Linguistic Gap Analysis — Book of Mark (Hmong)

0. Purpose and Method

This document performs the Phase 1 linguistic gap analysis mandated by the PRD for the Hmong Mark Language Package. It does not re-derive terms already fixed in the Romans baseline (assets/translation_memory.json) or already registered in analysis/08_core_glossary.md and assets/bible_term_registry.json (Mark). Instead it performs three analytical tasks the prior documents assumed but did not fully execute:

1. Build a doctrine vocabulary matrix — for each of the eight assigned Mark doctrines, what Hmong vocabulary is actually available, where it is weak, and what strategy Phase 2 generation must follow.
2. Map crowded semantic neighborhoods (multiple distinct source-concepts competing for the same Hmong root) that require lexical fencing, and separately catalogue true vocabulary gaps (no native referent at all).
3. Resolve, term-by-term, the transliteration vs. paraphrase question for every proper noun and technical term surfaced in Mark, and produce a ranked list of the letter’s highest-risk ambiguities for theologian review priority.

All renderings below reuse assets/translation_memory.json (Romans) and the Mark-specific entries already established in analysis/08_core_glossary.md / assets/bible_term_registry.json verbatim. No new Hmong string is coined here; this document only analyzes, ranks, and fences what already exists. Per the governing baseline, every doctrine below carries at least Critical or High tier, so every finding here routes to mandatory human theologian review before publication.

Citation note: No prior Hmong Language Package document fixed a destination-language book name for Mark. Following the same transliteration logic that produced “Loos” for Romans, and consistent with existing Hmong Bible translation tradition (Vaj Lug Kub Npaiv Npaum), the working citation form is Malakaus [chapter]:[verse] (e.g., Malakaus 10:45). This is flagged [TRANSLATOR REVIEW NEEDED: confirm "Malakaus" against the live HWB/YouVersion book-name convention before Phase 2 citation-format lock], parallel to the Romans package’s own pending re-verification note for its YouVersion Version ID.

---

1. Doctrine Vocabulary Matrix

For each doctrine: the available Hmong vocabulary (drawing on Sections A/B of the core glossary and the term registry), the specific weakness(es) in that vocabulary, and the recommended Phase 2 generation strategy.

1.1 Jesus as the Suffering Servant and Son of God — Critical

Available terms	Weakness	Recommended strategy
Leej Tub ntawm Vajtswv (Son of God, A7)	None lexically; risk is severance — hearers may take this honorific in isolation from the cross, since Hmong honorific titles (Leej) normally signal unqualified status/glory, not suffering.	Never present Leej Tub ntawm Vajtswv without proximate co-text anchoring to the Passion (esp. 15:39, spoken by the centurion at the cross). Mandatory pairing rule: this title’s climactic confession must always cite Malakaus 15:39 alongside any earlier occurrence (1:11, 3:11, 5:7, 9:7).
tus Tub ntawm Tibneeg (Son of Man, B1)	Deliberately humble classifier (tus, not Leej) risks the opposite failure: collapsing the title into “just a human being,” losing the Daniel 7 authority/glory pole.	Every occurrence requires the two-pole teaching note (humility + Dan. 7 authority) established in the core glossary; never gloss as a bare synonym for “Jesus.”
Khetos (standalone Christ title, B22)	Lexically stable (shares the transliteration root of Yexus Khetos), but its narrative hinge at 8:29 is immediately followed by the first Passion prediction and Peter’s rebuke — the doctrinal risk is a hearer taking “Christ” as a triumphalist title severed from suffering.	Teach 8:29 forward to 10:45 and 14:61-62 as a single arc; never let a lesson isolate the confession from the suffering it provokes.
Ntoo Khaub Lig (Cross, B24)	Fixed rendering, but shares no root with any Hmong funerary/ritual-wood vocabulary, so the term is lexically safe but requires positive construction of meaning from zero — a comprehension gap, not a collision, but a large one given the doctrine’s centrality.	Mandatory first-use explanatory note in every module introducing suffering-Messiah content; must be distinguished explicitly from ancestral funerary wood-structures per B24’s note.
nqi txhiv / theej cov neeg coob (Ransom / substitution, B2/B3)	Grounds the doctrine’s climax (10:45) but depends on two flagged Critical terms operating correctly together — see Doctrine 8 below for full treatment.	Cross-reference Doctrine 8’s matrix row; do not duplicate teaching content, but always co-teach Suffering-Servant doctrine with the Ransom saying as its capstone text.

