Gathering deposit for EA-CORPORA-01, the primary source corpora seated in the Crimson Hexagonal Archive. Fifteen seats totalling 126,714 ref-tagged lines across Greek, Hebrew, Latin, Classical Chinese and Sahidic Coptic: Josephus, Catullus, Theocritus, the Greek New Testament, the Septuagint, Longinus, Philo, Plato, the Hebrew Bible, Sappho 31 as a diplomatic layer, the literary papyri, the Slavonic Josephus as index and provenance, Homer, the Yi Jing, and the Coptic Gnostic corpus. Each seat holds as-fetched originals, flat ref-tagged normalized text, a commit-pinned source.json recording origin, edition and licence, and a SHA-256 manifest. The governing principle is originals wherever they exist: where only a translation is available the seat is not taken, and where a licence cannot be verified at the file itself the seat is not taken. Three absences are declared, dated and reasoned rather than papered over — the Apocryphon of John, the remaining Nag Hammadi codices, and the Qumran sectarian corpus. Data is published as a sidecar index at api/corpora.json with per-seat links; this record is the gathering, not the data.
Gathering record for fifteen seats. The individual seats 01/01 through 01/12 were deposited 2026-08-19 as separate records; 01/13 Homer, 01/14 the Yi Jing and 01/15 the Coptic Gnostic corpus are seated here for the first time. This record gathers all fifteen and publishes the sidecar.
| seat | corpus | lines | language | licence |
|---|---|---|---|---|
| 02 | catullus | 2,311 | — | CC BY-SA (Perseus packaging); underlying edi |
| 15 | coptic-gnostic | 2,997 | Sahidic Coptic | CC-BY 4.0 — declared in each file's TEI <ava |
| 04 | gnt-nestle1904 | 7,943 | — | public domain (1904 edition); repo tagging C |
| 09 | hebrew-wlc | 23,213 | — | CC BY 4.0 (OSHB); text public domain |
| 13 | homer | 27,903 | — | CC BY-SA (Perseus); underlying edition publi |
| 14 | iching | 1,509 | Classical Chinese (zh-Hant) | Project Gutenberg License; underlying text p |
| 01 | josephus | 12,423 | — | CC BY-SA (Perseus Digital Library); underlyi |
| 06 | longinus | 185 | — | CC BY-SA (Perseus); underlying edition publi |
| 05 | lxx-swete | 29,277 | — | CC BY-SA 4.0 (First1KGreek); underlying edit |
| 11 | papyri | 136 | — | CC BY (Duke Databank / papyri.info data) |
| 07 | philo | 7,901 | — | CC BY-SA 4.0 (First1KGreek); underlying edit |
| 08 | plato | 8,203 | — | CC BY-SA (Perseus); underlying edition publi |
| 10 | sappho | 21 | — | CC BY-SA (carrier); diplomatic layer CC-BY-4 |
| 12 | slavonic-josephus | 0 | — | CC-BY-4.0 (index layer) |
| 03 | theocritus | 2,692 | — | CC BY-SA (Perseus Digital Library); underlyi |
Total: 126,714 ref-tagged lines.
data/corpora/<corpus>/ original/ as-fetched, unmodified text/ flat ref-tagged lines: '<Work> <ref>\t<text>' source.json origin repo, pinned commit, edition, licence, verified loci MANIFEST.sha256 one line per file
The normalization is deliberately flat. A ref-tagged line survives compression, quotation and re-serialization in a way that nested markup does not; a reader or a machine that receives one line receives its address with it.
Originals wherever they exist. Where only a translation is available, the seat is not taken. The Yi Jing is seated in Chinese, not in Legge's English. The Coptic Gnostic corpus is seated in Sahidic, not in Meyer's rendering.
Licence verified at the file. For the Coptic Gnostic seat every one of the thirty-one TEI files was checked for its own <availability> header rather than trusting the repository README. All declare CC-BY 4.0.
Absence declared rather than papered over. Three gaps are recorded in the corpus source with the date searched and the reason.
The Apocryphon of John. The foundational Sethian text, surviving in four manuscripts — NHC II,1, III,1, IV,1 and BG 8502,2 — and the closest thing Gnosticism has to a canonical scripture. No openly licensed machine-readable Coptic edition exists. Coptic SCRIPTORIUM does not hold it. The one machine-readable edition found derives its Coptic from scanned PDFs of Meyer 2007 and Linssen, relicensed CC BY-SA by a party who does not hold those rights. Schmidt 1905 is public domain but is the Pistis Sophia volume, and the archive.org OCR strips Coptic script entirely. Till 1955, the Berlin Codex edition, remains in copyright.
Not seated rather than seated defectively. A corpus whose licence is asserted by a party who cannot grant it would place a rights defect inside the sovereign record — the failure this archive exists to document.
Nag Hammadi Codices I and III–XIII. Fifty-one of fifty-two tractates. Print editions only.
The Qumran sectarian corpus. Qumran-Digital publishes Open Access with a permissive robots.txt but is JS-rendered with no discoverable text endpoint; the Leon Levy Digital Library returns 403. Transcribed decades ago, publicly funded, and not openly machine-readable.
One text, four witnesses. A synoptic open edition of the Apocryphon of John is the narrowest and highest-value gap in the openly licensed Coptic record. The Qumran corpus is the largest. Both are stated here so that the gap is on the record with a date against it.
Perseus tlg003 is the Epigrams, not the Homeric Hymns. Catalogued as the Hymns in places; the grc1 file at commit a065c359 opens Ἄνδρες ἄγρης ἁλίης. Seated as Epigrams; the Hymns are a separate seat, not taken.
The Coptic seat is named coptic-gnostic, not nag-hammadi. Of fifty-two tractates only NHC II,2 is seated. Naming the directory for the library would assert a completeness the seat does not have, and future retrieval would inherit the overclaim.
∮ = 1