Executive Summary: 2 Corinthians (English → Hebrew) Language Package

Why it matters

2 Corinthians is Paul’s most autobiographical letter, and for a Hebrew-speaking audience it lands in the single most theologically loaded linguistic environment in this entire pipeline. The core passage (5:11-21) compresses four Critical-risk doctrinal moves — reconciliation, new creation, the imputation doctrine’s negative mirror, and Christ “made sin” so believers “become the righteousness of God” — into eleven verses, and the letter as a whole introduces the curriculum’s single biggest new collision point absent from the Romans baseline: תשובה (teshuvah, repentance), the central category of Rabbinic penitential theology. Get this package wrong and Hebrew-speaking readers — many of them Messianic Jewish believers navigating Jewish covenantal identity and faith in Yeshua simultaneously — will either import self-effected-atonement theology into Paul’s grace-secured doctrine, or hear supersessionist claims against the Temple and Israel that the text never makes.

Key findings

• 96 terms now populate the combined translation memory (28 inherited exactly from the Romans baseline, 68 newly documented for this letter) — 19 Critical-risk and 17 High-risk (36 total requiring mandatory theologian review at the term level).
• 24 doctrines are tracked in the risk registry across all 13 chapters: 6 Critical, 9 High, 7 Medium, 2 Low — 15 doctrines requiring human theologian review, 7 requiring native speaker review, 2 automated-only.
• Unlike most target languages, Hebrew’s challenge in this letter is not vocabulary absence but semantic crowding: nearly every image (tents, veils, tablets, betrothal, aroma, comfort) has a rich, apt existing Hebrew term — the risk is that term’s competing Modern Hebrew, Rabbinic, or Kabbalistic default sense pulling the doctrine off course.
• One term — צדקה (tzedakah) — now requires a genuinely novel, passage-locked dual license: forbidden for general righteousness (5:21) exactly as in Romans, but correct and required for the ch. 8-9 charitable collection and the Psalm 112:9 quotation (9:9-10). This is the single highest operational risk identified in the linguistic gap analysis.
• The core passage’s reconciliation doctrine (5:18-20) sits on a term, התפייסות, whose ordinary Modern Hebrew sense (a mutual quarrel-patch-up) is structurally backwards from Paul’s unilateral, grace-initiated meaning — requiring mandatory anchoring language at every occurrence.
• Chapter 6:16’s “temple of the living God” (מקדש) imports the baseline’s Critical church/Israel supersessionism risk into an image with even higher emotional stakes: the specific, destroyed, still-mourned Jerusalem Temple.
• Chapter 13:14’s Trinitarian benediction compresses every Critical Godhead term into one sentence directly engaging the Shema — the single most doctrinally dense verse in the letter, matched in weight only by the core passage itself.

Risks

• Highest severity: tooling failure to track passage-context for tzedakah’s dual license, silently corrupting either the righteousness doctrine or the generosity doctrine.
• High severity: מקדש (6:16) taught without the mandatory Romans 11 grafting-in framing — risks either trivializing the historical Temple’s loss or implying the Church replaces Israel.
• High severity: תשובה (7:9-10) taught without explicit framing against self-effected Rabbinic atonement — risks silently displacing Christ-secured salvation with a human work.
• High severity: הכתב (“the letter,” 3:6) misread as an attack on Torah’s own sacred text rather than the law’s condemning function apart from the Spirit.
• Medium-high severity: documented cult/sect distortions in circulation (Jehovah’s Witnesses’ subordinationist rendering of 13:14 and Tetragrammaton insertion at 3:17/5:11; Sacred-Name movement Adon-to-YHWH substitution; anti-missionary polemical over-supersessionizing of 6:16) that this package must actively fence out, not merely avoid by omission.
• Medium severity: sarcasm-blindness in the “Fool’s Speech” (11:1-12:13) — flat literal rendering of “super-apostles” and similar ironic constructions could invert Paul’s rhetorical intent.

Opportunities

• Hebrew’s Biblical and Rabbinic depth supplies unusually strong positive assets unavailable in most target languages: עֵרָבוֹן (deposit of the Spirit) is the actual etymological source of the Greek ἀρραβών; כבוד’s root in “heaviness” naturally reproduces Paul’s “weight of glory” pun (4:17); the shlichut legal concept substantively illuminates apostolic authority; and Isaiah 40:1’s נחמה, Genesis 1:27’s צלם, and Genesis 15:6’s ח-ש-ב root all supply direct, teachable scriptural bridges to 2 Corinthians’ own argument.
• The Jewish tzedakah tradition, properly fenced from the general-righteousness sense, gives chapters 8-9 unusual persuasive and cultural resonance rather than being a risk to manage.
• The letter’s own vocabulary (new covenant, the letter/Spirit, temple, teshuvah) offers a rare opportunity to teach continuity rather than rupture between Israel’s Scriptures and Paul’s argument, if — and only if — every Critical/High framing note in this package is consistently applied.

Recommended actions

1. Load assets/translation_memory.json (this package) together with the Romans baseline TM for every Phase 2 segment; never process 2 Corinthians content against the Romans TM alone.
2. Enforce the tzedakah dual-license rule via passage-context detection, not keyword substitution, with mandatory translator notes at every occurrence of either sense.
3. Route all 15 theologian-review doctrines — and especially 5:11-21, 3:1-18, 6:14-18, 7:8-11, 8:1-9:15, 11:1-15, 12:1-10, and 13:14 — through mandatory human theologian review before Phase 2 sign-off; no exceptions for repentance or the Trinitarian benediction.
4. Lock the fixed-phrase renderings of 12:9 (“my grace is sufficient for you”) and 13:14 (the Trinitarian benediction) at the start of Phase 2 and verify document-wide consistency at every subsequent use.
5. Brief all human reviewers on the cult/sect distortion catalogue in 05_translation_landscape.md Section 4 before review begins, so flagged back-translations are recognized immediately rather than requiring re-education mid-review.

Full detail: 07_semantic_analysis.md, 08_core_glossary.md, 11_doctrine_analysis.md, 04_comparative_theology.md, 06_linguistic_gap_analysis.md, 05_translation_landscape.md, assets/translation_memory.json, assets/bible_term_registry.json, assets/doctrine_risk_registry.json.