Alchemical Hands in the Hypnerotomachia Poliphili

Marginalia, Scholarship & Reception

← All Documents

README

README.md — 1,416 words

Hypnerotomachia Poliphili -- Digital Marginalia & Scholarship

A digital humanities project documenting the readership, marginalia, and scholarship of the Hypnerotomachia Poliphili (Venice: Aldus Manutius, 1499), based on James Russell's PhD thesis Many Other Things Worthy of Knowledge and Memory (Durham University, 2014).

Live site: [https://t3dy.github.io/HPMarginalia](https://t3dy.github.io/HPMarginalia)

What's Here

SectionContent
**Home**Marginalia gallery -- annotated folio images with lightbox viewer, filters by manuscript
**Marginalia**109 individual folio pages with images, hand attributions, annotation type badges, and Russell's commentary
**Scholars**60 scholar profiles (59 with overviews, 11 historical figures) spanning 1899-2024
**Bibliography**109 entries grouped by relevance (Primary, Direct, Indirect), with collection/review badges
**Dictionary**94 terms across 15 categories with significance prose, cross-links, and provenance badges
**Timeline**71 events spanning 1499-2024: editions, annotations, scholarship, art inspired by the HP
**Manuscripts**6 annotated copies studied by Russell, with hand profiles and concordance data
**Docs**Project documents (methodology, architecture, audits, critiques, planning)
**Code**47 Python pipeline scripts with full source and descriptions
**Edition**Digital edition prospectus with phased roadmap
**Essays**Russell's Alchemical Hands + Concordance Methodology, grounded in DB evidence
**About**Database statistics, data provenance, and rebuild instructions

Architecture

SQLite is the source of truth. Python scripts generate JSON and static HTML pages. No framework, no build tools, no JavaScript dependencies.


db/hp.db                    SQLite database (24 tables, schema v3)
scripts/                    Python pipeline (47 scripts)
  build_site.py             Unified site generator (all pages + JSON)
  migrate_v2.py             Schema migration (idempotent)
  seed_dictionary.py        Dictionary terms (idempotent)
  validate.py               QA checks and audit report
  ingest_perplexity.py      Web research ingestion
  init_db.py                Initialize base schema
  catalog_images.py         Parse image filenames
  build_signature_map.py    Aldine collation formula
  extract_references.py     Extract refs from thesis PDF
  match_refs_to_images.py   Match refs to images
  add_hands.py              Annotator hand profiles
  add_bibliography.py       Bibliography and timeline
  pdf_to_markdown.py        PDF text extraction
  chunk_documents.py        Semantic chunking for RAG
site/                       Generated static site
  index.html                Marginalia gallery (JS-driven)
  scholars.html             Scholars overview
  bibliography.html         Full HP bibliography
  about.html                Project info and stats
  scholar/*.html            60 individual scholar pages
  dictionary/index.html     Dictionary landing (15 categories)
  dictionary/*.html         94 individual term pages
  marginalia/index.html     Folio grid index
  marginalia/*.html         109 individual folio pages
  timeline.html             Chronological reception history (71 events)
  manuscripts/index.html    Annotated copies landing
  manuscripts/*.html        6 individual copy pages
  digital-edition.html      Edition prospectus
  russell-alchemical-hands.html  Essay on alchemical annotators
  concordance-method.html   Concordance methodology essay
  docs/index.html           Project documents index
  docs/*.html               14 document pages (rendered markdown)
  code/index.html           Pipeline scripts index
  code/*.html               16 script pages (with line numbers)
  data.json                 Gallery data export with confidence flags
  style.css                 Base styles (warm parchment palette)
  scholars.css              Scholar/paper card styles
  components.css             Shared component styles (extracted from inline)
  script.js                 Gallery lightbox and filters

How to Rebuild


# Full rebuild from database
python scripts/migrate_v2.py        # Schema migration (safe to re-run)
python scripts/seed_dictionary.py   # Dictionary terms (safe to re-run)
python scripts/build_site.py        # Generate all HTML + JSON

# Validation
python scripts/validate.py          # QA checks + audit report

# Full pipeline from scratch (requires PDFs + PyMuPDF)
python scripts/init_db.py
python scripts/catalog_images.py
python scripts/build_signature_map.py
python scripts/extract_references.py
python scripts/match_refs_to_images.py
python scripts/add_hands.py
python scripts/add_bibliography.py
python scripts/migrate_v2.py
python scripts/seed_dictionary.py
python scripts/build_site.py

Data Provenance

DataMethodConfidence
Signature mapDeterministic (Aldine collation formula)Verified
Image catalogingDeterministic (filename parsing)Verified
Siena image matchesFolio-exact (explicit r/v naming)HIGH
BL image matchesFolio-exact (assumes photo# = folio#)LOW
Scholar summariesLLM-assisted (Claude)Unreviewed
Dictionary definitionsLLM-assisted (Claude)Unreviewed (Draft)
Hand attributionsLLM-assisted (from Russell's prose)Unreviewed
Scholar/bibliography metadataLLM-assisted + web searchUnreviewed
Timeline eventsLLM-assistedUnreviewed

Content marked Unreviewed is displayed with review badges on the site.

BL matches are displayed with LOW confidence badges and "Unverified" warnings.

Key Database Tables

TableRowsPurpose
`images`674Cataloged manuscript photographs
`signature_map`448Signature-to-folio concordance
`dissertation_refs`282Folio references from Russell's thesis
`matches`610Image-reference matches (26 HIGH, 218 LOW, 366 MEDIUM)
`annotators`11Identified annotator hands (2 alchemists)
`bibliography`109HP scholarship with gap analysis
`scholars`60Scholar profiles (1780-present)
`dictionary_terms`37HP terminology glossary
`dictionary_term_links`76Cross-reference links between terms
`timeline_events`42Publication/scholarship timeline (1499-2024)
`document_topics`34Multi-value topic cluster assignments

Project Documents

See the [Docs tab](https://t3dy.github.io/HPMarginalia/docs/index.html) for full text, or browse in the repo:

DocumentDescription
`HPCONCORD.md`Concordance methodology (6-step pipeline)
`HPDECKARD.md` / `HPDECKARD2.md`Deckard boundary audits (deterministic vs. LLM tasks)
`HPMIT.md`MIT Electronic Hypnerotomachia site analysis
`HPMULTIMODAL.md`Multimodal RAG architecture study
`HPproposals.md`Content quality improvement proposals
`HPAGENTS.md`Agent usage analysis
`HPEMPTYOUTPUTFILES.md`Empty output files post-mortem
`MISTAKESTOAVOID.md`12 lessons learned
`HPRACHAEL.md` / `HPWEBAESTHETICS.md`Visual design logic audit
`HPONTOCRIT.md`Ontology critique (9 issues, gap between model and implementation)
`HPengCRIT.md`Prompt engineering critique (9 patterns identified)
`HPromptTRANSCRIPT.md`Transcript of all user prompts
`AUDIT_REPORT.md`Validation results

Progress Log

Phase 1: Foundation (Database + Gallery)

Phase 2: Scholarship Layer

Phase 3: Architecture Hardening (V2 Migration)

Phase 4: Dictionary + New Page Types

Phase 5: Alchemist Analysis

Phase 6: Design + Deployment

Audits Conducted

What Still Needs Human Review

Dependencies

No web framework required. The site is plain HTML/CSS/JS with zero runtime dependencies.