Translation Stance

This document states the translation commitments that guide every decision in the People's Open Bible. Declaring them is an act of honesty: critics can assess our output against our stated commitments rather than guessing at hidden biases.

Status: first draft. This document will be refined as the project matures. Input and disagreement are welcomed via GitHub issues.

Theological stance

The People's Open Bible does not commit to a specific denominational or creedal doctrinal stance. Where the source texts presuppose theological claims — the divinity of Christ, the resurrection, the personhood of the Holy Spirit, the unity of God — the translation reflects those claims as the texts present them, not as imposed commitments of the translator.

Readers from any tradition (Catholic, Eastern Orthodox, Protestant of any stripe, non-denominational, inquirer) are invited to evaluate POB against their own framework. Where a translation decision is influenced by a theological reading — for example, where a Greek construction is genuinely ambiguous and the English choice leans one way or the other — that choice is documented in the verse's theological_decisions block and the alternative rendering is preserved in a footnote.

We do not pick sides on denominational distinctives (Reformed vs. Arminian, infant vs. believer's baptism, real presence vs. symbolic, cessationist vs. continuationist, etc.). Those questions are for the reader and their faith community, not for the translation.

Translation philosophy

The People's Open Bible aims at optimal equivalence: the most faithful, intelligible English representation of the source's meaning, wording patterns, literary form, and rhetorical force. This is not an automatic midpoint between formal and dynamic translation. Readability serves source fidelity; convention and familiarity do not override it.

Specifically:

1. Original-language primacy with reader-conscious wording. Where long-standing English renderings obscure the source text's plain sense, we explain the source pressure openly. For doulos, the main text uses "servant" consistently while notes and the public audit trail preserve bonded-status, ownership, and total-allegiance nuance where it is historically or theologically salient (Rom 1:1, Phil 1:1).

2. Translate titles; use conventional name forms. Christological titles are translated to preserve meaning (Christos → "Messiah" when it functions independently as a title; Kyrios → "Lord"). In name-like constructions, use the familiar English forms "Jesus Christ" and "Christ Jesus." Personal names are otherwise transliterated by standard conventions. Thus "Jesus is the Messiah" retains the title, while Greek Iēsous Christos used as a personal-name construction is rendered "Jesus Christ."

3. Preserve theological tension rather than resolve it. Where the source is genuinely ambiguous on a contested interpretive question (e.g., pistis Christou — "faith in Christ" vs. "faithfulness of Christ"), we render one in the main text and preserve the alternative in a footnote, documenting the decision in the per-verse YAML.

4. Consistency of key terms, or documented exception. A given Greek or Hebrew word should receive a consistent English gloss across the translation, unless context demands variance — in which case the variance is documented per-verse with a rationale. The consistency_lint.py tool enforces this at commit time.

5. Readable modern English, not archaic or academic. The target reader is a literate 18–35-year-old English speaker. Vocabulary should be accessible without flattening theological weight. We avoid both ecclesiastical jargon ("propitiation" is footnoted, not required) and informality that undersells the text's gravity.

Source distinctions and optimal English

Source fidelity takes priority over mere familiarity or defensibility. A significant Greek or Hebrew wording pattern must receive an explicit comparison with the best source-transparent English candidate, even when a prior draft already discusses it in a footnote. Prior rationale is evidence, not a veto. Do not infer different meanings from morphology alone or invent theology to justify different glosses; preserve observable wording without adding claims. A concise source-qualified or hyphenated rendering is available when ordinary English otherwise conceals the source pattern. Record unresolved proposals for maintainer review; do not silently certify them as unchanged or auto-publish.

Approved scoped rendering: John 21:15–17 uses agape-love for ἀγαπάω and phileo-love for φιλέω, including the repeated question in verse 17. The ἀγάπη noun's general table entry below is not a rule erasing that verb pattern. Do not revert this approved POB/SPOB wording to generic “love”.

Contested terms

The following terms appear frequently and carry heavy theological weight. Our default rendering and rationale are recorded here. Per-verse variance is documented in the individual YAML files.

Repentance terminology (2026-10-13)

The contextual reassessment supersedes the October 11 reorientation default. Use repent / repentance for the moral-spiritual response where the passage supports it. Familiarity alone is not evidence of accuracy; explain the intended change of disposition and conduct rather than reducing repentance to sadness or a change of opinion. Reorientation can explain the concept but is not a mandatory replacement in reader text. No claim about measured reader reception or independent scholarly approval is made.

