Amos — Linguistic Gap Analysis (Hmong Language Package, Phase 1, Step 8)

Purpose. This document performs the PRD Phase 1 Step 8 linguistic gap analysis for the Amos curriculum, built on analysis/08_core_glossary.md and assets/bible_term_registry.json (Amos Phase 1 artifacts). It (a) matrixes each assigned doctrine against the Hmong vocabulary presently available to express it, naming specific weaknesses; (b) separates genuine missing-vocabulary gaps (comprehension problems) from crowded-semantic-neighborhood problems (collision/fencing problems) that require disambiguating strategy rather than new words; (c) records transliteration-vs-paraphrase decisions for untranslatable source material; and (d) ranks the letter’s highest-risk ambiguities for downstream Phase 2 enforcement. Nothing here contradicts the Romans baseline (assets/translation_memory.json); all baseline-fixed strings are treated as immovable and reused, per 12_ai_translation_requirements.md §2 rule 1.

---

0. Method and Governing Constraints

• Every term already fixed in the Romans baseline TM (Vajtswv, tus Tswv, txiaj ntsim, hwjchim, kev cawm dim, kev xaiv, Vaj Ntsuj Plig, mercy, hope, creation, nations_gentiles, boasting, etc.) is inherited, not re-analyzed for collision risk here; only its extension into new Amos contexts is assessed.
• Amos introduces one register-shifting fact the Romans package did not need to solve: the Tetragrammaton (YHWH) appears independently of Adonai/kyrios roughly 80 times, forcing a three-way divine-name system (Vajtswv / tus Tswv / Yawmsaub) that Romans, an epistle without OT divine-name citation density, never required in this proportion. This is this book’s single largest net-new structural gap.
• Amos is poetry-heavy covenant-lawsuit oracle, not epistolary argument; several gaps below are genre-driven (funeral dirge, woe-oracle, hymn fragment, wordplay) rather than doctrine-driven, and are noted as such.
• “Fencing” = documented disambiguating strategy (qualifier phrase, mandatory co-text, teaching note) applied to an existing, unavoidable Hmong word that already carries a competing traditional-religion sense. “Gap” = a genuine missing referent requiring explanatory gloss, not collision management.

---

1. Doctrine Vocabulary Matrix

1.1 Social Justice and Care for the Poor

Element	Content
Available Hmong terms	kev txiav txim ncaj (mishpat, justice), cov pluag / cov txom nyem (poor/needy), tsim txom / tsuj yuam (oppress/crush), Vajtswv txoj kev ncaj ncees (tzedakah, reused Romans string), kev txiav txim ncaj ntawm rooj loog (justice-at-the-gate), tej teev nqes cuav (false scales)
Weaknesses	(1) kev txiav txim ncaj risks being heard as abstract judicial correctness rather than Amos’s concrete demand — real courts, real weights, real wages — unless institutional co-text (“ntawm rooj loog,” at the gate) is retained. (2) cov pluag / cov txom nyem sits inside a crowded semantic neighborhood: Hmong traditional cosmology explains differential prosperity largely through hmoov (impersonal ancestor/spirit-assigned fate/fortune), which naturalizes poverty as deserved misfortune rather than the product of addressable human sin. (3) Reusing the exact Romans string for tzedakah risks silently importing the forensic/imputed sense from Justification-by-Faith into a text where the word means covenant-faithful social conduct toward neighbor — the opposite direction of confusion from the Romans baseline’s own concern (there, the risk was moralism creeping into imputation; here, the risk is a forensic reading flattening social ethics).
Recommended strategy	Fence, do not invent. Keep kev txiav txim ncaj permanently paired with concrete social-institutional co-text (courts, gate, wages, weights) at every occurrence in Amos content, distinct from any bare doctrinal abstraction. Extend the baseline’s hmoov prohibition (already Critical for grace/election in Romans) explicitly to poverty-causation language: never attribute cov pluag’s condition to fate/fortune vocabulary; always name a human oppressor-agent (tsim txom, tsuj yuam, muag...rau nyiaj me me). Attach a mandatory bidirectional disambiguation note wherever the shared ncaj ncees string appears in Amos content, explicitly stating: “In Loos (Romans), this phrase names a legal status given by faith; in Amos, it names right conduct done toward the poor. These are related but not identical; do not substitute one teaching for the other.”

