Translation Review & Accuracy

The Shared Word Translation (TSW) and Fluent Translation exist to make strong, openly licensed English Scripture translations freely available to everyone. Open access and careful review belong together. Both translations are provisional and remain open to correction.

What has been completed

Both translations have draft coverage of all 66 books and 1,189 chapters of the Protestant biblical canon. The translations were developed with substantial artificial-intelligence assistance and with written translation philosophies grounded in Hebrew, Aramaic, and Greek source texts.

Existing project audits include automated checks of book and chapter coverage, verse labels, structural integrity, certain textual differences, and consistency of notes and vocabulary. An earlier TSW publication audit found no remaining issues within the structural and apparatus checks it performed on that snapshot. This is not a claim that the translation contains no language or interpretive errors. Selected passages and books have also received more detailed source and editorial review, with corrections recorded in the project repository.

Important distinction: complete draft coverage and successful automated checks do not establish that every verse has been independently checked by a biblical-language scholar.

What is being reviewed now

The work continues carefully, passage by passage and book by book, across the English text, notes, and vocabulary. Review asks whether the rendering accurately preserves the source text’s participants, actions, negations, relationships, significant lexical meanings, rhetorical structure, and meaningful ambiguity. Textual variants and legitimate interpretive alternatives are distinguished from demonstrable translation errors. AI-assisted analysis can help identify questions but is not presented as independent scholarly review.

Translation choices and corrections should be documented with their passage location, source evidence, explanation, and version history. The project’s editorial direction remains the responsibility of Matthew J. Skolnik.

Planned review by volunteer scholars

A future stage of the project is to invite qualified volunteer scholars with expertise in biblical Hebrew, Aramaic, Greek, and translation to examine selected passages independently. This program is planned, not yet represented as completed.

The intended approach has two separate parts:

• Targeted review: examine passages involving significant manuscript variants, complex grammar, disputed readings, poetry, idioms, or issues identified in prior review. These passages are selected for their difficulty; their findings cannot be treated as the average error rate of the Bible.
• Representative review: independently select a reproducible sample spanning both Testaments, source languages, and literary genres. This makes it possible to assess the frequency of confirmed material problems within the sampled passages, with appropriate attention to sampling uncertainty.

Reviewers will be asked to consider the original-language evidence, passage context, and the published translation philosophy. A defensible alternative interpretation will not be labeled an error merely because it differs from a reviewer’s preference. Substantial disagreements should receive additional examination.

How accuracy will be reported

The project intends to publish dated review results for each translation, identifying the version reviewed, the number of passages sampled, the review method, the qualifications and independence of reviewers, and any confirmed corrections or unresolved concerns.

Reports will distinguish structural checks passed, source-language review coverage, independent scholarly sample coverage, and confirmed material translation errors. If a suitable representative sample becomes available, the project may report a sample-based material-issue rate with its denominator and uncertainty. Findings from deliberately difficult passages will be reported separately.

No numerical, corpus-wide translation-accuracy rate is currently claimed. A figure such as “99% accurate” would be misleading without an independently defined method and appropriate evidence.

Current recorded baseline

As documented in repository records available through September 17, 2026: both translations have complete draft coverage; the earlier TSW structural/apparatus audit passed its stated tests; Fluent’s editorial tracker recorded Genesis 1–50 as source self-reviewed with open issues; and independent editorial review and final approval were still pending. These records are historical checkpoints, not an assertion that every subsequent revision has been independently verified.

For further detail, consult the public project repository, the TSW translation philosophy, the Fluent translation philosophy, and the free-use license and attribution statement.

Open to correction

A translation is not made trustworthy simply by the credentials of its author, the sophistication of its tools, or the number of automated passes performed. Trust is strengthened by examining the evidence, documenting decisions, identifying mistakes, and correcting them. That remains an ongoing commitment of The Shared Word Project.