About — Acta Sanctorum

The Acta Sanctorum

The Acta Sanctorum is the largest critical collection of hagiographic texts ever assembled. Arranged by liturgical feast day from January through November, it fills sixty-eight folio volumes published between 1643 and 1940. The work collects, edits, and annotates the primary sources for the lives of the saints: passions, vitae, miracle collections, translation accounts, and liturgical records, drawn from Latin, Greek, and vernacular manuscripts across Europe.

The project was conceived in 1603 when Father Heribert Rosweyde, a Jesuit at Antwerp, proposed a systematic edition of hagiographic texts based on manuscript sources rather than the unreliable popular compilations then in circulation. Rosweyde died in 1629 before the work could begin in earnest. The task passed to John Bolland (Ioannes Bollandus, 1596–1665), who transformed Rosweyde's plan into something far more ambitious: not merely an edition of texts, but a critical apparatus that would evaluate, compare, and annotate every source for each saint's cult.

Bolland published the first two volumes — covering January — in 1643, with the assistance of Godefrid Henschenius. The work continued through successive generations of Jesuit scholars who came to be known as the Bollandists. As Hippolyte Delehaye wrote in 1920:

What has been achieved is an impressive heritage for those who must continue the work. It is true that among its admirers are many who are not unwillingly awed by the imposing number and size of the volumes of the Acta Sanctorum and who have a kind of superstitious respect for the Bollandists: “Sacrés ils sont” — there is no need to complete the quotation.

The publication reached November before the series was interrupted. December was never completed. The Bollandists continue their critical work today through the journal Analecta Bollandiana and the Subsidia Hagiographica series, both published from Brussels.

This Edition

40,336,564

Words of English translation

January–August complete · 244 days · 19,516 text sections
Translated from 29,336,259 words of Latin

This site presents a new English translation of the first eight months of the Acta Sanctorum — January (Volumes I–III), February (Volumes I–III), March (Volumes I–III), April (Volumes I–III), May (Volumes I–VII), June (Volumes I–VII), July (Volumes I–VII), and August — based on the Latin text digitized by the Ökumenisches Heiligenlexikon and supplemented by manual transcription from the Paris edition. The translation encompasses vitae, passiones, miracle collections, translation accounts, and the Bollandists' own critical commentary across all 244 published days.

The translation aims for clear, scholarly English prose that preserves the structure and reference apparatus of the original: section numbers, marginal annotations, footnote markers, and cross-references are retained. Proper names are given in their Latin forms. The Bollandists' annotations — often as valuable as the texts they accompany — are translated in full.

Editorial headnotes identifying each saint, the genre of text, and the relevant Bibliotheca Hagiographica Latina (BHL) number are being added progressively.

Text Types

The Acta Sanctorum contains several distinct genres of hagiographic writing. Understanding these categories, established by the Bollandists themselves and refined by later scholars like Delehaye and Aigrain, helps the reader evaluate what each text offers:

Vita: A biographical account of a saint's life. These range from near-contemporary records of great historical value to late medieval compilations heavily shaped by convention. The Bollandists typically present multiple vitae when they survive, allowing the reader to compare sources.
Passio: An account of a martyr's suffering and death. Early passiones based on court records (acta proconsularia) are among the most historically reliable hagiographic documents; later literary passiones are more formulaic.
Miracula: Collections of miracles attributed to a saint, usually organized around a shrine. These are valuable sources for social history, medical practice, and popular devotion.
Translatio: An account of the transfer of a saint's relics from one location to another. Translationes document the politics of relic veneration and often contain topographical and institutional detail found nowhere else.
Commentary: The Bollandists' own introductions, annotations, and critical discussions. These prefatory commentaries often surpass the texts they introduce in scholarly value, synthesizing evidence from martyrologies, liturgical calendars, and parallel sources to establish dates, identify persons, and evaluate authenticity.

