same collision risk viewed from the human-response side rather than the divine-action side. Recommend Critical risk tier for consistency with 1.2, and recommend the two doctrines share a single fencing strategy rather than developing independently (see §4 ranked risks below).

1.7 Rest in the Promised Land as a Type of Salvation Rest

Available term(s)	Source	Weakness	Recommended strategy
kev so / thaj av uas tau kev so (rest)	NEW, registry #23	Lexically clean (no traditional-religion collision) but doctrinally the easiest term in the book to under-teach — kev so is an entirely ordinary, low-register Hmong word for everyday physical rest/respite from labor or war, carrying zero inherent typological weight. Without deliberate teaching, “rest” will default to “peace after war,” a true but doctrinally thin reading that misses the Hebrews 3-4 typological trajectory toward eschatological salvation-rest.	Every occurrence must be paired with an explicit typological teaching note connecting this rest, by anticipation, to ultimate salvation-rest — but must NOT claim direct lexical equivalence to the REUSED Romans TM eternal_life entry; the connection is typological/canonical, not translational. Recommend a dedicated recurring sidebar/teaching-box format used identically at every occurrence (1:1 unit consistency, per Romans-style register rules).
lub nroog cawm siav (cities of refuge)	NEW, registry #24; built on REUSED salvation_save root	Deliberately built on cawm (save/rescue) to create a strong typological bridge, but this creates a structural tension with the Romans TM’s mandatory rule that kev cawm dim/cawm must always be anchored to Yexus Khetos as sole agent — cities of refuge are an OT legal-institutional mechanism, not Christ, and anchoring this occurrence to Christ would be anachronistic, while failing to anchor it risks the term being heard as a free-floating, unqualified “salvation,” diluting the mandatory-anchor rule’s force elsewhere in the curriculum.	Recommend an explicit exception-flag structurally documented (not silently applied): every occurrence of cawm/kev cawm dim in ch. 20 must carry an “OT TYPE ONLY — do not anchor to Yexus Khetos” marker distinct from the Romans TM’s ordinary mandatory-anchor rule, so future curriculum-writers do not mistakenly treat the absence of a Christ-anchor as a violation of the Romans baseline.
tsis tau tso tawm/ntes tag (land not yet fully possessed)	NEW, registry #54; cross-referenced from 1.1	Directly relevant here: “rest” is proclaimed (11:23; 21:43-45) even while some land remains unconquered (13:1,13) — the tension between “rest given” and “rest not yet total” is the linguistic seed of the doctrine’s typological reading.	Use this term explicitly as the textual anchor for teaching “rest” as already-inaugurated-but-not-yet-consummated, rather than treating rest and incomplete possession as an unexplained contradiction in the text.
qub txeeg qub teg (inheritance) cross-referenced from 1.1	REUSED root	Rest and inheritance are conceptually fused in Joshua (land = rest = inheritance); risk of doctrinal blur is low but recommend explicit sequencing in teaching materials (inheritance granted → land occupied → rest experienced) so the terms are not treated as fully synonymous.	Sequential teaching note recommended at Medium priority.

Doctrinal note: This doctrine’s risk profile is inverted relative to most others in this book: the words themselves are safe (no traditional-religion collision), but the doctrine is High risk precisely because safe, ordinary words are easy to leave untaught. Recommend High risk tier, driven by comprehension/typological-teaching-emphasis need, consistent with the Romans-baseline pattern already used for hope/kev cia siab.

---

2. Missing Vocabulary vs. Crowded Semantic Neighborhoods

2.1 Missing Vocabulary (no ready Hmong equivalent at any register)

These are concepts for which Hmong possesses no existing lexeme or fixed idiom, requiring either a newly-built periphrastic compound or a structured borrowing. They are distinguished from §2.2 below in that the risk is comprehension/construction, not collision with an existing competing meaning.

