2 Baruch (Syriac Apocalypse of Baruch) — scope and source strategy

This document covers the People's Open Bible's dedicated sub-phase for 2 Baruch, also called the Syriac Apocalypse of Baruch. It sits in the same pipeline family as 2ESDRAS.md: both are late Jewish apocalypses transmitted through layered daughter traditions, and both need a multi-witness workflow rather than the LXX-only pipeline used for the deuterocanonical books.

It complements EXTRA_CANONICAL.md (overall roadmap) and applies the three-zone source policy in REFERENCE_SOURCES.md.

Status: chapter-ready primary substrate now landed. The repo now has the full Ceriani primary OCR sweep, a bridged Syriac page corpus, tentative chapter buckets, targeted Kmosko control pages, and a scaffolded translation landing zone under translation/extra_canonical/2_baruch/. Verse-level refinement still remains ahead, but 2 Baruch is now ready for chapter drafting.

Why 2 Baruch belongs in the same family as 2 Esdras

2 Baruch and 2 Esdras are not the same textual problem, but they are close cousins from an engineering point of view:

• both are post-70 CE Jewish apocalypses
• both survive through translation history, not through a clean original-language base text inside our repo
• both require primary-witness translation plus secondary-witness apparatus, not reconstructed-source speculation
• both benefit from a per-verse (or per-section) provenance block listing which witnesses survive and where they diverge

So the right abstraction is not merely "the 2 Esdras pipeline," but a broader Semitic apocalypse multi-witness family, with 2 Esdras and 2 Baruch as the first two members.

Textual situation

The broad scholarly picture is:

1. Original composition: probably Hebrew or Aramaic, shortly after the destruction of the Second Temple.
2. Greek stage: the surviving Syriac points back to a Greek stage.
3. Primary surviving witness: the full work is preserved in Syriac.
4. Secondary witnesses: a few Greek fragments and a tiny Latin remnant survive, but not enough to replace the Syriac as the primary base text.

In other words: 2 Baruch is to Syriac what 2 Esdras is to Latin — the primary living witness is not the original language, but it is the base text we can responsibly translate.

Structure of the book

2 Baruch is usually divided into 87 chapters:

• Chs. 1–77 — the apocalypse proper
• Chs. 78–87 — the epistle / closing exhortation

That division matters for the pipeline because the back section should be modeled explicitly rather than treated as miscellaneous appendix matter.

Source editions (now pinned locally)

The following are the natural candidates for the 2 Baruch source stack. We are documenting them now so later source acquisition follows an already-audited plan.

Zone 1 primary / secondary (public domain or otherwise cleanly usable)

• Ceriani, Apocalypsis Baruch Syriace (1871) — primary Syriac base text
• Kmosko, Patrologia Syriaca edition (1907) — secondary Syriac / Latin-side scholarly witness
• Charles, The Apocalypse of Baruch (1896) — PD English reference only, for numbering / orientation, never as translation base
• Violet, Die Apokalypsen des Esra und des Baruch in deutscher Gestalt (1924, GCS 32) — German-form structural companion volume; useful for chapter/vision division and orientation, but not the same parallel-column witness layout as Violet 1910

Zone 2 consult (not reproduced)

• modern critical Syriac editions / commentaries
• modern translations in OTP / similar scholarly collections
• specialist studies on the Greek fragments and Syriac textual history

These may inform adjudication, but the same no-derivative-text rule applies here as everywhere else in POB.

Translation strategy

Adapted from METHODOLOGY.md and REFERENCE_SOURCES.md:

1. Syriac is the primary source text.
2. Greek / Latin remnants are apparatus, not replacement base texts.
3. The translator does not reconstruct a hypothetical Hebrew or Greek original.
4. Witness disagreements are exposed in the per-verse / per-section YAML with plain-English rationale.
5. Output remains fully auditable the same way our 2 Esdras and deuterocanonical work is.

Planned pipeline components

These are intentionally parallel to the 2 Esdras track:

Relationship to 2 Esdras work happening now

The current 2 Esdras work is already producing reusable pieces:

• local-PDF OCR pipeline design
• provenance sidecars per OCR page
• chapter-indexed cleaned primary-witness format
• multi-witness data model
• prompt-builder architecture

That means 2 Baruch should be included in the architecture now, even if its own transcription / loader work still starts later.

Kickoff materials now present

• sources/2baruch/README.md — source layout + witness roles + OCR notes
• sources/2baruch/MANIFEST.md — rehydration commands + pinned SHA-256 hashes
• sources/2baruch/CONTROL_WITNESSES.md — control-witness order + truth-check plan
• tools/prompts/transcribe_2baruch_ceriani_syriac.md — primary Syriac OCR prompt
• tools/prompts/transcribe_2baruch_kmosko_syriac_latin.md — secondary witness OCR prompt
• tools/prompts/transcribe_2baruch_violet1924_gcs32.md — German-form orientation / structure prompt
• translation/extra_canonical/2_baruch/ — chapter YAML draft landing zone scaffolded

Tentative timing

• Now: source acquisition scaffolded locally, with manifest + README + prompt docs in place, and the full Ceriani primary OCR sweep on disk
• Now: the chapter corpus is strong enough for real drafting to begin at the chapter level
• Next: keep spot-checking the remaining medium-confidence boundary pages (especially in the epistle ladder) while chapter drafting and later verse alignment proceed
• Drafting phase: after earlier higher-priority phases complete, using the already-proven family tooling

Why this is worth doing

• 2 Baruch is one of the most important Jewish apocalypses adjacent to 2 Esdras.
• It fits the same POB value proposition: transparent, auditable, open-license, witness-aware translation.
• Doing the architectural inclusion now saves rework later.