Net assessment: No missing vocabulary; the risk is entirely compositional — the four terms above must always be taught as a single, indivisible arc (glory-title → humble-title → suffering-confirmation → atoning-death), never any one in isolation.

1.2 The Kingdom of God Breaking In — High

Available terms	Weakness	Recommended strategy
lub Nceeg Vaj ntawm Vajtswv (Kingdom of God, A13)	Hmong Nceeg Vaj denotes a static, bounded royal territory/dynasty (a real historical Hmong kingdom concept), which naturally produces a nationalistic, one-time-enthronement expectation.	Never use without contextual signal of the already/not-yet tension; must be reinforced by B10 at every early-ministry occurrence (ch. 1, 4).
twb los ze zuj zog lawm (kingdom at hand / ēngiken, B10)	The Greek perfect tense’s “already begun, presently effective” nuance has no single Hmong grammatical equivalent (Hmong verbs are not tense-marked this way); the phrase is a workable periphrasis but is easily flattened in casual retelling to a simple future (“the kingdom is coming”).	Lock this exact periphrasis as fixed collocation, never substitute a simple future construction; explicitly teach the already/not-yet tension every time the Kingdom is introduced, tying forward to James and John’s static-kingdom misunderstanding at 10:37.

Net assessment: This is fundamentally a grammatical gap (no native perfect-tense equivalent), not a religious-collision risk — but the consequence of under-translating the grammar is a doctrinal distortion (over-eager political/nationalistic kingdom expectation), which is why the doctrine is tiered High despite the underlying issue being linguistic rather than religious-collision-based.

1.3 Faith and Discipleship amid Fear — High (per core glossary; escalates in core-passage proximity)

Available terms	Weakness	Recommended strategy
kev ntseeg / ntseeg (Faith, A12)	Broad generic-belief verb; must always be object-anchored to Christ (inherited constraint from Romans).	No new weakness beyond the inherited one; reinforce anchoring at every Markan miracle-faith pericope.
kev ntshai (Fear, B14)	Stable lexically, but Hmong daily life is structurally organized around fear of malevolent spirits and the ritual vigilance that manages it (offerings, taboos, calling back souls). Simply telling a hearer “do not fear” without naming the replacement source of security risks being heard as “abandon protective vigilance with nothing to replace it,” which is dangerous advice inside the traditional worldview.	Every “do not fear” saying must be paired with an explicit statement of what security replaces the abandoned ritual vigilance — namely, trust in Jesus’ demonstrated authority (B5/#59) over the very storms, spirits, and afflictions that generate fear. This pairing is not optional flavor text; it is the doctrine’s load-bearing safety mechanism.
koj txoj kev ntseeg tau cawm koj / kho koj zoo (faith-that-saves/heals, B16)	Deliberate Greek double-sense (heal/save) has no single Hmong verb carrying both; the chosen compound phrase is workable but risks a translator silently resolving to only one sense out of a desire for clarity.	Explicit prohibition (already logged) against resolving the dual sense; teaching note must present both readings together at every occurrence (5:34, 10:52).

Net assessment: No missing vocabulary; the central risk is pastoral-doctrinal, not lexical — an under-taught “fear” term could functionally leave converts feeling spiritually unprotected rather than differently protected.