How This Translation Is Made

Every section of English on this site is drafted with the help of a large language model — Claude, from Anthropic — working from the Latin, and is then checked by a procedure that has been written down, tested, and revised across more than fifteen thousand sections of text. The first half of that sentence is what people ask about. The second half is what decides whether the translation is any good, so it is set out here in full.

Setting the Latin

The Latin is not read off page images. It is taken from the Ökumenisches Heiligenlexikon’s digitization of the Antwerp edition, which is a transcription rather than an OCR read and is correspondingly cleaner than anything a direct read of seventeenth-century folio type would produce. The scraped pages carry the host site’s own furniture — German navigation, headers, footers — and that material is stripped before chunking. It is not always stripped cleanly, so a footer-leak scan runs over every day before it is accepted.

The Latin for each day is then cut into sections of roughly 1,500 words on paragraph boundaries. Where the digitization is defective or absent, the passage is transcribed by hand from the Paris edition. Each day carries a citation to its page in the Paris edition, linked to the scan on the Internet Archive, so that any sentence here can be checked against the printed folio without taking this site’s word for anything.

The English

The English is literal and it is parallel: section for section, and marker for marker. Section numbers, marginal annotations, footnote letters, and cross-references stand where they stand in the Latin rather than being smoothed into the prose. Proper names are given in their Latin forms, which is what makes an entry findable against the Bibliotheca Hagiographica Latina and against the Bollandists’ own indices.

The Bollandists’ commentary and annotations are translated in full, to the same standard as the texts they accompany. That is deliberate. Those notes carry much of the scholarly value of the Acta — the dating arguments, the manuscript comparisons, the judgments on authenticity — they have never been put into English, and they are the first thing an abridgment drops.

How the Work Is Batched

The unit of work is the calendar day, not the section. One agent takes one or two complete days — typically seventy to ninety-six sections — and only days too large for that are sharded across several. August 25 alone runs to 451 sections and needs seven.

Whole-day ownership is a verification decision rather than a scheduling one. When one worker owns a whole day, the check that every Latin section produced exactly one English section is arithmetic; when a day is split across workers, gaps hide at the seams. Across the full August run, thirty-five agents took all thirty-one days — 2,912 sections — with thirty-five clean completions, no gaps, no footer leaks, and valid metadata on every file.

January through August are now complete: 244 days, 19,516 text sections. August was published incrementally as it was translated, rather than held back until whole as earlier months had been — a change of practice made because the saint-splitting pass that divides a day into individual entries produces slightly different divisions once the surrounding days exist. That meant a number of early-published August entries moved to different addresses once the finished month was re-split as a whole; the day-by-day index reflects the final addresses, and no text published during the partial run was withdrawn.

What Is Checked, and When

Before a day is accepted it passes a set of scripted audits: a one-to-one count of Latin sections against English sections, a scan for untranslated host-site furniture, metadata validation on every file, and a duplicate-translation scan. That last one reports false positives by design — the Greek Acts printed with the Bollandists’ facing Latin look like duplicates to any similarity measure — and the known ones are enumerated in advance so that a real duplication is not lost among them.

At month’s end the whole month is re-split into saint entries and swept. A corpus-wide integrity pass over all 308 entries of 25,000 words or more found exactly one genuine mis-merge, an April 22 entry with five foreign saint headers buried inside it; the re-split otherwise held. A separate pass over the footnote apparatus of a single February dossier found 111 of 114 references binding correctly to their definitions and three running on past the block that governs them.

Several of these checks exist because an earlier judgment failed. Midway through the August run five headers were “corrected” on a premise that turned out to be wrong — the detector already handled the form in question, and 333 published files across January through July use it. The edits were reverted from snapshot and verified identical to the originals. A register of known open defects is kept with the project and consulted before new work rather than after.