Context still controls. Retain regret, remorse, relenting, reconsideration, change of mind, or turning back where those fit the source and local direction. Preserve contested readings in Job 42:6 and Hebrews 12:17, the alternate turning reading in Jeremiah 31:19, and distinct turn/return verbs in Acts 3:19 and 26:20. Use "repentance toward God" in Acts 20:21 because the sentence explicitly gives that direction; do not insert it into every occurrence of metanoia.

Keep sorrow and repentance distinct in 2 Corinthians 7:9-10. Retain the relationship between repentance and its fruits, rather than equating the two. Use "I repent" in Luke 17:4 without requiring proof of completed reform before forgiveness. Preserve baptismal grammar without deciding symbolic versus causal interpretations by a vocabulary substitution.

Do not add intensifiers. Keep source-backed emphasis such as the existing "completely" and "from their whole heart" in Hermas 21:4. Keep the corrected "more joy ... than" in Luke 15:7. Extra-canonical writers retain their own conceptual frameworks, including adverse changes of mind; do not harmonize them to a canonical Godward formula. Historical English witnesses and source transcriptions remain untouched; reader adaptations retain their disclosure.

All 397 previously changed records have an explicit new decision, including retained exceptions and the four October 12 follow-up records. Earlier wording, notes, plans, and review history are preserved rather than erased. See the full decision ledger and exact, hash-bound plan in config/terminology/repentance-contextual-2026-10-13.json. These remain editorial drafts, not independent scholarly certifications.

What the translation is not

To prevent misunderstanding:

• Not a paraphrase. Every English rendering is tied to a specific source-text word or phrase, documented in the per-verse YAML.
• Not a study Bible. Footnotes document translation decisions, not doctrinal application or devotional commentary.
• Not a denominational product. The translation is intentionally cross-denominational. Contested readings are preserved in footnotes, not smoothed away.
• Not a finalized translation. Every verse is currently an AI-drafted rendering, released openly with full rationale. There is no formal scholarly review process at this time.
• Not a first-century English approximation. We aim at modern English that honors source-language meaning, not at archaism.

Translating into other languages

The People's Open Bible's long-term intention is to be translated into every major world language. Scripture belongs to the whole church; producing POB only in English would contradict the project's mission.

Each target language is its own translation — drafted directly from the original-language sources (Hebrew OT, Greek NT, Swete LXX for the deuterocanonical books) rather than from the English POB. Translating from English would compound the distance from the originals and inherit choices that make sense in English but not elsewhere. Every target language deserves its own encounter with the Hebrew and Greek.

Each language goes through several revision passes before it is considered stable:

1. AI-drafted first pass — a frontier multilingual model produces the initial verse-by-verse draft with full per-verse provenance, following the same methodology used for English (DOCTRINE.md translation philosophy §§1–5, per-verse YAML schema, doctrine_reference citations where relevant).
2. AI-reviewer revision — a second, distinct model reads the draft and flags awkward phrasing, dropped nuance, or places where the target language's idiom has drifted from the source's force. Revisions are discrete commits.
3. Native-speaker community review — before a language is declared stable, speakers of that language who are also familiar with the biblical source languages (or who have access to published lexicons in that language) audit the text and raise issues publicly.
4. Public issue window — a minimum public-review period with open GitHub issues before any language moves from "preview" to "stable."

Supported languages will be disclosed publicly based on model performance and feasibility. Not every language has an AI model capable of scholarly-accurate translation yet, and not every language has a sufficient community of native-speaking reviewers accessible to the project. We will publish the current supported- language list, and the ready-but-not-yet-stable list, as we reach the capability threshold for each. Languages that do not yet meet the bar are named as such, not silently omitted.

Every language published under this program carries the same CC-BY 4.0 license, the same per-verse provenance, the same auditability. There is no tier system across languages — every supported language gets the full People's Open Bible.

Change log for this document

Material changes to this document are themselves commits. Prior versions of the stance are preserved in git history.

• 2026-04-21: Removed creedal affirmations (Apostles' Creed, Nicene Creed, Chalcedonian Definition) and replaced with a theological-stance section that declines specific creedal commitment. Translation philosophy and contested-terms table preserved unchanged; those are methodology, not creedal commitment, and the project actually runs on them. Added a "Translating into other languages" section stating POB's long- term intention to ship in every feasible major language, with the full revision pipeline and transparent supported-language disclosure. Per-verse YAML citing the old doctrine_reference values is preserved as historical record; those are not retroactively edited.