Concept	Gap description	Resolution adopted
Cherem (devoted-to-destruction ban)	No Hmong category names “an object/population irrevocably set apart to total destruction as an act of divine judgment/consecration,” distinct from both sacrifice and ordinary warfare/plunder.	Built periphrastic compound (registry #6); flagged Critical; requires the heaviest teaching-note apparatus in the curriculum (see §1.2, §4).
Berith as bilateral, oath-sealed covenant (distinct from unilateral promise)	Hmong lus cog tseg names a solemn pledge but does not encode the bilateral, mutually-binding, ratified-by-oath character of berith; no prior Romans-context need for this distinction ever arose.	Built new compound kev cog lus (ntawm Vajtswv thiab nws Haiv Neeg) (registry #16); flagged Critical for TM-revision priority — this gap will recur in any future OT curriculum.
Chesed as covenant-loyalty-kindness (distinct from both grace and mercy)	Hmong has words for kindness, loyalty, and compassion separately, but no single fixed idiom capturing the specific “steadfast love that persists because of a prior bond/oath” sense chesed carries.	Built compound kev hlub ruaj khov (registry #11); flagged Medium but recommend elevation to High given the doctrine-4 centrality of Rahab’s chesed-request/receipt narrative.
The “Captain of the LORD’s host” theophanic military figure	No existing Hmong Christian vocabulary addresses a divine/angelic military commander distinct from God the Father and appearing bodily; nearest available cultural category (spirit-general) is precisely the wrong one.	Built compound tus Thawj Coj ntawm Vajtswv Cov Tub Rog (registry #4); flagged Critical; genuinely novel construction with no fallback idiom.
Cities-of-refuge legal-institutional asylum mechanism	No equivalent Hmong customary-law institution for state-designated asylum cities protecting an accidental killer from kin-vengeance pending judicial process (Hmong clan-vengeance custom has no formal asylum-city analogue, only negotiated clan settlement/blood-price).	Built compound lub nroog cawm siav (registry #24) leaning on the cawm root for pedagogical bridge-building, accepting the resulting tension documented in §1.7 above.

2.2 Crowded Semantic Neighborhoods (existing Hmong word, multiple competing referents, requires fencing)

These are cases where a serviceable Hmong word or root already exists but is already claimed — fully or partially — by a competing traditional-religion referent, an existing REUSED Romans TM sense, or another NEW Joshua sense. Each requires explicit fencing (qualifiers, teaching notes, or deliberate avoidance of a tempting-but-wrong shortcut).