1.4 The Necessity of the Cross — Critical

Available terms	Weakness	Recommended strategy
Ntoo Khaub Lig (Cross, B24)	See 1.1 above; additionally, the metaphorical extension “take up your cross” (8:34) requires the hearer to grasp that this is a daily, costly, self-denying posture modeled on (but categorically distinct from) Christ’s unique atoning death. Collapsing the two senses (call-to-discipleship cross vs. Christ’s own once-for-all cross) is a real risk given they share the identical Hmong noun.	Fixed disambiguation rule: any occurrence of “take up the/his cross” (8:34) must carry a co-text marker (e.g., “raws li Yexus tau ua” — “after the pattern Jesus set”) distinguishing imitative self-denial from the unrepeatable atoning act at 15:24-37.
tso nws tus kheej tseg / txoj sia (deny self / psychē, B25)	Psychē here means one’s vital self/whole-life-orientation, not one’s spirit (ntsuj plig, A6/#6); a translator instinct to reach for the more “spiritual-sounding” ntsuj plig must be resisted.	Maintain the fixed lexical separation already established in the core glossary; txoj sia for psychē is locked, ntsuj plig is reserved exclusively for anthropological pneuma.
lub khob (ntawm kev txom nyem) (Cup of suffering, #47)	Metaphorical cup-as-destiny has no native Hmong idiom; comprehension gap requiring a first-use gloss, but crucially must remain lexically identical at both 10:38-39 and 14:36 to preserve Mark’s deliberate echo between James/John’s request and Jesus’ own Gethsemane submission.	Lock a single fixed phrase across both occurrences; flag any drafting variant as a continuity violation, not a stylistic choice.

Net assessment: The doctrine’s chief risk is intra-lexical collision — the same Hmong noun (Ntoo Khaub Lig) must bear two theologically distinct weights (imitative call, unique atonement) within the same sixteen chapters, and the cup-metaphor’s echo effect can be silently lost through uncoordinated drafting across chapters 10 and 14.

1.5 Servanthood versus Worldly Greatness — High

Available terms	Weakness	Recommended strategy
qhev (Servant/Slave, doulos, A10)	Inherited High-risk term; here reinforced because 10:44 applies it aspirationally to every disciple’s definition of greatness, not merely as Paul’s self-designation — this intensifies the risk that hearers recoil from an image of literal chattel bondage being held up as the goal of discipleship.	Catechetical note must explicitly state: this is the voluntary, honored self-giving Christ Himself models (v.45), commended as the paradoxical summit of greatness, not a degrading condition to escape.
tus uas pab lwm tus (general servant, diakonos ethical sense, B26)	Must not be confused with the Romans-baseline church-office term tus pab tu pawg ntseeg (Phoebe’s recognized office) — using the office-term here would be anachronistic (no church yet exists in the narrative) and would wrongly formalize what is meant as a universal ethical posture.	Maintain the lexical fence between the two diakonos senses established in B26; flag any drafting reuse of the Romans office-term inside Mark content.
kav lawv nkaus li tus tswv kav nws qhev / swb hwjchim rau lawv (lord it over, B30)	Deliberately reuses the reverent tswv root and the flagged hwjchim root in a condemned, negative example — the single highest internal-consistency risk in this doctrine, since these same roots also build the reverent titles tus Tswv (A2) and the qualified Vajtswv hwjchim (B5).	Every occurrence of this phrase must carry a “this is the rejected worldly pattern” framing marker in the same sentence or immediately adjacent clause; never allow the phrase to stand as a neutral description of “ruling.”

Net assessment: No missing vocabulary; the doctrine’s risk is entirely about negative reuse of positive/sacred roots — Mark deliberately repurposes vocabulary elsewhere reserved for God and Christ to describe the rejected worldly model, and Hmong readers must not let that negative context bleed backward into their reverence for tus Tswv or hwjchim elsewhere.

1.6 The Messianic Secret — Medium-High (literary-structural, escalates near core passage)

Available terms	Weakness	Recommended strategy
txwv tsis pub qhia rau leej twg (silencing command, B23)	Not a lexical collision at all — a purely literary-structural device (progressive revelation strategy) that Hmong has no difficulty expressing, but which risks being mistaken for a comprehension-level vocabulary problem and therefore under-taught as narrative strategy.	Every occurrence requires a narrative-strategy teaching note (why Jesus restricts disclosure at this stage of the story), not a vocabulary note; explicitly connect to 8:29-30 (Peter’s confession immediately silenced) and to James/John’s premature triumphalism in the core passage (10:35-37).
tus Tub ntawm Tibneeg (Son of Man, B1) and Khetos (B22)	Both titles are the very content being “kept secret” for portions of the narrative; the doctrine’s risk is that a translator or teacher resolves the ambiguity too early by over-explaining, flattening the narrative tension Mark deliberately builds.	Sequence teaching notes to match the narrative’s own pacing — do not front-load full Christological explanation before 8:29, even though the vocabulary itself is already fixed and available from ch. 1.

