Executive Summary — Gospel of Mark | English → Punjabi

Why it matters

Mark is Punjabi’s first narrative-Gospel curriculum built on this pipeline, and it lands in the most religiously contested vocabulary space of any language in the pipeline: Sikh-majority Indian Punjab, with a substantial Hindu minority, sitting alongside a Muslim-majority Pakistani Punjabi tradition in a different script entirely. The core passage, Mark 10:35–45, concentrates this curriculum’s hardest terms — ransom, servanthood, Son of Man, Christ’s true Lordship versus worldly domination — into eleven verses that must carry the whole book’s theological weight. Get this passage’s vocabulary wrong and the doctrine of substitutionary atonement can be silently inverted into a Hindu self-performed-penance frame; get the “lord it over” contrast wrong and Jesus’ entire rebuke to the disciples collapses into incoherence.

Key findings

• 74 enforced glossary terms, of which 67 carry Critical or High doctrinal risk (9 Critical + 13 High inherited from the Romans baseline; 16 Critical + 29 High newly introduced by Mark’s narrative content).
• 30 doctrines tracked, matching doctrine_risk_registry.json exactly: 9 Critical, 14 High, 6 Medium, 1 Low — 23 of 30 require human theologian review, 6 require native speaker review, only 1 (Marriage and Creation Order) clears with automated review alone.
• Mark introduces a second mandatory-gloss term beyond the baseline’s ਮੁਕਤੀ (salvation): ਛੁਟਕਾਰੇ ਦਾ ਮੁੱਲ (ransom, Mark 10:45) must never be rendered with the Hindu penitential term ਪ੍ਰਾਸਚਿਤ, which would relocate Christ’s atoning act onto the sinner — the precise doctrinal inversion this curriculum exists to prevent.
• Mark’s far higher address-term frequency (“Teacher,” 13 occurrences vs. none in Romans) surfaces a new forbidden-substitution risk (ਗੁਰੂ) that the Romans baseline never had to enforce at this scale, plus a subtler risk (ਰੱਬੀ for “Rabbi”) the baseline never encountered at all.
• The Greek text’s own wordplay — κύριος/ἐξουσία (Christ’s true Lordship/authority) sharing roots with κατακυριεύω/κατεξουσιάζω (worldly domination, Mark 10:42) — has no Punjabi lexical equivalent risk, which is fortunate, but requires active, disciplined fencing to preserve rather than accidentally flatten the contrast Jesus draws.

Risks

• Highest-stakes single term: Ransom (λύτρον, 10:45) — Critical, no Romans precedent, requires the curriculum’s second mandatory clarifying gloss.
• Highest-stakes single passage: Mark 10:35-45 itself — the core passage carries five Critical-or-High terms in eleven verses (glory, cup, baptism, servant/slave, ransom, substitution, Son of Man) and anchors two of the curriculum’s nine Critical doctrines.
• Highest collision risk with Sikh tradition specifically: Servanthood (ਸੇਵਕ/ਦਾਸ) genuinely overlaps with the deeply positive Sikh value of ਸੇਵਾ (Guru ka Langar); left unanchored to Christ’s own pattern in 10:45, it risks reading as a virtue Sikhism already fully supplies, silencing the passage’s specifically Christological claim.
• Highest collision risk with Hindu tradition specifically: Ransom/atonement vocabulary (ਪ੍ਰਾਸਚਿਤ risk) and Gehenna (ਨਰਕ’s temporary-rebirth-cycle connotation) both risk quietly importing a merit/karma framework into passages built on substitutionary, once-for-all categories.
• Structural risk unique to Mark’s genre: The Messianic Secret’s narrative arc (silence commands at 1:25 through open confession at 15:39) can be lost if Phase 2 segment-by-segment translation resolves Jesus’ identity too early or too plainly in any single verse, since Mark’s rhetorical strategy depends on the delay itself.

Opportunities

• ਸੰਗਤ and ਮਰਜ਼ੀ carry forward and extend the baseline’s positive-bridge-term pattern: ਮਰਜ਼ੀ (“ਰੱਬ ਦੀ ਮਰਜ਼ੀ”) gives Gethsemane’s submission-to-God’s-will language ready-made warmth across Sikh, Hindu, and Muslim-Punjabi speech communities alike, with no competing doctrinal baggage to fence off.
• ਦਾਸ’s devotional precedent in Punjabi religious poetry (“Nanak das”) gives Mark 10:44’s total-self-subjection language dignity rather than degradation, provided the caste-reversal point (subjection commanded toward “all”) is stated explicitly.
• Punjabi’s fortunate lexical separation of ਪ੍ਰਭੂ/ਅਧਿਕਾਰ from ਹਕੂਮਤ ਕਰਨਾ (unlike Greek’s deliberate shared root) means the core passage’s contrast is easier, not harder, to keep clear in Punjabi — a rare case where the destination language’s structure does translators’ work for them, provided they resist stylistic variation that would accidentally re-link the roots.

Recommended actions

1. Load the extended translation_memory.json (74 terms) and 12_ai_translation_requirements.md before any Phase 2 work begins; treat every Critical/High term exactly as enforced, with zero substitution tolerance.
2. Route Mark 10:35-45 and all fourteen other Critical-doctrine passages to human theologian review as a standing rule, not a case-by-case decision — this passage alone justifies dedicated reviewer time given its five-term density.
3. Build the ransom gloss and the Gehenna gloss into Phase 2 tooling as automated triggers (parallel to the existing mukti-gloss trigger), so neither can be silently omitted by a busy reviewer.
4. Brief native speaker reviewers specifically on the ਸੇਵਾ/ਸੇਵਕ/ਦਾਸ fencing rule before Mark 9–10 review begins, since this is the doctrine most likely to be approved too quickly precisely because it sounds right to a Punjabi ear for the wrong (generic-virtue) reason.
5. Carry the Comparative Theology document’s Sikh/Hindu distinction forward into every reviewer briefing for this book: Mark’s exorcism, atonement, and Gehenna material each independently risk collapsing Sikh and Hindu categories together if reviewers are not reminded these are two distinct traditions with two distinct collision profiles.

---

This summary is derived from 07_semantic_analysis.md, 08_core_glossary.md, 06_linguistic_gap_analysis.md, 11_doctrine_analysis.md, assets/translation_memory.json, assets/bible_term_registry.json, and assets/doctrine_risk_registry.json. All figures above are drawn directly from those documents; no new risk determinations are made here.