Hmong root/word	Competing occupants	Fencing strategy
kevcai (custom/rite root)	(a) REUSED Txoj Kevcai (Mosaic Law); (b) kev cai dab qhuas (ancestral spirit-worship customs, forbidden collision already documented in Romans); (c) NEW kevcai thiab kev txiav txim (statutes/ordinances, 24:25); (d) NEW Kev Cai Hla Dhau (Passover, ch. 5); (e) NEW kev txiav (raws li Txoj Kevcai Mauxes) (circumcision, REUSED). Joshua concentrates five kevcai-rooted terms across chs. 1, 5, 8, 23, 24 — a higher density than anywhere in Romans.	Recommend a book-level “kevcai family” sidebar distinguishing all five occupants at the first ch. 5 cluster-point (circumcision + Passover fall in the same narrative unit) and reinforcing it at the ch. 24 cluster-point (Law + statutes/ordinances). Never allow any of the five to appear unqualified.
hwjchim (power)	REUSED Romans TM power, already Critical, mandatorily qualified Vajtswv hwjchim. Joshua does not introduce a new sense but raises the stakes: the conquest narratives are the most concrete, visible demonstrations of Vajtswv hwjchim anywhere in Scripture (sun standing still, walls falling, hailstones), making the shamanic-power collision risk more vivid and more tempting to leave implicit (“obviously God is powerful here, no need to state it”) than in Romans’ more abstract doxological contexts.	Reuse Romans qualifier rule unmodified but flag every conquest miracle-narrative occurrence (6:20; 10:11-14) for explicit qualifier enforcement, since narrative vividness increases (not decreases) the temptation to drop the qualifier for stylistic flow.
pehawm (worship/serve, cultic)	REUSED Romans TM worship_serve_cultic, already High, mandatorily object-anchored. Joshua 24:14-15 is textually denser in this verb (four occurrences in two verses) than any single Romans passage, and it directly juxtaposes pehawm Vajtswv against pehawm [other gods] in adjacent clauses — the highest-density pehawm/pehawm-dab contrast anywhere in this Language Package.	Recommend register-level: every one of the four 24:14-15 occurrences must independently carry its object, with no elliptical shortcuts (“serve Him… and them” collapsing to a single anchored verb) — this is the passage’s entire rhetorical structure and cannot be economized for smoother Hmong prose.
xaiv (choose)	REUSED Critical-tier Romans formula kev xaiv (Vajtswv xaiv raws li nws lub siab nyiam) = God’s sovereign election of persons; NEW Joshua sense xaiv seb koj yuav pehawm leej twg = human covenantal choice of whom to serve (24:15). These are theologically distinct categories (divine sovereign election vs. human responsible choice) that happen to share the same base verb.	Registry #35 already builds the human-choice sense around the serve verb specifically to avoid the fixed election-formula shape; recommend this be treated as a mandatory structural rule (never abbreviate 24:15’s “choose” to bare kev xaiv) rather than a stylistic preference, given the Critical status of the underlying Romans formula.
dab (spirit)	REUSED Romans TM caution (dab as false/lesser spirit, god entry’s rejected-alternative note); NEW Joshua sense lwm tus dab / lwm tus vajtswv (other gods, 24:14-15, registry #31) deliberately leans into this collision as the passage’s evangelistic point, rather than avoiding it.	Unlike every other dab-adjacent caution in this Language Package (which exists to prevent collision), this is the one place collision is intentional and necessary — the passage requires hearers to recognize their own ancestral dab-service in “the gods your fathers served.” Recommend explicit documentation (already present in registry #31) that this entry is the sole deliberate exception to the general dab-avoidance pattern, so future reviewers do not “correct” it into a softened rendering.
av (land/earth)	NEW Joshua sense thaj av uas Vajtswv cog lus tseg (Promised Land, doctrine 1.1); competing resonance with ancestral-land veneration and land-blessing ritual (hu plig av, land-spirit propitiation before farming/building).	See §1.1 above; never allow bare thaj av for the Promised Land referent.
fij (sacrifice/offering)	REUSED Romans TM caution (living_sacrifice, propitiation_atoning_sacrifice, both already fenced against shamanic appeasement); Joshua’s lub thaj (altar, registry #27) and cherem (registry #6) both sit adjacent to this domain without directly using the fij lexeme, but risk bleed-through if teaching materials casually describe cherem-destruction or altar-building using fij vocabulary for narrative color.	Explicit prohibition recommended: never describe cherem (judicial destruction) using fij vocabulary (offering/sacrifice), and always anchor lub thaj with the “rau Vajtswv” qualifier exactly as the registry already specifies, to prevent three adjacent-but-distinct domains (offering, altar, ban-destruction) from blurring into one generalized “ritual” impression.

---

3. Transliteration vs. Paraphrase Decisions

Joshua introduces substantially more proper-noun and place-name vocabulary than Romans (an epistle with comparatively few named locations) required. The following decisions govern the whole book and should be treated as binding precedent for any future Old Testament narrative curriculum in this pipeline.

