Executive Summary — 2 Peter — English → Dutch Language Package

Why it matters

2 Peter is short — three chapters, 61 verses — but it is the densest doctrinal text this pipeline has processed for Dutch: 11 of its 14 core doctrines (79%) and 42 of its 84 tracked terms (50%) are Critical or High risk, requiring mandatory human theologian review before any segment ships. The letter sits at the intersection of the Netherlands’ two defining religious realities: a doctrinally precise Reformed tradition (Canons of Dort, Heidelberg Catechism) that gives Dutch unusually exact vocabulary for election, patience, and salvation register — and a fast-growing secular “spiritualiteit zonder religie” culture that is primed to misread this letter’s single most distinctive verse (1:4, “partakers of the divine nature”) as license for self-divinization. Getting this letter wrong in Dutch does not just blur meaning; it can invert Christ’s character (the despotēs/“despoot” trap), collapse Scripture’s authority (graphē/“geschrift” trap), or hand a proof-text to New Age and LDS appropriation (theia physis/“goddelijke natuur”). The core passage, 1:16-21, alone activates three of the letter’s seven Critical doctrines simultaneously.

Key findings

• Doctrine risk profile: 7 Critical, 4 High, 3 Medium, 0 Low across 14 doctrines — every single doctrine in this letter requires at least native-speaker review; none clear automated-only review, unlike Romans’ 6 Low-risk doctrines.
• Term risk profile: of 84 terms tracked (28 inherited unchanged from the Romans baseline, 56 newly introduced by 2 Peter), 11 are Critical and 31 are High risk.
• Three forbidden-cognate/substitution traps unique to this letter: δεσπότης → never “despoot” (means “tyrant” in modern Dutch — would imply Christ is oppressive); αἵρεσις → never “sekte” (sensationalizes Peter’s warning about corrosion from within the church); γραφή → never bare “geschrift” (strips Scripture’s God-breathed, canonical claim).
• Two register-split terms mirror the baseline’s “behoud” pattern and require the same reviewer-confirmation discipline: σωτήρ (Zaligmaker/Heiland/Redder) and μακροθυμία (geduld/lankmoedigheid).
• Two genuine missing-vocabulary gaps exist where Dutch, like English, collapses a Greek distinction Peter relies on: γνῶσις vs. ἐπίγνωσις (both “kennis”) and φθορά vs. ἀπώλεια (both tending toward “verderf”) — both require per-occurrence Greek-term tracking in the segment cache, not a one-word fix.
• The core passage (1:16-21) is the letter’s highest-concentration risk zone, simultaneously activating Apostolic Eyewitness Authority, Sonship and Deity of Christ, and The Reliability and Inspiration of Scripture — all three Critical or High.

Risks

• Divine Nature and Escaping Corruption (1:4, Critical): highest cross-cultural misreading risk in the curriculum. “Goddelijke natuur” can be read through a Dutch New Age/mindfulness lens (“wij zijn allemaal goddelijk”) or an LDS “exaltation” lens — both denying the Creator-creature distinction this doctrine actually protects.
• Assurance of Calling and Election (1:10, Critical): directly continuous with Dutch Christianity’s single most historically loaded doctrinal flashpoint (the Canons of Dort vs. the Remonstrants); “verkiezing” also collides with the everyday political-election sense.
• Lordship and Sovereign Ownership of Christ (2:1, Critical): the despotēs/“despoot” forbidden-cognate trap is a live risk for any automated system or inattentive translator defaulting to transliteration.
• The Certainty of Christ’s Return / Day of the Lord (1:16, 3:4, 3:10, both Critical): consistency failures in “wederkomst” or “dag van de Heere” (including a single lowercase slip) sever the letter’s own structural argument and its central eschatological claim.
• Scripture Distortion vs. Scripture Reliability (3:16, High): Dutch university historical-critical culture risks conflating “some readings distort Scripture” with “Scripture itself is unreliable” — a move Peter explicitly does not make.
• Watchtower (Jehovah’s Witnesses) doctrinal patterns circulate widely enough in Dutch churches and door-to-door literature that 1:1 (“God and Savior”) and 2:1 (despotēs) require text-level, not just footnote-level, disambiguation to resist subordinationist readings.

Opportunities

• Reuse, don’t reinvent: 28 baseline Romans terms (God, Jezus, Heer, Zoon van God, genade, geloof, gerechtigheid, behoud, verkiezing, roeping, etc.) transfer to 2 Peter with zero modification, giving learners moving between the two curricula a stable, familiar vocabulary base.
• ἀρετή/“deugd” is a genuine improvement opportunity, not a risk to avoid: unlike δικαιοσύνη, this Greek word is correctly and confidently rendered “deugd” — reviewers should be told explicitly not to apply Romans-baseline caution here.
• Dutch’s existing “voorbede”/“voorspraak” and “zending” clarity (noted in the Romans baseline) has an analog here: Dutch’s dedicated theological vocabulary for “zondvloed” (the Noahic flood) and “wederkomst” (the Parousia) already gives this curriculum cleaner starting terminology than several other language pairs in this program.
• The letter’s OT judgment precedents (angels, flood, Sodom) and its Day-of-the-Lord/Yom-YHWH continuity offer strong, ready-made bridges to Dutch readers with Jewish or Islamic background, useful for supplementary interfaith-aware teaching material.

Recommended actions

1. Route all segments touching the 11 Critical/High doctrines to human theologian review by default; do not allow automated-only clearance anywhere in this letter (0 doctrines qualify for it, versus 6 in Romans).
2. Hard-code the three forbidden substitutions (despoot, sekte, geschrift) and the two forbidden inspiration-verb substitutions (gedicteerd, geïnspireerd) into Phase 2 validation before any segment is marked complete.
3. Add per-occurrence Greek-term tracking fields to the segment cache schema for the two missing-vocabulary gaps (γνῶσις/ἐπίγνωσις and φθορά/ἀπώλεια) before Phase 2 begins.
4. Mandate a standing theologian-reviewed gloss template for every occurrence of “goddelijke natuur” (1:4) that explicitly excludes both ontological-deification and New Age self-divinization readings.
5. Enforce identical rendering of “wederkomst” (1:16; 3:4; 3:12) and capitalized “dag van de Heere” (3:10, and cross-referenced at 2:9/3:7 as “dag van het oordeel”) at every occurrence via automated consistency checking, not reviewer memory alone.
6. Carry forward the baseline’s register-split reviewer-confirmation workflow (established for “behoud”) to this letter’s two new register-split terms, σωτήρ and μακροθυμία.

Bottom line: 2 Peter concentrates more doctrinal risk per verse than any curriculum this pipeline has processed to date for Dutch. The Language Package extends the Romans baseline cleanly with no contradictions, but Phase 2 execution must budget theologian-review capacity accordingly — this is not a curriculum where automated-only clearance is a realistic option for any doctrine.