chapter-level decision records generated from the live translation registry
The evidence beneath the English
Every consequential choice
leaves a record.
The published library was developed through direct Greek analysis, lexical work, and manuscript comparison. Where surviving witnesses materially disagree, the competing Greek readings and their effect on English remain open to the reader.
public provenance schema, designed to remain machine-readable and auditable
standardized inspectable Greek reference edition
Textual basis
SBL Greek New Testament
Edited by Michael W. Holmes. The project's standardized, publicly inspectable Greek reference edition. It is not presented as retroactive proof that every existing chapter originated from one uniform printed edition.
This is an honest reference claim, not a retroactive claim that all completed work began from one uniform printed edition. Material manuscript forks remain independently documented at the reading they affect.
Translation record
Jude 1
John707 recordsMark525 recordsMatthew987 recordsLuke929 recordsActs1146 records1 Thessalonians67 recordsGalatians138 records1 Corinthians357 records2 Corinthians250 recordsRomans340 recordsPhilippians84 recordsPhilemon22 recordsColossians74 recordsEphesians121 records2 Thessalonians36 records1 Timothy86 records1 Peter90 records2 Peter61 records1 John107 records2 John10 records3 John13 recordsJude23 recordsRevelation324 records
jude.1.jude122RescueText
ἐλεᾶτε/ἐλέγχετε; διακρινομένους ...
TEXTBOUNDARYPASTORALTRANSLATIONlow confidence
How many groups are addressed, and how should each be treated?
The wording and punctuation of verses 22–23 form one of Jude’s most difficult textual problems. NA28 and Tregelles distinguish three groups: some doubting, others rescued from fire, and others shown fearful mercy. Westcott-Hort can be punctuated as two groups, while the Byzantine text rearranges the clauses and emphasizes saving those who make divisions.
Why this default
The three-group reading best accounts for the changing verbs and objects without forcing the fire-rescue clause to refer back to the doubters; its low confidence remains visible because the transmission is unusually tangled.
Evidence
ExternalSBLGNT apparatus at 22–23: ἐλεᾶτε — WH, NA28; ἐλεεῖτε — RP; ἐλέγχετε — Tregelles. The following clause arrangement supports the Tregelles/NA28 three-group, WH two-group, and RP rearranged profiles represented in the options.
InternalThe decision record relies on the stated lexical, grammatical, or translation rationale.
Project defaultAnd have mercy on some who are doubting; save others by snatching them from fire; and have mercy on others with fear
Project default · Tregelles/NA28 three groups
WitnessesTregelles · NA28 three-group profileAlternativeAnd have mercy on those who are doubting; save them by snatching them from fire; and have mercy on others with fear
Westcott-Hort punctuation · two groups
WitnessesWestcott-Hort two-group punctuationAlternativeAnd rebuke those who are disputing; save others with fear by snatching them from fire
Byzantine arrangement
WitnessesByzantine tradition (RP)Record integrity
One registry drives both the reader and this record.
The manifest is regenerated from the translation source before every production build. Its source digest is 7c4cbee907fbc0e84976d6e78a615134e2288f8197ce8abcf67b0452155ddc84.
Decision-level revision history begins with this ledger. Earlier development is not reconstructed into false precision; future material changes can be appended to the relevant record with their reason and date.