Net assessment: This doctrine has essentially no vocabulary gap — its entire risk is pedagogical sequencing, which is a curriculum-design concern rather than a lexical one, but must still be logged here because mis-sequencing produces the same net effect (doctrinal flattening) as a lexical failure.

1.7 Jesus’ Authority over Sickness, Sin, and Nature — Critical

Available terms	Weakness	Recommended strategy
hwjchim qualified Yexus tus kheej muaj hwjchim (Jesus’ divine authority, B5/#59)	The single most dangerous bare root in the entire Mark package: unqualified, it is lexically identical to a shaman’s (txiv neeb) ritual power to negotiate with spirits. Three distinct qualified senses now compete for the same root across the curriculum: (a) Christ’s own inherent authority (this doctrine), (b) delegated civil authority (Doctrine 10, ch.12-13,15), (c) the condemned “lording over” worldly example (Doctrine 5, ch.10).	Absolute rule: hwjchim NEVER appears bare in generated content. Each of the three senses requires its own fixed qualifying frame, and no lesson may allow two senses to appear in the same paragraph without an explicit differentiating clause.
dab / dab qias (demon/unclean spirit, B6/#56)	Exact indigenous term for the very spirits Hmong shamans ritually negotiate with; unavoidable and correct, but the single highest-stakes contrast-teaching requirement in the whole curriculum.	Every exorcism narrative must explicitly state Jesus expels dab by sovereign word alone — instantly, permanently, without negotiation, payment, or appeasement ritual — directly and repeatedly contrasted with ua neeb shamanic practice. This contrast, not the term itself, is the actual doctrinal content requiring delivery.
mob ruas (leprosy, B7)	Comprehension gap: no Hmong purity-exclusion category attaches to this specific ailment.	Explanatory note on Mosaic ritual-exclusion law; not a lexical fix.
raug zam txhaum (forgiveness of sins, #44)	Distinct from the Romans-baseline forensic-imputation vocabulary (raug suav tias ncaj ncees); risk is a translator conflating “forgiveness” (debt-release) with “justification” (righteous-status-reckoning) since both involve raug + a passive construction.	Maintain lexical separation explicitly; a footnote cross-referencing the Romans baseline’s distinct term is recommended at first occurrence (2:5-10).

Net assessment: This doctrine contains the curriculum’s single highest-density crowded-neighborhood problem (hwjchim serving three theologically incompatible senses) and its single highest-stakes cultural collision (dab/exorcism vs. shamanism). Both require structural, non-negotiable disambiguation rules, not case-by-case judgment calls.

1.8 The Ransom for Many — Critical (anchors the core passage)

Available terms	Weakness	Recommended strategy
nqi txhiv (Ransom, B2/#49)	Built on the Romans-baseline kev txhiv (redemption) root, intensified with nqi (“price”) to isolate the specific price-paid nuance of lytron against the more general apolytrōsis (“redemption” as an act). This is a genuine positive bridge (Hmong debt-redemption custom — buying back a pawned person/item), but the bridge itself is also the danger: Hmong debt-redemption is a repeatable, negotiable, market transaction between parties of relatively equal standing, whereas Christ’s ransom is unilateral, sufficient, and never repeated.	Every occurrence must carry the “once for all, sufficient, non-negotiable” qualifying clause established for the Romans baseline’s hilastērion entry; explicitly forbid any teaching that frames the ransom as a negotiated price a hearer or family could have paid themselves.
theej cov neeg coob (substitution/anti pollōn, B3/#50)	Reuses the theej root from the Romans baseline’s propitiation entry — good internal consistency, but risk is a translator softening theej (substitution) toward a merely benefactive “for the benefit of” sense, losing the strict “instead of” force that makes this a doctrine of vicarious substitution rather than generic benevolence.	Lock the fixed collocation muab nws txoj sia ua nqi txhiv theej cov neeg coob as a single unit; forbid the weaker “rau ntau tus” (merely “for many”) as a substitute, per the existing forbidden-substitution note.
qhev (servant/slave, A10) — doctrinal ground	See Doctrine 5; here it functions as the premise (“whoever would be first must be slave of all”) that the Ransom saying (v.45) supplies as its climactic warrant — Jesus grounds His own ethical instruction in His own coming self-sacrifice.	Teach v.43-44 and v.45 as a single argumentative unit; never excerpt the Ransom saying without its immediately preceding servanthood teaching, or vice versa.
Vajtswv Txoj Kev Cog Lus (Covenant, B4/#40)	Not in the core passage itself (14:24) but doctrinally inseparable from the Ransom, since the Last Supper’s covenant-blood language is the liturgical enactment of the same substitutionary logic the Ransom saying states propositionally. Risk: Hmong blood-oath covenant custom is reciprocal and renewable; Christ’s covenant is unilateral and final.	Cross-teach 10:45 and 14:24 together; the “once for all” qualifier used for nqi txhiv must also govern Vajtswv Txoj Kev Cog Lus.