What Is Known to Be Wrong

A corpus of this size has defects, and it is more useful to name them than to imply there are none. Sixteen days of January Volume II carry two complete independent translations of the same saints, and which of the two is authoritative is not yet settled — any word count keyed on the day number alone doubles those days. One set of secondary Acts in February runs I, II, IV, and whether Chapter III is missing from the translation or from the source has not been established. Three footnote references in that same dossier still display without their definitions. These are open, they are logged, and they are fixed when there is reason to touch that month again.

What We Do Not Correct

Plain scrape artifacts are repaired. Anything genuinely uncertain is flagged rather than guessed. Where a citation lost its numeral in transmission it renders as an empty bracket rather than a plausible number, because an invented column reference is worse than a visible gap.

The Bollandists’ own text stands as printed. Where a source contradicts itself — five hundred soldiers in one paragraph and seven hundred in the next, fifty sheep against five hundred — both readings are kept; the Bollandists flag these themselves, and it is their edition, not an improvement on it. Paragraph numbering that jumps in the folio jumps here. Long editorial brackets in which the Bollandists supply material absent from their own manuscript are preserved as brackets rather than absorbed into the text. Where a dropped footnote letter was restored because the surrounding sequence made it unambiguous, that was reported rather than done silently.

Why Translate This Way

Beneath the question of human against machine lies a prior question that settles it: nobody was doing this. No one had attempted the full Acta Sanctorum in English before. The cost is measured in human lifetimes and has not moved since the nineteenth century, so the corpus stood no closer to English this year than it would a thousand years from now. A corpus of this size is not translated slowly. It is not translated at all.

Two other corpora in the same condition are being attempted the same way, by the same hand and under the same procedure: Bonaventure’s Opera Omnia at bonaventure.wrootpress.com, and Migne’s Patrologia at migne.app. None of the three had been attempted whole before.

So the aim is not a perfect translation. It is to unlock the language gate to the tradition. Someone with no Latin can now reach the dossier, follow what the Bollandists are arguing about a saint, and see exactly where to check them.

There will be errors, including serious ones — as there are in Migne, and in the Bollandists, and in every edition of this kind ever assembled. The model carries a fault forward, a line or a spacing or an artifact of the transcription, much as a clerk copying his exemplar carried forward the mistake set in front of him. Working this way has made me feel nearer to the medieval copyists than to a modern translator.

A Working Draft

Everything here is a draft and should be cited as one. Corrections are wanted, not merely welcomed. If you read Latin and find something wrong — a mistranslation, a dropped section, a garbled citation — write to wilson@wrootlabs.com or use the form at the foot of this page. It will be fixed, and the correction recorded. Inquiries from scholars who would like to review or contribute are welcome at the same address.

This site is static HTML with no JavaScript. The translations are maintained as markdown files and compiled to HTML by a build script. It is typeset in Cormorant Garamond, EB Garamond, and Source Sans 3.

In Print

The apparatus that makes this site useful for reference makes it awkward for reading. The Acta Library is a print companion for those who want the texts themselves: the Bollandists' commentary set aside, and everything the Church preserved of one saint bound as a single reading.

The first volume is Saint Agatha of Catania — the early Acts of her passion, an anonymous Greek recension, the retelling of Simeon Metaphrastes, the oration of Methodius of Constantinople, the history of her body's return from Constantinople, two collections of her miracles, and the hymns of her feast, all drawn from the fifth of February. It is published in English and in Spanish; an Italian edition is in preparation. Further volumes will follow.

There is also Acta Sanctorum: A Daily Devotional, three hundred and fifteen lives from January 1 to November 10, a page to each.

Nothing here sits behind any of that. The translation on this site is free and stays free.

Rights and Reuse

The Latin is public domain and free without condition. The English translation, the editorial apparatus, and the structured text are © 2026 Wilson Pruitt, licensed CC BY‑NC 4.0 — free to quote, teach, post, mirror, and build on, with credit; not free to sell. Commercial use is available by permission and given readily. Full rights statement.

About This Translation