Executive Summary — Mark → Pashto Language Package

Why it matters

Mark is the first full Gospel-narrative curriculum built on this Pashto Language Package, extending a Romans baseline that was epistolary and forensic into a genre that is narrative, oral, and action-driven. The core passage, Mark 10:35-45, sits at the fusion point of five of this curriculum’s eight doctrines (Servanthood, Suffering Servant/Son of God, Necessity of the Cross, Jesus’ Authority, Ransom for Many) and lands directly on this readership’s two deepest fault lines: Islamic tawhid/crucifixion-denial theology, and Pashtunwali’s honor-shame, servant-is-shameful social default. Getting this book right matters more, not less, than Romans did, because Mark’s climax (10:45’s ransom saying) is the single highest-doctrinal-density verse assessed anywhere in this pipeline so far.

Key findings

• 98 enforced terms now live in the Mark translation_memory.json: 32 inherited exactly from the Romans baseline, 66 newly minted for Mark’s own narrative vocabulary (Son of Man, ransom, unclean spirit, “I am,” necessity-of-suffering, cross, Gehenna, blood of the covenant, cry of dereliction, and the full supporting cast).
• 20 Critical-risk terms (9 inherited + 11 new) and 35 High-risk terms (12 inherited + 23 new) — 55 of 98 terms (56%) — require absolute glossary enforcement and theologian-level scrutiny at every occurrence.
• 31 doctrines are tracked in the Mark doctrine risk registry: 13 Critical, 13 High, 5 Medium, 0 Low — an unusually top-heavy risk profile compared to Romans, driven by Mark’s narrative embedding of deity/sonship claims (transfiguration, “I am,” forgiving sins) and its historical-event claims (crucifixion, resurrection) that Romans’ abstract vocabulary never had to carry.
• 26 of 31 doctrines require mandatory human theologian review; only 5 require native-speaker-only review, and zero are automated-review-only — every doctrine in this book touches either a Quranic collision or a Pashtunwali cultural default.
• Mark introduces two genuinely new fencing rules with no Romans-baseline precedent: (1) واک (authority) must never share vocabulary with پر مخکې واکمني کول (worldly domination, 10:42), or Christ’s unique authority collapses into the abusive pattern he condemns; (2) ناپاک روح (unclean spirit) must never drop its qualifier, or it collapses toward روح القدس (Holy Spirit).

Risks

• Ransom (10:45, فدیه): the single highest-stakes lexical decision in the book. Islamic فدیه defaults to a lesser compensatory payment; Pashtunwali بدل (explicitly forbidden as the ransom-noun) would import revenge-cycle logic into a text whose logic is gracious substitution. Mistranslation here corrupts the curriculum’s own doctrinal anchor.
• Crucifixion and resurrection (ch. 15-16): direct, sustained collision with Qur’an 4:157’s denial that Jesus was killed or crucified — the most consequential shared risk with the surrounding Islamicate culture, requiring oral/eyewitness apologetic framing, not lexical caution alone.
• Son of Man (11 occurrences): the literal Pashto calque risks silently confirming, rather than correcting, the Quranic “merely human prophet” reading unless the Daniel 7 authority-content is actively taught at every major occurrence.
• Servanthood/slave vocabulary (9:33-37; 10:42-45): direct clash with Pashtunwali’s honor code (nang); “slave of all” (10:44) has no ready cultural or Islamic-devotional parallel and risks being flattened into generic humility rather than taught as a deliberate inversion.
• Gehenna and the cry of dereliction: both risk importing entire pre-existing theological frameworks (Quranic multi-level hell cosmology; an actual Trinitarian rupture) that the Pashto term or literal rendering does not, on its own, prevent.

Opportunities

• Nanawatai (ritual submission accepted in place of retaliation) remains, as in Romans, a genuine and now doubly-important cultural bridge for the Ransom for Many doctrine — Mark 10:45 and 14:24 both benefit from its calibrated, explicit use.
• Legion’s Roman-military-occupation resonance (5:9) lands with real force for a population with its own generational experience of foreign occupation — an asset for teaching decisive spiritual liberation.
• Sola (peace) and honor-shame mockery/vindication dynamics (15:16-32) continue to be assets, not merely risks, for a readership fluent in both conflict-resolution vocabulary and honor-shame social reasoning.
• Mark’s oral, narrative genre is itself an opportunity: this is the most audio/oral-delivery-ready book in the pipeline so far for a readership with lower average literacy and strong safety incentives to favor oral transmission.

Recommended actions

1. Route Mark 10:35-45 (the entire core passage), the Last Supper (14:22-24), Gethsemane (14:32-42), and all of Mark 15 to human theologian review as complete units, not merely term-by-term — per the escalation rules in analysis/12_ai_translation_requirements.md.
2. Lock the Son of Man, Ransom, and Authority/Domination fencing rules before batch-processing any chapter, since these three decisions have the widest downstream footprint across the book (11, 2, and 2+ occurrences respectively, cross-referenced repeatedly).
3. Pair every crucifixion and resurrection occurrence with named-eyewitness narrative material rather than relying on doctrinal assertion alone, consistent with this readership’s oral-culture orientation and literacy profile.
4. Sequence Phase 2 batch processing so that chapters 1-9 (which stabilize Son of Man, authority, servant/slave, and unclean-spirit vocabulary) are locked and version-confirmed before chapters 10, 14, and 15 (the highest Critical-density chapters) are processed.
5. Carry forward the Romans baseline’s forbidden-substitution list unchanged and add the six new Mark-specific forbidden substitutions (جن, کفر, روایت, بدل-as-noun, unqualified جهنم, softened ego eimi) to every reviewer’s checklist from day one of Phase 2.