Are there digitized databases of Cayce readings with metadata for quantitative analysis?
Executive summary
Yes — there are multiple digitized corpora of the Edgar Cayce readings that can be searched and mined, ranging from the Association for Research and Enlightenment’s (A.R.E.) online reading database and specialized health database to CD‑ROM editions and large Internet Archive “mega collections” with full‑text search files [1] [2] [3] [4] [5]. However, public availability of structured, research‑grade metadata (standardized fields suitable for large‑scale quantitative analysis) is uneven: some resources provide searchable text and reading identifiers, while full metadata access and reuse are constrained by archival controls and membership/licensing rules [6] [7] [8].
1. What exists: digitized readings and searchable engines
The A.R.E. hosts what it calls the complete set of 14,306 readings and an online searchable readings database, and it explicitly markets topic and sample reading search tools that expose reading identifiers and excerpt text [1] [2]. Historically the readings were packaged on CD‑ROM with a built‑in search engine — editions described in library catalog records and publisher notes — and those CD‑ROM versions have been the basis for research projects and derivative digital compilations [4] [9] [10].
2. Metadata and searchability: what the records show
The digitized resources commonly include reading numbers and topical indexing — the Explore Readings site demonstrates individual reading references such as “5618‑5” and offers topical listings across thousands of subject entries, which functions as basic metadata useful for targeted queries [2]. The Edgar Cayce health database aggregates roughly 8,000 health‑related readings into topical categories and explanatory material, which is effectively curated metadata for that domain [3]. Yet the publicly visible descriptions stop short of documenting a standardized, machine‑readable metadata schema (e.g., fields for date, recipient demographics, transcription confidence) in the sources provided [3] [2].
3. Downloadable corpora and third‑party collections for analysis
Independent digital collections exist: Internet Archive hosts large “mega collections” and individual scanned volumes with OCR/hOCR and searchtext files (example: hOCR JSON and searchtext dumps), which researchers have used as primary data sources for keyword extraction and frequency analysis [5] [11] [10]. Those extracted text files and hOCR outputs enable offline quantitative work (text mining, word frequencies, concordances) even when formal metadata schemas are absent, because the readings’ numeric identifiers and the OCR’d full text provide anchors for analysis [5] [10].
4. Access limits, rights, and archival controls affecting research
The Edgar Cayce Foundation/A.R.E. retains copyright and requires authorization for publication and reuse of materials, and some collections are explicitly governed by reference policies and restrictions on identifying recipients — a practical and legal limit on wide republication or sharing of enhanced metadata [6] [8]. A.R.E. also positions the full online searchable database behind membership or member‑only access in some branches of its network, meaning friction for researchers who need bulk downloads or programmatic APIs [7] [1].
5. How researchers have actually used digitized readings
Published student and institutional projects cite CD‑ROM editions and A.R.E. data as sources; one example study on cancer used the CD‑ROM readings as its primary data, extracting text via structured keyword searches and conducting quantitative frequency analysis, demonstrating that available digital formats have already supported basic quantitative research [10]. At the same time, those projects worked around limited metadata by relying on keyword queries, manual curation, and the reading identifiers embedded in transcriptions rather than on a comprehensive metadata API [10].
6. Bottom line — ready for large‑scale quantitative work, with caveats
The readings are digitized and searchable in multiple forms — A.R.E.’s online and health databases, CD‑ROM editions, and large OCR’d archive collections supply the raw material and reading identifiers necessary for computational analysis [1] [2] [3] [4] [5]. Researchers seeking a fully standardized, machine‑readable metadata package (complete fields for dates, recipient attributes, transcription provenance, and reuse licenses) will find that publicly documented, downloadable metadata is limited or governed by A.R.E. policy; practical quantitative work is possible today but often requires negotiation for access, custom metadata creation, or reliance on OCR/text dumps and the reading numbers present in existing transcriptions [6] [7] [10].