Net assessment: This is the curriculum’s theological center of gravity. All four terms above must function as a single, tightly fenced semantic cluster (nqi txhiv + theej + qhev + Vajtswv Txoj Kev Cog Lus), each reinforcing the others’ “unilateral, sufficient, once-for-all” qualifier — a failure in any one term’s disambiguation compromises the entire doctrine, since Hmong hearers have ready-made cultural analogues (debt-redemption, blood-oath covenant) for every term in the cluster, each of which is a repeatable/negotiable practice that must be explicitly overridden.

---

2. Semantic Neighborhood Fencing

This section catalogues Hmong lexical items where multiple distinct source-concepts compete for the same root (crowded neighborhoods requiring fencing) and, separately, source-concepts with no native referent at all (true vocabulary gaps).

2.1 Crowded Semantic Neighborhoods (fencing required)

Root / Family	Competing senses	Fencing strategy
hwjchim	(1) shamanic ritual power [forbidden bare use]; (2) Christ’s own inherent divine authority (B5/#59, this doctrine’s anchor); (3) delegated civil authority (#60); (4) the condemned “lording over” worldly example (B30)	Never bare. Each sense requires its fixed qualifying frame (Yexus tus kheej muaj hwjchim; cov muaj hwjchim kav teb chaws; explicit “rejected worldly pattern” marker for B30). No two senses may co-occur in one paragraph without an explicit differentiator. This is the single most overloaded root in the entire Mark package.
ntsuj plig / plig family	(1) Vaj Ntsuj Plig = Holy Spirit [Vaj-prefix mandatory, never dropped]; (2) ntsuj plig (bare) = anthropological human spirit (A6/#6); (3) txoj sia = psychē/vital self (B25/#35, NOT spirit)	Three-way fence, not two-way: Vaj Ntsuj Plig (divine), ntsuj plig (human spirit), txoj sia (life/self) must never be interchanged. The blasphemy-against-the-Spirit passage (3:28-29) and “deny himself… whoever loses his life” (8:35) sit only five chapters apart and use two different members of this family — a single drafting slip could conflate the unforgivable sin with ordinary self-denial.
kevcai root family	(1) Txoj Kevcai (ntawm Vajtswv/Mauxes) = divine Law [qualifier mandatory]; (2) tej kab ke qub ntawm cov txwj laus = tradition of the elders (B18/#38) [deliberately does NOT share the kevcai root]; (3) kev cai raus dej = baptism (A28/#51)	Mark 7:8-9 depends on senses (1) and (2) remaining visibly distinct — this is the opposite disambiguation direction from most of the package (usually the risk is under-qualifying; here the existing strategy of using a different root for tradition is correct and must not be “corrected” toward consistency by a well-meaning drafter reaching for kevcai vocabulary for tradition too). Sense (3) baptism escalates to High specifically at its metaphorical extension in the core passage’s near neighborhood (10:38-39, being “baptized” into suffering) and must not be read as a second literal water-rite.
dawb huv / qias family	(1) dawb huv = sanctification/ritual-and-moral purity (inherited Critical risk from Romans baseline); (2) qias / ua qias = ritual defilement, ch.7’s antithesis (B19/#39)	Mark 7:14-23 is the direct textual resolution of the Romans baseline’s anticipated collision — Jesus Himself relocates defilement from external/ritual to internal/moral. This passage should be flagged as a cross-book teaching asset: use it explicitly when teaching the Romans sanctification doctrine to resolve the dawb huv ritual-purity ambiguity at its source.
tswv / qhev family	(1) tus Tswv = reverent divine/messianic title (A2, Low risk); (2) qhev = honored voluntary bondservant self-designation (A10, High risk); (3) tus tswv kav nws qhev = the condemned worldly-domination pattern (B30)	Sense (3) deliberately weaponizes senses (1)-(2)‘s roots for a negative example; requires the explicit “rejected pattern” marker at every occurrence (see 1.5 above) so reverence for (1)-(2) is not contaminated.
fij (sacrificial-offering) family [inherited from Romans, reactivated in Mark]	(1) Christ’s own atoning death (typologically present throughout Mark’s Passion, though the term itself surfaces more explicitly in Romans/Hebrews-style vocabulary); (2) shamanic spirit-appeasement (fij dab)	Mark itself does not heavily use the fij noun directly (the Ransom/substitution vocabulary of B2/B3/#49/#50 carries the weight instead), but any curriculum cross-references to the Romans baseline’s living_sacrifice or propitiation entries when teaching Mark’s atonement doctrine must preserve the existing Romans-level fence between (1) and (2).
theej root	(1) Romans baseline’s qhov theej txhoj kev txhaum (propitiation); (2) Mark’s theej cov neeg coob (substitution, B3/#50)	Both uses are intentionally consistent (same root, compatible sense) — this is a positive cross-book reinforcement, not a collision, and should be explicitly taught as such: Mark 10:45 uses the same substitutionary logic Romans 3:25 names propositionally.
cawm / kev cawm dim family	(1) general salvation (A15/#16, inherited High risk); (2) physical healing in the dual-sense faith sayings (B16/#17)	Not a true collision but a deliberate Markan wordplay; must be taught as intentional ambiguity, not resolved.

