Executive Summary — Gospel of John Kannada Language Package

Why it matters

Kannada already has a mature Bible translation tradition and an established core-term glossary (inherited from the Romans package: ದೇವರು, ಯೇಸು, ಕರ್ತ, ಪವಿತ್ರಾತ್ಮ, ರಕ್ಷಣೆ, ದೇಹಧಾರಣೆ, ಪುನರುತ್ಥಾನ). John’s Gospel, however, introduces a doctrinal concentration — the Logos prologue, the new birth, the seven “I am” statements, the Father-Son unity claims, and the Paraclete discourses — that stress-tests this vocabulary harder than Romans ever did. The core passage, John 3:1–21, is the single highest cumulative-risk stretch of text in this entire pipeline: it contains the book’s most dangerous default-mistranslation risk (ಪುನರ್ಜನ್ಮ for “born again”), its most quoted verse (3:16, with two Critical-risk terms inside one sentence), and the doctrinal seed of a risk — Aikya-merger collision — that recurs at greater intensity later in John 10, 14, and 17.

Key findings

• 113 glossary terms are now tracked for this curriculum: 47 inherited exactly from the Romans baseline, 66 newly defined for John.
• 77 terms are Critical or High risk (32 Critical + 45 High) and require human theologian review at every occurrence — a substantially higher concentration than Romans, driven by John’s dense Christological self-disclosure material (Logos, “I am” statements, Father-Son unity).
• 25 doctrines are documented for full-book coverage; 18 require theologian review (11 Critical + 7 High), 7 require native-speaker review (Medium), and 0 are automated-only — John’s doctrine set carries no Low-tier doctrine at all, unlike Romans.
• The single highest risk in the whole book is not in the core passage’s most-cited verse but in John 10:30 / 17:11,21-23: “I and the Father are one” and its extension to believers. This is the most natural verse in the Gospel for a Lingayat-background reader to reinterpret through Veerashaiva Aikya categories (the devotee’s identity dissolving into Shiva) — a risk the Romans baseline had already flagged for salvation and lordship but which John’s unity and “abide” (μένω) language make acute in new ways (chs. 10, 14, 15, 17).
• The core passage itself concentrates five of the twenty highest-ranked ambiguities identified in the linguistic gap analysis: γεννηθῆναι ἄνωθεν (“born again,” #2 overall), μονογενής (“only begotten,” #4), the πνεῦμα/wind pun (#5), ὑψόω (“lifted up,” #6), and the κόσμος/κρίσις cluster of 3:16-21 (#7-9).

Risks

• ಪುನರ್ಜನ್ಮ intrusion — the default Kannada word for reincarnation is the most likely translator error in the entire curriculum, striking at the core passage’s central doctrine (New Birth and Regeneration by the Spirit).
• Aikya-merger drift — a fluent, individually-correct-sounding Kannada rendering can still drift toward Veerashaiva merger theology by the end of a passage, especially across John 10:30, 14:9-11, 15:4-10, and 17:11-23; this is a trajectory risk, not a single-word risk, and requires whole-passage review, not just term-by-term checking.
• Jiva/moksha absorption of “eternal life” — ನಿತ್ಯಜೀವ’s root word ಜೀವ is the Vedantic/Jain technical term for the soul awaiting liberation; without explicit anchoring to John 17:3, the term can be silently reinterpreted as a moksha-adjacent claim.
• Mārga-pluralism softening — John 14:6’s exclusivity clause is the sharpest exclusivity claim in the Gospel, set directly against Karnataka’s culturally prestigious plural “path” (mārga) vocabulary; softening pressure here is high precisely because the surrounding culture is generally hospitable to plural-paths language.
• Avatar/Sharana absorption of Incarnation and Sonship — ಅವತಾರ remains forbidden, but the Lingayat-specific risk (Christ received as an exceptionally realized human Sharana figure who attained Aikya, rather than the eternal Son) is a distinct and separately-documented failure mode requiring its own translator vigilance.

Opportunities

• Kannada’s existing doctrinally-fenced baseline vocabulary (ರಕ್ಷಣೆ, ಪುನರುತ್ಥಾನ, ದೇಹಧಾರಣೆ, ಪವಿತ್ರಾತ್ಮ) transfers to John with zero renegotiation, giving this curriculum a strong running start compared to a fresh-language build.
• The genuine partial convergence between Lingayat anti-caste, anti-idol monotheism and biblical monotheism (documented in the comparative theology analysis) can be used pastorally as a bridge in teaching material — provided every such bridge is paired with an explicit note on where the traditions diverge (Aikya-merger vs. retained personal relationship).
• John’s own text supplies its own doctrinal safeguards in several places (17:3 defines eternal life relationally; 20:28 supplies the climactic deity confession; 19:30 supplies the finished-work confession) — these self-defining verses should be used as anchor cross-references throughout translator notes rather than requiring external doctrinal argument.
• The Haridasa bhakti tradition’s “dasa” (servant) self-designation, while requiring careful handling at John 15:14-15, offers a genuine, well-loved point of positive cultural contact for teaching Christian servanthood once reframed with grace-not-earned emphasis.

Recommended actions

1. Route John 3:1-21 as a single unit for theologian review before any sub-segment is approved in Phase 2, given its documented status as the highest cumulative-risk passage in the book.
2. Extend the Romans baseline’s forbidden-substitution enforcement to explicitly cover ಪುನರ್ಜನ್ಮ for new-birth language (not only resurrection) and ಐಕ್ಯ for abide/unity language (not only salvation/lordship) — both now encoded in analysis/12_ai_translation_requirements.md.
3. Require mandatory translator notes at all 15 first-occurrence terms identified in analysis/08_core_glossary.md Section 3, with the Father-Son unity cluster (10:30; 17:11,21-23) treated as the single highest-priority note in the whole book.
4. Pilot-test the ಮಾರ್ಗ rendering (John 14:6) and the ದಾಸ/ಸ್ನೇಹಿತ contrast (John 15:14-15) with native reviewers from both Haridasa-bhakti and Lingayat backgrounds before Phase 2 finalization.
5. Carry all 66 new John-specific terms and all 47 inherited Romans terms into Phase 2 exactly as recorded in the updated assets/translation_memory.json; do not re-derive renderings mid-pipeline.