1.2 Judgment on Empty Religious Ritual

Element	Content
Available Hmong terms	khoom fij kub tag / khoom fij qoob loo / khoom fij sib raug zoo (burnt/grain/peace offerings), tej hnub kevcai noj haus (feasts), tej rooj sib txoos dawb huv (solemn assemblies), kuv ntxub, kuv tsis lees yuav (I hate, I despise), suab nrov ntawm nej tej nkauj (mere noise of songs)
Weaknesses	(1) The entire offering vocabulary (fij) sits in the single most crowded semantic neighborhood in the whole Amos-Romans combined library: fij is simultaneously (a) Christ’s unique, sufficient atonement (qhov theej txhoj kev txhaum, Romans Critical), (b) ethical living-sacrifice devotion (khoom fij uas muaj txoj sia nyob, Romans High), (c) shamanic spirit/ancestor appeasement (fij dab, folk practice), and now (d) technically-correct-but-relationally-rejected Israelite cultic offerings (Amos). A hearer’s default diagnostic instinct for “an offering that didn’t work” is folk-religious: assume ritual error and try again with a better offering. Amos’s point is the opposite — the ritual was flawless, and God rejects it anyway because it is offered by unjust hands. This is a genuinely novel collision angle not present in the Romans baseline, which must be fenced separately from both existing fij fencing strategies. (2) tej hnub kevcai noj haus reuses the kevcai root already flagged for Law/baptism in Romans; the new risk here is the opposite polarity — not “biblical law collapsed into ancestral custom” but “God rejects festivals as such,” when in fact Amos rejects festivals-divorced-from-justice, not festival-observance itself.
Recommended strategy	Fence with a directional contrast note distinct from the Romans fij-fencing notes already established: every Amos lesson touching D3-D5 (burnt/grain/peace offerings) must state explicitly that the offerings were ritually correct but relationally worthless, not ritually flawed, to preempt the shamanic “try a different/better offering” instinct. Cross-reference, but do not merge, with Romans’ living_sacrifice and propitiation fencing language, since those solve a different confusion (Christ’s offering assimilated into repeatable appeasement) than this one (correct offering treated as insufficient due to injustice, not technique). For tej hnub kevcai noj haus, mandatory co-teaching of 5:24 in the same unit as 5:21-23, so festival-rejection is never presented without its justice-restoration resolution.

1.3 God’s Sovereignty over All Nations

Element	Content
Available Hmong terms	Yawmsaub uas kav tag nrho tej tub rog ntuj (LORD of hosts), lwm Haiv Neeg (nations, inherited), kev fav xeeb (pesha, transgression/rebellion), transliterated nation names, the 9:7 exodus-of-nations and Cushites clauses (descriptive, no fixed short form)
Weaknesses	(1) “LORD of hosts” risks assimilation into Hmong shamanic spirit-army mobilization (txiv neeb ritual warfare, in which a shaman negotiates with and commands spirit forces for a client) — a genuine crowded-neighborhood collision, since both concepts involve a deity/spirit-figure commanding an army of unseen beings. (2) kev fav xeeb (treason/rebellion) is a genuine vocabulary gain over the Romans baseline’s kev txhaum (general sin), but risks being under-used or treated as a mere synonym rather than the stronger, sovereignty-specific term it needs to be, if translators default to the more familiar baseline word. (3) The 9:7 material (Cushites-leveling, exodus-of-nations claim) has no fixed short-form rendering at all — it is currently descriptive/discursive only, which is appropriate for its one-time rhetorical force but creates inconsistency risk if different lessons paraphrase it differently.
Recommended strategy	Fence “LORD of hosts” with mandatory teaching contrast: YHWH’s hosts are unilaterally, sovereignly commanded — not a negotiable ritual resource invocable by a religious specialist on a client’s behalf. This is a new fencing pattern (unilateral command vs. negotiated invocation) distinct from the baseline’s hwjchim/governing-authority fencing (delegated vs. supernatural), and should be logged as its own contrast category. Reserve kev fav xeeb exclusively for covenant-treason contexts (ch.1-2, ch.3’s pesha-adjacent material) and never let it drift into interchangeability with kev txhaum without an explicit note that pesha is the stronger term. For the 9:7 material, adopt one settled descriptive rendering per clause (already logged in the registry) and require verbatim reuse across all curriculum units rather than ad hoc re-paraphrase — promote to a fixed string at next TM-revision review.