Item	Decision	Rationale
Personal names (Rahab, Achan, Caleb, etc.)	Transliterate, following existing Hmong Bible phonological conventions (e.g., /r/→/l/: Rachav → Lahaj).	Consistent with the Romans-baseline pattern for Yexus Khetos, Povlauj; proper names are transliterated, never paraphrased or replaced with descriptive epithets, to preserve genealogical/historical continuity (cf. Matthew 1, Hebrews 11 citing Rahab by name).
Ethnic/tribal group names (Amorites, Levites, etc.)	Transliterate, following the established cov Yudas / cov Kilis pattern.	Direct precedent already set in the Romans TM; extending the same convention avoids introducing a second naming pattern within one Language Package.
Place names with no ongoing theological function (Jericho, Ai, Gilgal, Shechem)	Transliterate the name itself; paraphrase/gloss any theologically-loaded wordplay embedded in the name (e.g., Valley of Achor/“Trouble,” registry #55) in a footnote rather than attempting to reproduce the pun in Hmong.	Preserves geographic/historical concreteness while still surfacing authorial wordplay for teaching purposes without forcing an artificial Hmong pun.
Technical covenant/legal terms with no single-word Hmong equivalent (berith, cherem, chesed)	Paraphrase (build a periphrastic Hmong compound), never transliterate.	Consistent with the Romans-baseline rejection of transliterated loanwords for theologically load-bearing terms (hilastērion, baptism) on the grounds that an opaque loanword would prevent the necessary contrast-teaching; the same logic applies with even greater force to Hebrew technical terms, which have no existing foothold in Hmong Christian usage at all (unlike Greek NT terms, which at least benefit from decades of prior Hmong Bible translation tradition).
Theophanic/quasi-angelic figures (Captain of the LORD’s host)	Paraphrase (descriptive compound), never transliterate a Hebrew title and never borrow existing Hmong spirit-military vocabulary.	A transliterated Hebrew military title would be doctrinally inert (comprehension failure) while borrowing existing spirit-general vocabulary would produce doctrinal collision; the periphrastic descriptive compound is the only option that threads both hazards, at the cost of requiring sustained teaching-note support.
”Abba”-style retained loanword precedent (from Romans)	Not extended to Joshua. No Hebrew term in Joshua carries the same untranslatable-intimacy rationale that justified retaining “Abba” as a loanword in Romans 8.	Recommend against inventing a Hebrew-loanword precedent in this book merely because one exists for Aramaic “Abba” in Romans; each loanword decision must be independently justified, not applied as a general rule.

---

4. Ranked List of Highest-Risk Ambiguities

Ranked by combined severity (doctrinal centrality × collision specificity × difficulty of fencing), consistent with the risk-tier definitions inherited from the Romans baseline.