2.2 True Vocabulary Gaps (no native referent, comprehension-level only)

These terms have no competing traditional-religion referent — the risk is purely that the underlying concept is culturally unfamiliar, requiring explanatory scaffolding rather than lexical disambiguation:

• mob ruas (leprosy/ritual exclusion, B7) — no Hmong purity-exclusion disease category.
• chaw kub taws txhaum mus ib txhis (Gehenna, B27/#70) — no native afterlife-judgment topography; must avoid drift toward ancestral-spirit-realm imagery.
• yam qias vuab tsuab uas ua rau chaw pe hawm nyob do cuas (Abomination of desolation, B34/#71) — no Danielic apocalyptic referent exists; requires dedicated prophecy-background teaching.
• Hnub Xanpataus (Sabbath, #41) — transliteration; no native weekly-rest-day institution.
• sab tes xis / sab tes laug (right hand/left hand seats of honor, #63) — partial bridge via Hmong court/clan-elder seating custom, but the specific royal-coregency sense requires a note; low collision risk, moderate comprehension gap.
• tsob ntoo txiv aliv (olive tree, inherited Romans entry — not present in Mark but relevant to any cross-book teaching) — botanically unfamiliar; not applicable to Mark directly but noted for completeness given the shared TM.

Assessment: True gaps are comparatively low-stakes in this book — Mark’s linguistic terrain is dominated by crowded neighborhoods (existing vocabulary doing double or triple duty) far more than by empty ones. This should shape Phase 2 effort allocation: fencing existing roots matters more here than inventing new ones.