1.4 The Day of the Lord as Judgment

Element	Content
Available Hmong terms	Hnub ntawm Yawmsaub (Day of the LORD), Hnub kev puas tsuaj (day of calamity), kev tsaus ntuj / kev pom kev, kev tsaus nti / kev ci ntsa iab (darkness/gloom pairs), kuv yuav tsis dhau lawm (finality formula), txoj hlua ncaj (plumb line)
Weaknesses	(1) Hnub ntawm Yawmsaub is a genuinely double-hazard term: a flattened rendering loses Amos’s shocking reversal (expected national vindication becomes judgment on Israel itself, 5:18-20), and independently risks conflation with unrelated vocabulary from historical Hmong millenarian/messianic movements (folk end-times “new king” prophetic traditions), which could cause the term to register as generic apocalyptic folklore rather than YHWH’s specific covenant-lawsuit verdict. This second hazard has no parallel anywhere in the Romans baseline and is unique to Amos/OT-prophetic material. (2) The finality formula (“I will not pass by them again”) must be fenced against the same appeasement-reversal logic flagged for offerings (§1.2): a hearer accustomed to negotiable spirit-anger may assume a sufficiently large or correct future offering could still reverse the verdict, when the text insists it cannot.
Recommended strategy	Fence Hnub ntawm Yawmsaub with a mandatory reversal-teaching note at every occurrence (not merely first occurrence) given the doctrine’s Critical tier: state explicitly what Israel expected (vindication over enemies) versus what the text delivers (judgment on Israel). Separately and explicitly disambiguate from Hmong millenarian-movement end-times vocabulary — this requires a dedicated cultural-context footnote in Phase 2 lesson templates, flagged here for [TRANSLATOR REVIEW NEEDED: potential conflation with Hmong messianic-movement end-times folklore] pending native-speaker/theologian review. Pair the finality formula with explicit “no ritual reverses this” language drawn from the same fencing logic as §1.2.

1.5 The Restoration of David’s Fallen Tent (Messianic Hope)

Element	Content
Available Hmong terms	lub tsev pheeb suab ntawm Daviv uas vau lawm (fallen booth/tent of David), kuv yuav kho dua tshiab, thooj li thaum ub (rebuild as of old), cov seem ntawm Edoos thiab tag nrho lwm Haiv Neeg (remnant + nations), kuv yuav cog lawv...tsis muab lawv dob dua li lawm (planting permanence), tej roob yuav muaj cawv txiv hmab ntws saum roob (wine-stream abundance)
Weaknesses	(1) The sukkah (temporary hut/booth) vs. bayit (grand house/dynasty) distinction is lexically fragile in Hmong: lub tsev pheeb suab (temporary field-hut, an authentic Hmong agrarian-lifecycle referent) is available and accurate, but nothing prevents a translator or reader from defaulting to a generic lub tsev (house) gloss that erases the deliberate humiliation-imagery of the Hebrew, since “restoration” language psychologically pulls toward “grand rebuilding” rather than “modest hut repaired.” (2) The doctrine’s climactic promise has zero standalone vocabulary problem but a significant cross-reference-management problem: its theological payload only lands correctly if actively cross-linked to baseline-fixed Romans terms (son_of_god, heir, kingdom_of_god) and to the Romans 9-11 Future-of-Israel doctrinal thread (all_israel_will_be_saved, grafted_olive_tree) — omitting these links leaves 9:11-15 as a bare, disconnected political-restoration oracle rather than the Acts 15:16-17 messianic anchor it is.
Recommended strategy	Lock lub tsev pheeb suab (never a generic lub tsev) as the fixed rendering for sukkah in this context, with mandatory footnote on first occurrence in every module explaining the deliberate hut/palace contrast. Require every Phase 2 lesson unit touching G1-G5 to include explicit cross-reference callouts to the four named baseline Romans entries — this is a structural/pedagogical fencing requirement, not a lexical one, and should be encoded as a template requirement, not left to per-lesson discretion.