1. Cherem / devoted-to-destruction ban (registry #6; doctrines: Holy War, Obedience-and-Conquest). Highest risk in the book: a wholly new construction, with no prior Hmong Christian usage to lean on, addressing the single most morally and theologically difficult concept in Joshua, with live collision risk against ritual curse-transfer and spirit-appeasement destruction customs. Critical.
2. Berith as bilateral covenant, distinct from unilateral promise (registry #16; doctrine: Covenant Renewal at Shechem). A genuine vocabulary gap inherited structurally from the Romans baseline (which never needed this distinction); the core passage (Josh 24:14-15) depends on the covenant/promise distinction being clearly maintained, and no existing fixed idiom protects against collapse. Critical.
3. “Other gods” / lwm tus dab as direct, intentional collision with ancestral/household spirit-service (registry #31; doctrine: Covenant Renewal at Shechem / Obedience). Uniquely, this is a deliberate collision rather than one to be fenced against — mishandling it in either direction (softening it into “idols” generically, or failing to make the ancestral-spirit-service parallel explicit) neutralizes the core passage’s entire evangelistic force. Critical, with an inverted fencing logic from every other entry on this list.
4. Vajtswv tau ua rog pab cov Yixalayees / the LORD fought for Israel, and its entanglement with obedience-triggers ritual-technique risk (registry #5; doctrines: Holy War, Obedience-and-Conquest). Risk of producing an implicit “obey the ritual instructions and God will act” formula indistinguishable in surface grammar from shamanic ritual-efficacy patterns. Critical.
5. Captain of the LORD's host (registry #4; doctrine: Courage through God’s Presence, secondarily Holy War). Genuinely novel theophanic vocabulary with a specific, named collision (spirit-general invocation) and no fallback idiom; contained to a single chapter (5) but doctrinally weighty (widely read as Christophanic). Critical.
6. Txoj Kevcai reused in the ch. 24 ceremonial-ritual setting (registry #17; doctrine: Covenant Renewal at Shechem). Not a new collision, but a context where the existing Romans-Critical caution is under heightened pressure because the surrounding narrative furniture (assembly, oath, monument) is itself genuinely ritual in form. High, escalating toward Critical in ch. 24 specifically.
7. thaj av / Promised Land and ancestral land-veneration resonance (registry #21; doctrine: God’s Faithfulness to the Land Promise). Strong positive bridge that is simultaneously a syncretism risk; unlike most entries above, the danger here is a positive resonance being over-extended rather than a negative one being confused. High.
8. pehawm density in Joshua 24:14-15 (registry #34; doctrine: Covenant Renewal at Shechem). Not a new collision (inherited Critical/High from Romans) but the highest-density concentration of the pehawm/pehawm-dab contrast anywhere in the Language Package, raising the stakes on qualifier discipline. High.
9. kev so / Rest, under-teaching risk (registry #23; doctrine: Rest in the Promised Land as a Type of Salvation Rest). The book’s clearest example of a safe word carrying unsafe theological thinness — no collision, but high risk of the typological doctrine simply evaporating if the word is treated as ordinary vocabulary needing no comment. High.
10. lub nroog cawm siav / Cities of refuge and the cawm Christ-anchor exception (registry #24; doctrine: Rest in the Promised Land as a Type of Salvation Rest). Structural tension with the Romans TM’s mandatory Christ-anchoring rule for cawm; risk is procedural (curriculum-writers misapplying or over-applying the Romans rule) rather than referential. Medium-High.
11. kevcai-family density across chs. 1, 5, 8, 23, 24 (multiple registry entries). No single occurrence is high-risk in isolation, but the cumulative density is unmatched in Romans and creates cumulative fencing fatigue risk (qualifiers dropped for stylistic economy after repeated use). Medium, escalating with volume.
12. kev hlub ruaj khov / chesed vs. grace vs. mercy three-way distinction (registry #11; doctrine: Rahab and the Inclusion of the Nations by Faith). A genuine, newly-surfaced three-way semantic neighborhood (grace/mercy/chesed) with no precedent requiring this distinction in Romans. Medium, recommend elevation to High given centrality to doctrine 4.
13. xaiv human-choice vs. divine-election formula adjacency (registry #35; doctrine: Covenant Renewal at Shechem). Well-fenced already by design (built around “serve” rather than a bare election-shaped noun), but the fencing depends entirely on consistent enforcement rather than inherent lexical safety. Medium.
14. lus twv cog tseg vs. lus cog tseg (oath vs. promise), one-syllable distinction (registry #12; doctrine: Rahab and the Inclusion of the Nations by Faith). Low collision risk but genuine risk of the distinction being lost in oral delivery or careless summary paraphrase given how minimal the surface marker is. Medium.

---

5. Summary Recommendations for Phase 1 Step 9 (Doctrine Risk Registry) and Phase 2 Enforcement

1. Cherem, berith, and the “other gods” collision (items 1-3 above) should anchor the Critical tier of the forthcoming Joshua doctrine risk registry, each requiring mandatory theologian review with no automated-review bypass, consistent with the Romans-baseline routing rule.
2. The courage/presence pairing and the pehawm object-anchor rule should be encoded as structural, atomic-unit enforcement rules in the Joshua Phase 2 requirements document, exactly mirroring the Romans TM’s hwjchim→Vajtswv hwjchim and kev cawm dim→+Yexus Khetos enforcement pattern.
3. A single canonical cherem teaching note should be authored once and cross-referenced by both the Holy War and Obedience-and-Conquest doctrine units, rather than independently drafted twice, to prevent doctrinal drift between the two doctrines’ treatments of the same term.
4. The cities-of-refuge Christ-anchor exception must be explicitly and separately documented in the Phase 2 requirements as a structural exception to the Romans cawm-anchoring rule, not left for curriculum-writers to infer.
5. New TM-candidate entries requiring priority formal review: kev cog lus (ntawm Vajtswv thiab nws Haiv Neeg) (covenant), khoom/tej uas raug muab cais rau Vajtswv kom puas tsuaj tag nrho (cherem), tus Thawj Coj ntawm Vajtswv Cov Tub Rog (Captain of the LORD’s host), and kev hlub ruaj khov (chesed) — these four carry the highest downstream risk if left as provisional glossary entries rather than ratified, permanent TM entries before Phase 2 generation begins.