Ched-ai/voynich-manuscript-metadata

Voynich Manuscript Metadata

Dataset Summary

This dataset contains structured metadata about the Voynich Manuscript (Beinecke MS 408), a famous 15th-century codex held at Yale's Beinecke Rare Book & Manuscript Library. The dataset includes three tables: pages, folios, and quires, providing comprehensive codicological information.

Dataset Structure

Configurations

This dataset has three configurations:

• —pages: Page-level metadata (226 records)
• —folios: Folio-level metadata (102 records)
• —quires: Quire-level metadata (18 records)

Pages Configuration

{
  "page_id": "f1r",
  "folio_id": "f1",
  "side": "recto",
  "quire_id": "qA",
  "currier_language": "A",
  "hand": "1",
  "illustration_type": "herbal",
  "section": "herbal_a",
  "has_text": true,
  "line_count": 31,
  "is_foldout_panel": false,
  "panel_number": null
}

Pages Fields

Folios Configuration

{
  "folio_id": "f1",
  "folio_number": 1,
  "quire_id": "qA",
  "is_foldout": false,
  "panel_count": null,
  "is_missing": false,
  "pages": ["f1r", "f1v"]
}

Folios Fields

Quires Configuration

{
  "quire_id": "qA",
  "quire_number": 1,
  "folio_count": 8,
  "is_complete": true,
  "folios": ["f1", "f2", "f3", "f4", "f5", "f6", "f7", "f8"],
  "section_primary": "herbal_a"
}

Quires Fields

Manuscript Sections

Dataset Creation

Source Data

Metadata extracted from:

• —IVTFF transcription file headers (ZL3b-n.txt)
• —voynich.nu section classifications
• —Davis (2020) scribal hand identifications
• —Beinecke Library catalog

Processing

Built by the Voynich Computational Analysis Toolkit (VCAT).

Considerations

Limitations

• —Scholarly disputes: Some classifications are contested
• —Section boundaries: Not all scholars agree on section divisions
• —Hand identification: Based on Davis 2020; other scholars may differ
• —Pre-1.0 schema: May evolve in future versions

Attribution

All metadata includes implicit attribution to the ZL transcription and voynich.nu. Disputed classifications are noted in the source documentation.

Related Datasets

This dataset is part of the Voynich Computational Analysis Toolkit (VCAT):

• —voynich-eva: EVA transcription text
• —voynich-transcription-mismatch: Cross-transcription comparison

Datasets can be joined on page_id, folio_id, or quire_id.

Quick Start

from datasets import load_dataset

# Load pages metadata
pages = load_dataset("Ched-ai/voynich-manuscript-metadata", "pages")

# Load folios metadata
folios = load_dataset("Ched-ai/voynich-manuscript-metadata", "folios")

# Load quires metadata
quires = load_dataset("Ched-ai/voynich-manuscript-metadata", "quires")

# Count pages by section
from collections import Counter
sections = Counter(p['section'] for p in pages['train'])
print(sections)

# Find foldout pages
foldouts = [p for p in pages['train'] if p['is_foldout_panel']]
print(f"Foldout panels: {len(foldouts)}")

Licensing

Processing code and dataset structure: MIT License

Underlying data: Compiled from IVTFF transcription headers (Zandbergen-Landini), voynich.nu section classifications, Davis (2020) scribal hand identifications, and the Beinecke Library catalog. Released as a research resource: cite the original scholars, and verify rights independently for commercial applications. See SOURCES_LICENSE.md for full provenance details.

Citation

@misc{voynich-manuscript-metadata,
  author = {VCAT Contributors},
  title = {Voynich Manuscript Metadata},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/Ched-ai/voynich-manuscript-metadata}
}

References

• —Zandbergen, R. voynich.nu - The Voynich Manuscript. https://www.voynich.nu/
• —Davis, L. F. (2020). How Many Hands Wrote the Voynich Manuscript? Manuscript Studies, 5(2), 271-305.
• —Beinecke Rare Book & Manuscript Library. MS 408: Voynich Manuscript. Yale University.

Contact

• —Repository: GitHub
• —Issues: GitHub Issues