---

3. Transliteration vs. Paraphrase Decisions

Term	Decision	Rationale
Talitha koum (5:41)	Transliterate (Talitha khumas) + footnote	Follows established Abba Aramaic-retention precedent; preserves Mark’s own bilingual narrative technique (he retains the Aramaic and then translates it for his Greek audience — the Hmong text should mirror this dual-layer structure, not collapse it).
Eloi Eloi lema sabachthani (15:34)	Transliterate + Mark’s own translation retained	Identical rationale; Mark himself models the retain-then-translate pattern, so Hmong should not “help” by translating only, since that erases Mark’s deliberate literary device highlighting Jesus’ full participation in dereliction.
Hosanna (11:9-10)	Transliterate (Hauxana) + gloss	Established convention; low risk, functions as a liturgical acclamation already familiar in Hmong Christian hymnody patterns.
Abba (14:36)	Transliterate + gloss (already fixed in Romans baseline; this Mark verse is its source text)	No change; Mark 14:36 is definitionally authoritative for this entry.
Beelzebul (3:22)	Transliterate (Npe-exenpub) + gloss	Proper name; paraphrase (e.g., “prince of demons” alone) would lose the specific accusation’s rhetorical target in the divided-kingdom argument that follows.
Legion (5:9)	Transliterate (Legioos) + gloss noting Roman military-unit echo	Proper name chosen by the demoniac himself to convey overwhelming multiplicity; paraphrase would flatten the deliberate irony (Rome’s occupying power vs. Jesus’ superior authority).
Corban (7:11)	Transliterate + explanatory parenthetical (Khaupas (ib yam khoom uas twv cia rau Vajtswv))	A technical legal-vow term with no Hmong equivalent institution; bare transliteration alone would be opaque, so the parenthetical gloss is mandatory, not optional.
Golgotha / place names (15:22, etc.)	Transliterate, standard practice	Not flagged as high-risk; follows ordinary proper-noun transliteration convention already used for Yudas, Kilis, etc. in the Romans baseline.
Son of Man (huios tou anthrōpou)	Paraphrase (tus Tub ntawm Tibneeg), NOT transliteration	A title, not a proper name; transliteration would produce an opaque, unparsable string, whereas the paraphrase preserves the ordinary-humanity register the title requires while remaining flaggable for the Daniel 7 teaching overlay.
Cross (stauros)	Established Hmong Christian compound (Ntoo Khaub Lig), NOT a fresh transliteration	Existing Hmong Bible tradition already fixed this compound; introducing a transliterated loanword here would sever the term from decades of existing catechetical usage, violating the “no sectarian/unregulated-stream vocabulary” rule from the governing requirements document.
Gehenna	Paraphrase/description, NOT transliteration	A bare transliteration (“Kekena” or similar) would be entirely opaque to a Hmong hearer with no access to the term’s Hinnom-Valley background; the descriptive rendering carries at least partial sense while awaiting the mandatory explanatory teaching note.
Ransom (lytron)	Paraphrase built on existing TM root (nqi txhiv), NOT transliteration	A transliterated loanword was explicitly rejected in the core glossary (B2) as opaque and defeating the needed contrast-teaching against shamanic appeasement; the paraphrase strategy allows the necessary “once-for-all” qualifier to be grammatically attached, which a bare loanword would not permit.
Sabbath	Transliterate (Hnub Xanpataus)	No native weekly-rest institution exists to paraphrase against; transliteration is the only viable option and carries no collision risk (contrast with Ransom, where paraphrase was necessary specifically to enable contrast-teaching).

General principle surfaced by this analysis: transliteration is preferred for proper nouns (names of people, places, and Aramaic quotations Mark himself retains) and for institutions with no native referent to paraphrase against (Sabbath); paraphrase — built, wherever possible, on already-fixed TM roots for internal consistency — is preferred for theological technical terms (Ransom, Son of Man, Gehenna) precisely because paraphrase permits the grammatical attachment of mandatory disambiguating qualifiers that a bare loanword would foreclose.

---

4. Ranked List of Highest-Risk Ambiguities

Ranked by (a) doctrinal centrality, (b) proximity to the core passage (Mark 10:35-45), and (c) severity of potential inversion if mishandled. All items below require mandatory human theologian review per the governing routing rule; ranking determines review priority, not whether review is required.