---

2. Missing Vocabulary vs. Crowded Semantic Neighborhoods

2.1 Genuine Missing-Vocabulary Gaps (comprehension problem — explanatory gloss sufficient, no collision)

These require a footnote/teaching aid but carry no risk of doctrinal inversion if left unglossed (though comprehension would suffer):

Term	Gap type	Resolution
Plumb line (anakh)	Unfamiliar construction tool	Illustrative note; already logged with descriptive phrase txoj hlua ncaj (ntsuas kev ncaj)
Nazirite (nazir)	No native consecration-vow analogue	Explanatory gloss modeled on baseline’s circumcision pattern
Olive-adjacent: Bashan cattle, ivory houses, horns of the altar	Unfamiliar geography/architecture/luxury materials	Footnote-level; comprehension-only, per registry entries H4, H5, D8
Solemn assemblies (atzarot)	No exact liturgical-calendar equivalent	Descriptive phrase sufficient (tej rooj sib txoos dawb huv)
Harps/lyres (nevalim)	Unfamiliar instrument	Descriptive phrase + optional illustrative note
Sieve, summer-fruit basket, planting/streams imagery	Agrarian imagery	No gap at all — these map with unusual ease onto a Hmong highland rice/orchard farming context and should be treated as a translation asset, not a risk

Conclusion: Amos has comparatively few pure comprehension gaps relative to collision risks, because its dominant imagery (agriculture, herding, weights/measures, courts, family/clan inheritance) overlaps strongly with traditional Hmong material culture. The gap profile of this book is collision-dominant, not vocabulary-scarcity-dominant — the opposite emphasis from, e.g., a book set in a purely urban Greco-Roman institutional world.

2.2 Crowded Semantic Neighborhoods Requiring Fencing (collision problem — lexical choice is correct/unavoidable, disambiguating strategy is mandatory)

These are the load-bearing risk zones of the whole Amos package. Each is a case where the correct, unavoidable Hmong word already carries a live traditional-religion sense that must be actively fenced off, not replaced:

Crowded neighborhood	Competing occupants	Fencing strategy required
Divine-name space (Vajtswv / tus Tswv / Yawmsaub)	Generic deity term, generic lordship title, and covenant proper name must now coexist at high density (~80x for YHWH alone) in a single book, a proportion the Romans baseline never tested	Absolute non-interchangeability rule; Yawmsaub reserved exclusively for YHWH, never substituted with tus Tswv or Vajtswv even for stylistic variation (cf. §5, Rank #1 below)
fij (sacrificial-offering) domain	Christ’s atonement (Romans Critical) / living-sacrifice ethics (Romans High) / shamanic spirit-appeasement (fij dab) / technically-correct-but-rejected Israelite cultic offerings (Amos, new)	Four-way fence; Amos’s contribution is the new “correct-ritual-rejected-anyway” diagnostic problem, requiring a distinct teaching note from the other three uses (see §1.2)
hwjchim (power/host-command) domain	Shamanic ritual power (Romans High) / delegated civil authority (Romans High) / YHWH’s unilateral cosmic army-command (Amos, new)	Extend baseline’s mandatory-qualifier rule; Amos’s “LORD of hosts” sense must carry its own distinguishing note (unilateral command vs. negotiated invocation), not simply reuse the “Vajtswv hwjchim” qualifier verbatim without the hosts-specific contrast
kevcai (custom/rite) root	Mosaic Law (Romans Critical) / baptism (Romans High) / ancestral spirit-worship customs (kev cai dab qhuas) / Israelite festival calendar (Amos, new)	Amos’s polarity is inverted from Romans’: the danger is not conflating biblical practice with ancestral custom, but concluding ritual observance itself (rather than injustice) is what God condemns; requires its own note distinct from the baseline’s Law-qualifier strategy
hmoov (fate/fortune) — permanently forbidden zone	Grace/election (Romans Critical, permanently forbidden) / poverty-causation (Amos, newly extended prohibition)	Extend the existing permanent prohibition by explicit analogy rather than treat as a separate rule; log as a formal TM-revision extension, not a new independent prohibition, to keep the rule unified across both curricula
dab / spirit-invocation domain	Idol (Romans High, inherited) / illicit cult-site oath-swearing (Amos, new: “your god, O Dan,” “the way of Beersheba”)	Oath-formula usage is structurally identical (not merely analogous) to Hmong ancestor/spirit oath-invocation customs; requires explicit “this is the same structure, condemned” teaching rather than a softer comparative note
ncaj ncees (righteousness) string	Forensic imputation (Romans Critical) / OT covenant-faithful social conduct toward neighbor (Amos, new — different direction of risk)	Bidirectional bad-transfer risk in both directions (see §1.1); requires a note usable in either curriculum’s lesson templates
Millenarian/end-times folk vocabulary space	Hmong 19th-20th century messianic-movement end-times folklore / biblical “Day of the LORD”	Newly identified in this analysis; not previously flagged in the Romans baseline since Romans has no comparable “the Day” apocalyptic-popular-expectation motif of this density; requires dedicated [TRANSLATOR REVIEW NEEDED] flag pending native-speaker consultation on which specific folk-vocabulary items pose the sharpest risk

Key finding: Every crowded neighborhood in Amos is an extension of an already-fenced Romans neighborhood, with one exception (millenarian/end-times folklore, newly surfaced). This means Phase 2 generation can mostly reuse existing fencing infrastructure (qualifier-phrase enforcement, mandatory co-text, teaching notes) but must add book-specific contrast notes at each extension point, since the direction or nature of the risk sometimes inverts (e.g., ncaj ncees, kevcai, fij) rather than simply repeating.

