Provenance & License

Source

Everything derives from a single scan: the Internet Archive item ahumandocumenta04mallgoog, a Google-digitized copy of the one-volume Chapman & Hall "New Edition" (1892) of W. H. Mallock's A Human Document.

This is deliberately the same edition Tom Phillips used for A Humument, so this project's page numbers map page-for-page onto both the printed book and Phillips's work — printed page N = humument page N = A Humument page N, for N in 1–367.

The multi-volume "three-decker" first edition has different pagination and does not line up with A Humument; this repo deliberately uses only the one-volume 1892 edition.

How the data was made

The scanned PDF is rasterized, OCR'd locally with Apple Vision (via ocrmac — no cloud APIs), deskewed and aligned, contrast-normalized to clean B&W, annotated with NLP features, and analyzed for whitespace geometry. The full process is described in The Pipeline.

data/humument.db is the canonical OCR artifact. Because Apple Vision OCR is nondeterministic, re-running the pipeline will not reproduce it byte-for-byte — so the database is versioned in the repository (via git-lfs) rather than treated as regenerable.

License

The 1892 novel is in the public domain. This project's code, the derived data, and the packaging are MIT-licensed — the humument repository and all three npm packages (humument, humument-data, humument-images).

See the LICENSE file for the full text.

Citing / building on it

If you make something with these tools, a link back to the repository is appreciated.