1. Bare hwjchim (any sense) appearing without its qualifying frame — the single highest-risk lexical event possible in this package; an unqualified occurrence instantly and silently converts Christ’s unique authority, delegated civil rule, or condemned worldly domination into indistinguishable shamanic ritual power. Affects Doctrines 1.7, 1.5, and the Doctrine 9-parallel civil-authority material.
2. nqi txhiv / theej cov neeg coob (10:45) taught without the “once-for-all, non-negotiable” qualifier — collapses the Ransom for Many into a negotiated debt-redemption transaction the hearer could imagine paying themselves or a relative paying on their behalf, inverting the doctrine’s unilateral, sufficient character at the letter’s single most theologically dense clause.
3. dab / dab qias exorcism narratives taught without the sovereign-word-vs-ritual-negotiation contrast — without this contrast, Jesus’ exorcisms assimilate into the existing ua neeb shamanic framework as simply a more powerful practitioner, rather than a categorically different mode of authority.
4. Conflation of Vaj Ntsuj Plig, bare ntsuj plig, and txoj sia — a single dropped Vaj prefix or a misapplied ntsuj plig for psychē risks confusing the divine Person with a human soul or a human soul with one’s vital self, with acute danger at the blasphemy-against-the-Spirit passage (3:28-29) where the stakes are the doctrine of the one unforgivable sin.
5. qhev (10:44) taught without the voluntary-honored-bondservice catechetical note — risk of the servanthood doctrine being either rejected outright (as endorsing literal degrading slavery) or, conversely, sentimentalized into generic “helpfulness” that loses Jesus’ deliberately shocking self-application of a slave’s status to Himself and to greatness itself.
6. kav lawv nkaus li tus tswv kav nws qhev / swb hwjchim rau lawv (10:42) appearing without its “rejected worldly pattern” marker — risk of retroactively tainting the reverent titles tus Tswv and qualified Vajtswv hwjchim by proximity to a condemned example built from the same roots.
7. twb los ze zuj zog lawm (1:15) flattened to a simple future — produces the identical nationalistic, immediate-enthronement misunderstanding James and John display at 10:37, undermining the Kingdom-breaking-in doctrine at its very introduction and setting up (rather than defusing) the core passage’s central error.
8. Ntoo Khaub Lig used identically for both “take up your cross” (8:34, imitative) and Christ’s own cross (15:24-37, atoning) without a disambiguating co-text marker — risks either cheapening Christ’s unique atonement into one more instance of costly discipleship, or overclaiming ordinary disciples’ suffering as itself atoning.
9. lub khob (cup of suffering) rendered inconsistently between 10:38-39 and 14:36 — a continuity failure (not a collision) that nonetheless erases Mark’s deliberate echo linking James/John’s naive request to Jesus’ own costly submission, weakening the core passage’s rhetorical force.
10. kev ntshai (“do not fear”) taught without naming the replacement source of security — a pastoral-doctrinal risk rather than lexical collision, but one capable of leaving converts feeling stripped of protective vigilance with nothing substituted, a serious real-world discipleship hazard given the cultural weight of spirit-fear in Hmong life.
11. qias/ua qias (7:14-23) taught as a mere hygiene comprehension gap rather than the direct resolution of the dawb huv collision anticipated in the Romans baseline — a missed teaching opportunity rather than an active distortion, but ranked here because failing to exploit this cross-book asset leaves the Romans sanctification doctrine’s known collision risk unresolved when Mark itself supplies the resolution.
12. raug zam txhaum (2:5-10) conflated with the Romans-baseline forensic-justification vocabulary — a lower-frequency but real risk of blurring debt-release (forgiveness) with forensic status-reckoning (justification), two related but distinct actions.

---

5. Recommendations for Phase 2 TM-Revision and Doctrine Registry Construction

1. Adopt all Section 2.1 fencing rules as structural generation constraints (not stylistic guidance) in the Phase 2 system prompt, modeled on the existing hwjchim/Txoj Kevcai/Vaj Ntsuj Plig qualifier-enforcement pattern already present in 12_ai_translation_requirements.md §2.
2. Elevate item #1 of the ranked list (bare hwjchim) to the same “STOP generation” escalation tier already defined for Critical-tier term substitutions in the governing requirements document §9 — this book multiplies the root’s risk threefold relative to Romans.
3. Confirm the “Malakaus” citation convention (or its correct established alternative) before Phase 2 citation-format lock; log as [TRANSLATOR REVIEW NEEDED] per this document’s header note.
4. Build explicit cross-reference teaching hooks between Mark 7:14-23 and the Romans baseline’s sanctification/dawb huv entry, and between Mark 10:45/14:24 and the Romans baseline’s hilastērion/theej entry, so the two books’ Language Packages reinforce rather than duplicate disambiguation effort.
5. Carry forward all twelve ranked ambiguities into analysis/11_doctrine_analysis.md (Mark) and assets/doctrine_risk_registry.json (Mark) as named translation-risk descriptions, preserving this document’s ranking order as the review-priority sequence for human theologians.