---

3. Transliteration vs. Paraphrase Decisions

Item	Decision	Rationale
YHWH (Tetragrammaton)	Established-tradition proper name (Yawmsaub), not a new transliteration	Follows existing Hmong Bible translation tradition; inventing a phonetic transliteration of the Hebrew consonants would break with the reading community’s existing Scripture and create a false sense of a “different god” from the OT they already know
Foreign nation names (Damascus, Gaza, Tyre, Edom, Ammon, Moab, Philistia, Aram, Cush)	Transliteration, following the baseline’s cov Yudas/cov Kilis convention	These are historical proper nouns with no semantic content to paraphrase; transliteration is the only faithful option and matches established pattern
Beit-El / Beth-Aven wordplay (“house of God” → “house of nothing”)	Transliterate the place name (Npe-Es) + mandatory footnote, do NOT attempt a Hmong pun	No Hmong phonetic or semantic construction reproduces the Hebrew pun; forcing an invented Hmong wordplay would risk sounding contrived or, worse, theologically distorted if the invented pun drifted in meaning. Footnote-based explanation is the safer, honest option, consistent with the baseline’s treatment of untranslatable material (e.g., Abba, Leej Txiv retention pattern)
Qayits/qets pun (“summer fruit” / “the end”)	Paraphrase literally + mandatory footnote, do NOT attempt a Hmong pun	Same rationale as above; unlike Beit-El (a proper name that can at least be transliterated), this pun has no proper-noun anchor at all, making literal paraphrase plus explanation the only faithful path
Nazirite (nazir)	Transliterate + explanatory gloss (tus Naxi (uas khi lus fij rau Vajtswv))	Modeled directly on the baseline’s successful circumcision pattern: transliteration preserves the proper-noun/technical-term feel while the parenthetical gloss supplies the missing cultural content
Sukkah (“booth/tent of David”)	Paraphrase using an authentic Hmong referent (lub tsev pheeb suab, temporary field-hut), NOT a transliteration	Unlike Nazirite, a real Hmong referent exists that captures the “humble, temporary structure” sense accurately; transliterating would forfeit this valuable cultural bridge for no gain
”LORD of hosts”	Paraphrase (Yawmsaub uas kav tag nrho tej tub rog ntuj), NOT a transliteration of “Sabaoth”	A Sabaoth-style loanword (as used in some English liturgical traditions) would be opaque to Hmong readers and would forfeit the needed contrast-teaching opportunity against shamanic spirit-army mobilization; the descriptive paraphrase makes the contrast teachable
Cows of Bashan, ivory houses	Paraphrase with transliterated place-name retained (cov nyuj Npasaas)	Balances comprehension (concrete animal/luxury imagery, paraphrased) against loss of geographic specificity (place-name retained via transliteration)

General rule surfaced by this analysis: transliterate proper nouns (names of people, nations, places, technical Hebrew office-titles like Nazirite) that have no semantic content to lose; paraphrase everything else, especially wherever the paraphrase itself creates a teachable contrast against a competing traditional-religion concept (LORD of hosts, sukkah). Never invent a Hmong wordplay to solve an untranslatable Hebrew pun — flag and footnote instead, per the two [TRANSLATOR REVIEW NEEDED] items already logged in the registry (Beit-El/Beth-Aven; qayits/qets).

---

4. Cross-Reference Requirements Surfaced (Structural Gaps, Not Lexical)

Several items in this analysis are not vocabulary gaps at all but mandatory cross-reference/template requirements that Phase 2 generation must enforce structurally:

1. Every Amos 5:24 (kev txiav txim ncaj / Vajtswv txoj kev ncaj ncees) lesson must carry the bidirectional Romans/Amos disambiguation note (§1.1).
2. Every Amos 9:11-15 (Davidic-tent restoration) lesson must cross-link the four named baseline Romans entries plus the Future-of-Israel doctrinal thread (§1.5).
3. Every Amos 5:21-23 (rejected offerings) lesson must carry the “correct-but-rejected” contrast note distinct from Romans’ fij-fencing notes (§1.2, §2.2).
4. Every Amos 3:14, 4:4-5, 5:4-6 (Bethel/Gilgal/Beersheba cult-site) lesson must retain the explicit “seek the LORD, not these sites” polemical contrast (registry #48).

---

5. Ranked List of Highest-Risk Ambiguities

Ranked by combined severity (doctrinal centrality × collision specificity × occurrence density):

1. Divine-name interchange (Yawmsaub vs. tus Tswv vs. Vajtswv). Highest risk by sheer density (~80+ occurrences of YHWH alone) combined with total inversion potential: any silent substitution collapses the covenant-specific, proper-name force of the entire prophetic-lawsuit structure into generic deity-talk. No equivalent density risk exists anywhere in the Romans baseline. Escalation: human theologian + native speaker, mandatory pre-publication audit of every divine-name occurrence.

2. Rejected-but-technically-correct offerings (D3-D5, fij domain). Highest doctrinal-inversion risk: if hearers apply the shamanic “ritual error → try again” diagnostic, they will misread God’s decisive rejection as fixable through better ritual performance, precisely inverting 5:21-24’s argument that only justice (not more or better ritual) can resolve God’s rejection. Escalation: human theologian review; this is the doctrinal engine of the core passage itself.

