Editorial policy
How Divine Reader works.
Trust in a sacred-text library comes from being able to check. This page explains where every text comes from, how it is checked, what the labels mean, and what AI does and does not do here.
Where the texts come from
Divine Reader publishes public-domain and openly licensed editions. Each text's page names the edition, translator, year, licence and source. Today the library contains:
- The Torah — JPS Tanakh, Gender-Sensitive Edition (Jewish Publication Society, 2023), CC BY-NC 4.0; Hebrew (Miqra according to the Masorah) (Masoretic text, Leningrad Codex tradition), CC BY-SA 4.0.
- The Bible — King James Version (Church of England translators, 1611, 1769), Public domain (outside the United Kingdom).
- The Qur'an — Yusuf Ali (English) (Abdullah Yusuf Ali, 1934), Public domain in the United States; Arabic (Uthmani script, Hafs), Tanzil Project terms — text reproduced unmodified with attribution.
- Bhagavad Gita — Edwin Arnold, The Song Celestial (Sir Edwin Arnold, 1885), Public domain.
- The Tipitaka (selections) — Plain-English retelling (Divine Reader editorial retelling, after public-domain and free-distribution translations, 2026), Retelling © Divine Reader; underlying translations acknowledged on the Sources page.
- Sikh Daily Prayers (Nitnem) — English translation (English translation attributed to Dr. Sant Singh Khalsa), Rights status under review — shown with attribution; removed on request of the rights holder.
- Tao Te Ching — James Legge (1891) (James Legge, 1891), Public domain.
- The Kojiki (selections) — Condensed English retelling (Divine Reader editorial retelling, drawing on Basil Hall Chamberlain's 1882 translation, 2026), Retelling © Divine Reader; Chamberlain (1882) is public domain.
How accuracy is maintained
- Canonical order. Books, surahs, chapters and banis appear in their traditional order, never alphabetically.
- Nothing silently edited. Translators' marginal notes and footnotes are separated from the text and shown as labelled translator's notes, off by default.
- Scope labels. A text marked Complete contains every chapter of the work. Selections means excerpts of a real translation. Retelling means a condensed modern paraphrase written for orientation — not a translation, and not to be quoted as scripture.
- Corrections in public. Known issues and fixes are listed on the Sources and licences page.
Four kinds of text, always distinguished
| Kind | How it appears |
|---|---|
| Original sacred text | In its own script (Hebrew, Arabic), marked as the original, shown alongside the translation when you switch it on. |
| Translation | The main reading text. Its translator, edition and licence are named on every page. |
| Translator's notes and retellings | Notes are collapsed by default and labelled “Translator's note”. Retellings carry a notice at the top of every page. |
| AI-generated explanation | Never on a Divine Reader page. Explanations live in The Scripture Guide, are labelled as AI, and cite the passage they discuss. |
What AI does and does not do
Nothing in the reading text, headings, notes or search results is generated by a language model. Search is a plain full-text index of the editions above. If you follow “Ask The Scripture Guide”, you leave Divine Reader for a separate tool whose answers are clearly labelled as AI-generated.
How Divine Reader is funded
Reading is free and requires no account. Divine Reader is supported by The Faith Vault. Texts under non-commercial licences (such as the JPS Tanakh via Sefaria) are never placed behind a paywall or paired with rewards. If optional supporter features are added in future, the sacred texts themselves will remain free.
Report an error
If a verse is wrong, misattributed or mislabelled, write to contact@thedivinereader.com with the passage reference. Corrections are logged on the Sources page.