The SIBI Translation Process

The Scattered Israelites Bible (SIBI) is translated through a multi-stage pipeline. Each stage has a specific purpose, and every stage is visible to the reader. The goal is zero bias, maximum faithfulness — leaning toward the source language (Hebrew/Greek) rather than toward natural English.

The stages are:

1. Manuscript Import — the raw Hebrew or Greek text, morphology, and Strong's references
2. SIBI-P1 — root-faithful rendering of each word in isolation (no context)
3. SIBI-P2 — context-aware rendering that starts from P1 and only adjusts where necessary
4. Manual Review — human verification of every rendering before final approval

Before any translation begins, the source text must be imported word by word from the original manuscripts.

Each word is stored as a WordOccurrence with a unique position in its verse, linked to its morphological analysis, Strong's reference, and lexical information. This is the foundation everything else builds on.

P1 is the heart of the SIBI. Each Hebrew or Greek word is translated in complete isolation — based purely on what the word itself means, not on the verse around it.

What P1 considers

1. Root meaning — the consonantal root (e.g., אמן = firmness, faithfulness) and its core semantic field
2. Morphology — grammatical gender, number, person, and verbal stem (Qal, Niphal, Hiphil, etc.)
3. Semantic range — the full range of meanings the word can carry, drawn from the SILEX (Scattered Israelites Lexicon)

What P1 deliberately ignores

• The surrounding verse context
• How KJV, ESV, NASB, or any other English translation renders the word
• English readability or naturalness

Why isolate words?

Context awareness is where translation bias enters. When a translator reads the verse and chooses an English word that "fits," they are often influenced by theological tradition, denominational preference, or the gravitational pull of familiar translations. By rendering each word without that influence, P1 creates an unbiased baseline that exposes the original meaning.

P1 examples

P1 renderings are stored in the SILEX (Scattered Israelites Lexicon) and are generated by an AI model that has been given only the word's root, morphology, and semantic range — no verse context, no existing translations.

67824 unique word forms rendered so far.

One of the most significant examples of how common translations obscure the original text involves the Divine Name Yahweh (יהוה, YHWH).

Strong's assigns two different numbers to what is actually the same word:

The consonantal text is identical in both cases — יהוה. The only difference is the vowel pointing added by the Masoretes, Jewish scribes who preserved the Hebrew text between the 6th and 10th centuries CE.

Why the different pointing?

The Name יהוה was considered too sacred to pronounce aloud. When reading Scripture, Jews substituted Adonai in its place. The Masoretes encoded this practice by writing the consonants יהוה with the vowels of Adonai.

But when the text already contained אֲדֹנָי (Adonai) immediately before יהוה — as in the phrase אֲדֹנָי יהוה — reading both as "Adonai" would produce "Adonai Adonai," an awkward repetition. To avoid this, the Masoretes pointed יהוה with the vowels of Elohim instead, signaling the reader to say "Elohim" in that position.

What common translations do

When a reader sees "Lord GOD" in the KJV, the underlying Hebrew is Adonai Yahweh — two different words with distinct meanings. The SIBI restores this distinction. The vowel difference between H3068 and H3069 is a reading convention added centuries after the text was written, not part of the original inspired text.

P2 takes the P1 rendering and asks: does this root-faithful word work in the verse? The default answer is yes. P2 only modifies P1 when there is a clear, specific reason — never for English readability, never to match common translations.

P2 is a multi-layer process

The critical principle

SIBI leans toward the source language, not toward natural English. A P1 rendering like "royal-decree" is better than "law" because it preserves a distinction the Hebrew makes. The AI is given a strict rule: do not flatten distinct Hebrew terms into the same common English word. If P1 uses a specific, root-meaningful term, keep it — even if a simpler English word exists.

What P2 keeps

• P1 fits the verse context (even if awkward in English)
• P1 uses a root-meaningful or distinctive word
• P1 preserves a distinction between Hebrew synonyms
• Common function words (and, the, to, in, etc.)
• P1 verb form matches the morphology

What P2 may change

• Remove parenthetical notes like "(as reason or ground)" — keep the core word
• Resolve slash alternatives like "reaching/touching" — pick the most appropriate one for the context
• Remove literal morphology markers like "(masculine singular)"
• Adjust verb tense only if grammatically impossible in the narrative
• Choose a different sense from the semantic range only if P1 picked the wrong one for this specific context

Every P2 rendering is subject to manual review before it is considered final. The translation team uses the SIMAP (translation comparison tool) to examine each verse word by word, comparing Common, P1, and P2 side by side.

Manual overrides can be applied to any word. Once a rendering is marked as manually reviewed, it is permanently locked — no automated process can overwrite it. This ensures that human judgment has the final word.

19 manual overrides so far.

The translation process is supported by several purpose-built tools, all accessible on this site:

Most Bible translations are the product of committees whose deliberations are never published. Readers are asked to trust that each word was rendered faithfully, with no way to verify.

The SIBI inverts this. Every translation decision is traceable:

1. The manuscript word (Hebrew/Greek surface form)
2. Its morphological analysis
3. Its Strong's reference and full lexical entry
4. The P1 rendering (root-faithful, no context)
5. The P2 rendering (context-aware, with rationale)
6. Whether the rendering was AI-generated or manually reviewed
7. The rationale for any change from P1 to P2

This is not just a translation — it is an open record of how every word of Scripture was brought into English.