3. Hnub ntawm Yawmsaub (Day of the LORD) conflated with Hmong millenarian-movement end-times folklore. Newly surfaced in this analysis; no prior fencing exists anywhere in the baseline. Risk is severe because it could cause the entire Day-of-the-Lord doctrine to be assimilated into folk-apocalyptic expectation rather than heard as a specific covenant-lawsuit verdict. Escalation: [TRANSLATOR REVIEW NEEDED: possible collision], native-speaker consultation required before any Phase 2 content on this doctrine is finalized.

4. hmoov (fate) applied to poverty-causation (C2/#17). Direct extension of an already-Critical baseline prohibition into new territory; risk is that Amos’s entire social-justice argument (poverty as the fruit of specific, addressable human injustice) collapses into fatalistic acceptance if any content slips into fate-vocabulary framing, even inadvertently through idiom choice. Escalation: human theologian review, treat as an extension of the existing permanently-forbidden rule.

5. “LORD of hosts” assimilated into shamanic spirit-army mobilization. Specific, nameable collision (unilateral divine command vs. negotiated ritual invocation of spirit-forces) with no precedent fencing strategy in the Romans baseline (the closest analogue, hwjchim, solves a related but distinct problem). Escalation: human theologian review for the new contrast-note wording before first use.

6. Oath-swearing by illicit cult-site deities (8:14) structurally identical to Hmong ancestor/spirit oath customs. Distinguished from #5 by being a structural identity claim rather than a loose-analogy claim — the practice condemned is the same kind of act Hmong hearers may themselves practice, raising the pastoral stakes considerably. Escalation: human theologian review.

7. Bidirectional ncaj ncees (righteousness) transfer between Romans’ forensic sense and Amos’s covenant-conduct sense. Real risk but more containable than #1-6 because the fencing mechanism (a disambiguation note) is already well-understood from the Romans baseline’s own analogous cautions; risk is under-application (translators forgetting to add the note) rather than an unsolved collision. Escalation: native-speaker review at minimum, theologian if any lesson merges the two senses without comment.

8. Beit-El/Beth-Aven and qayits/qets wordplays. Lower doctrinal stakes than #1-7 (these are literary devices, not load-bearing theological vocabulary), but flagged because silent omission of the mandatory footnote would produce a comprehension failure at a rhetorically climactic moment (8:1-2’s vision, and the repeated Bethel irony). Escalation: native-speaker review sufficient; footnote presence is a checklist item, not a doctrinal-review item.

9. kevcai festival-rejection polarity inversion (D1, “God hates ritual” misreading). Real but manageable risk, since the corrective (mandatory pairing with 5:24) is a straightforward template rule rather than an unsolved lexical problem. Escalation: native-speaker review, escalate to theologian only if a lesson unit omits the 5:24 pairing.

10. Sukkah-vs-bayit flattening (G1, “fallen tent” read as “fallen house”). Real risk of losing deliberate humility-imagery, but the correct Hmong term (lub tsev pheeb suab) already exists and is culturally authentic; the risk is translator/writer default drift toward a generic lub tsev, which is a discipline/consistency problem more than an unsolved linguistic gap. Escalation: native-speaker review with a hard-lock on the fixed string.

---

6. Governing Note

This analysis is a Phase 1 artifact feeding directly into the Amos Phase 2 AI Translation Requirements document (parallel in function to the Romans baseline’s 12_ai_translation_requirements.md). All fencing strategies, transliteration decisions, and ranked risks above must be encoded as enforceable rules in that downstream document, following the same structural pattern the Romans baseline established (mandatory qualifier phrases, forbidden-substitution tables, escalation-routing tables). No recommendation here overrides any Romans-baseline-fixed string; every extension proposed is additive and requires the same formal TM-revision review process specified in 12_ai_translation_requirements.md §2 rule 4 and §9 before any [NEW] item in 08_core_glossary.md or assets/bible_term_registry.json (Amos) may be promoted to permanent translation-memory status.