A Doorway for Scholars · For Researchers
For dissertations, articles, critical editions, and classrooms. The Conway text is verbatim, every work has a stable identifier and a permanent URL, the revision history is public, and the corpus is exposed through standard library-harvesting protocols.
Chapter I
Machine-readable surfaces
Five endpoints. Free, CC0 metadata, underlying texts in the public domain.
§ 1
OAI-PMH 2.0
A static OAI-PMH 2.0 responder under /api/oai-pmh/, with the four standard verbs.
Identify— repository metadata + admin contact.ListMetadataFormats— currentlyoai_dc(Dublin Core) only.ListIdentifiers— every work's OAI identifier, year-stamped.ListRecords— full Dublin Core records for all 177 works./api/oai-pmh/get-record/<slug>.xml— per-work GetRecord (e.g.the-gods.xml).
Repository identifier scheme: oai:filthylittleatheist.com:conway:vol-N:<slug>. Because the responder is static, parameter handling (?from=, ?until=, resumptionToken) is not honored — harvesters should pull the full ListRecords in a single pass.
§ 2
JSON APIs
Equivalent data in JSON, indexed for direct programmatic use.
/api/works.json— every work with title, slug, year, volume, category, URL, word count, and the per-work text URL./api/downloads.json— manifest of every per-work export (txt / json / bib / ris) with byte sizes./api/work-history.json— Git-derived revision history for every transcription, plus per-slug feeds at/api/work-history/<slug>.json./api/concordance.json— the named-entity concordance summary; per-work cross-references at/api/concordance/<slug>.json./api/glossary.json— the editorial glossary feed used by the in-text tooltips./api/concept-graph.json— nodes + edges for the concept-category co-occurrence graph (also rendered as SVG).- Per-volume JSON at
/api/volume-N.json(e.g.volume-1.json) lists each volume's works in canonical order.
§ 3
Per-work downloads
Every work page has a Download & export modal behind the Share button on the floating reader toolbar. Each work ships in four formats.
.txt— plain UTF-8 body, unformatted..json— structured record with the stable Dresden identifier, year, volume, category, license, full body, and word count..bib— BibTeX entry suitable for TeX..ris— RIS for Zotero, EndNote, Mendeley, and most reference managers.
Download URLs follow /works/<slug>/downloads/<slug>.<ext> — predictable enough to script.
§ 4
Sitemaps + feeds
/sitemap.xml— XML sitemap for crawlers./feed.xml— main RSS (essays + works)./works/feed.xml— works-only RSS./this-day/feed.xml+/api/this-day.json— every dated event from the timeline as Atom + JSON Feed.
Chapter II
How to cite
The cite-dialog on every work page produces APA, MLA, Chicago, and BibTeX citations ready to copy. The full guide lives at /citations/.
§ 5
Stable identifiers
Every work carries a permanent Dresden identifier: conway:vol-N:<slug> (e.g. conway:vol-1:the-gods). The identifier appears in the JSON-LD CreativeWork graph on each work page, in the OAI-PMH record, and in the BibTeX/RIS exports. URLs at /works/<slug>/ are guaranteed stable: redirects are kept on slug changes, and the per-work revision history records every textual change.
§ 6
Per-paragraph permalinks
Every paragraph in every work has an ID of the form #pN — for example /works/the-gods/#p42. The cite-dialog's "Cite at" toggle lets you pick "Whole work" or "Current paragraph"; the chosen anchor is appended to the URL in every citation style. When the printed Conway pagination is captured into the source text, the same dialog will surface "Current page" using #page-volV-NNN anchors automatically.
§ 7
Author identity
The JSON-LD links Paine to the standard authority files.
Citations should name Thomas Paine as author and The Works of Thomas Paine, Conway Edition (G. P. Putnam's Sons, New York, 1900–1902) as the source of record. The on-site URL is the digital-edition pointer.
§ 8
Version history
Every work has a sibling page at /works/<slug>/history/ that lists every recorded revision to the transcription, drawn from the public Git history. Substantive corrections are dated, attributed, and visible at the commit level. If your citation needs to specify the exact version of the text you used, the history page lists every commit hash; cite the short SHA in a footnote.
Chapter III
Bulk access
Three ways to take the corpus offline.
§ 9
Manifest harvest
Iterate the works[] array in /api/downloads.json and fetch each format you want. Predictable URLs, no auth, no rate limit.
curl -s https://filthylittleatheist.com/api/downloads.json \
| jq -r '.works[].downloads.txt' \
| xargs -I {} curl -s -O "https://filthylittleatheist.com{}"§ 10
OAI-PMH dump
One-pass Dublin Core dump for library-grade harvesters: /api/oai-pmh/list-records.xml. The full corpus in a single response.
§ 11
Source repository
Clone github.com/jonajinga/filthy-little-atheist and read src/works/*.md directly. Each file is the canonical text plus the metadata that drives the on-site rendering.
Chapter IV
Open questions worth pursuing
The corpus is transcribed and cross-referenced; it is not exhaustively interpreted. Five places where the machinery could accelerate someone else's work.
§ 12
Reconstructing the lecture circuit
The seed dataset at /lecture-tour/ covers eighteen verified appearances. The newspaper evidence for the full thirty-year tour (1869–1898) is recoverable from regional digital newspaper collections; the data model accepts contributions.
§ 13
Variant readings of the major debates
The Field, Gladstone, Manning, and Black exchanges each ran first in The North American Review, then as standalone pamphlets, then in the Conway text. A proper variant collation has never been published.
§ 14
Reception history
Citations of Paine in late-19th-century freethought, suffrage, and labour periodicals are scattered and largely undigitised. The per-work concordance would carry a lot of weight here.
§ 15
Influence on twentieth-century atheism and humanism
Bertrand Russell, Mark Twain, Susan Jacoby, Christopher Hitchens — the line of citation runs but has never been mapped.
§ 16
Linguistic and rhetorical analysis
The corpus is now machine-readable in plain text and structured JSON, which makes it suitable for stylometric, sentiment, and topic-modelling work. Per-work word counts are in /api/works.json.
Chapter V
Get in touch
If you need more than what's on the surface — source TXT files used for ingestion, an early look at unfinished editorial work, coordination on textual variants — write to me.
§ 17
Research contact
Use the contact form and mention this page. A footnote naming the archive (Filthy Little Atheist, filthylittleatheist.com) is appreciated but not required. The text is public domain; the commentary is CC BY-SA 4.0.