Imagines Coronatae

A machine-readable catalogue of sacred images crowned under papal authority, compiled from a complete sweep of the official gazettes of the Holy See — Acta Sanctae Sedis (ASS, 1865–1908) and Acta Apostolicae Sedis (AAS, 1909– ) — and from the one digitized volume of the Vatican Chapter's own coronation dossiers, Madonne coronate, tomo IV (1689–1714) — plus, kept visibly distinct, the coronations reported by six secondary works on the same archive: Bombelli's illustrated Raccolta of the crowned images of Rome (1792), Briccolani's printed register of 1800, the Zander/Magister catalogue of the Fabbrica's painted copies (2011), Vrabelová's study of Central Europe (2013), Balzamo's of Italy (2023), and — the bulk of everything after 1800 — the Basilici–Bigliazzi database of 2025, a modern compilation that is taken in as a lead to be checked rather than as a documented crowning.

Where	What
registry/by-country.md	Browsable registry, grouped by country
registry/chronological.md	Browsable registry, ordered by date of crowning
data/imagines-coronatae.json	The catalogue — one record per image, each carrying all its sources
data/imagines-coronatae.csv	One row per image, flat
data/attestations.json	The evidence layer — one record per act in ASS/AAS
data/attestations-acsp.json	The evidence layer for the Chapter archive — one record per dossier in Madonne coronate IV
data/attestations-lit.json	Secondary literature — coronations reported, not read at first hand, each with the author's archival citation
data/bombelli-1792-raccolta.json	Transcription of Bombelli's 1792 Raccolta, tomi II–IV — 74 Roman images with the day of crowning, the crown's cost and the plate's caption
data/briccolani-1800-serie.json	Transcription of Briccolani's 1800 Serie — the earliest printed register, 256 entries 1631–1793
data/zander-magister-2011-catalogue.json	The 96 entries of the Zander/Magister 2011 catalogue, each with its dossier citation
data/basilici-bigliazzi-2025-db.json	The Basilici–Bigliazzi database of 1,737 crownings, 1631–1981, as fetched; the 1,420 they vouch for are in the catalogue as LIT
docs/BASILICI-BIGLIAZZI.md	What their database adds, and where it disagrees with the other sources (data/basilici-in-catalogue.csv row by row)
data/basilici-rome-alignment.json	Their 133 Roman schede aligned by hand to the catalogue's church strings
data/sources.csv	One row per (image, source) pair

What is in it

1,578 crowned images in 50 countries, documented by 2,359 source citations: 400 from 147,576 pages of the two gazettes (ASS 1–41, 1865–1908; AAS 1–118, 1909–2026); 34 from the 713 transcribed folios of Madonne coronate IV — the Chapter's dossiers for 34 images crowned between 1689 and 1716, only three of which the gazettes ever mention; and 1,925 from secondary literature reporting coronations from volumes of the same archive that are not online — 256 from Briccolani's register of 1800 (every Chapter coronation to 1791, by year), 81 from Bombelli's Raccolta of 1792 (74 Roman images with the day of the crowning, and seven separate crowns for the Child), 107 from the Zander/Magister catalogue of 2011 (89 images with the day of crowning and the volume and folios of the dossier), 30 from Vrabelová 2013 on Central Europe 1717–1786, 31 from Balzamo 2023 on Italy and Rome 1631–1798, and 1,420 from the Basilici–Bigliazzi database of 2025 — the schede they mark as verified crownings, 1631–1981, with the day of the crowning for nearly all.

482 images are attested by more than one source — the most-cited is Savona's Nostra Signora della Misericordia, with nine. Crownings run from 1500 to 2005, far earlier than the gazettes themselves, because a later act often recites when an image was first crowned. 1,213 images rest on secondary literature alone; 909 of them on the Basilici–Bigliazzi database and nothing else, which is what a catalogue of the Chapter's practice after 1800 looks like while the archive's later volumes stay offline: leads, each marked low, each saying in its notes what it rests on.

By strongest evidence available for each image:

Type	Images
chapter_decree	955
papal_coronation_act	514
retrospective_attestation	69
papal_personal_coronation	34
papal_legate_deputation	4
norms	2

A papal_coronation_act from the Basilici–Bigliazzi layer means a crowning they attribute to a papal brief (B.P.), not an act read in the gazettes; the source's series tells the two apart.

Most-represented countries:

Country	Images
Italy	786
Spain	204
France	132
Poland	93
Mexico	70
Belgium	32
Argentina	24
Philippines	24
Colombia	18
Ukraine	18
Peru	16
Malta	14

Confidence: 233 high, 217 medium, 1,128 low. medium usually means either that the country was inferred from the Latin name of the diocese rather than stated outright (every such case says so in the source's notes), or that the only source is a secondary work citing the dossier by folio; low is a register that gives a year and nothing else, or a modern compilation that cites no folio at all.

One record per image, not per act

An image crowned once may be mentioned in a dozen later acts. Nossa Senhora da Conceição Aparecida is recited in AAS 23 (1931), AAS 46 (1954) and AAS 59 (1967) — that is one image with three sources, not three crowned images. So each record here is an image, and every act attesting it is kept in that record's sources array. Nothing is discarded by consolidation: 2,359 source citations sit inside 1,578 image records, and data/attestations.json, data/attestations-acsp.json and data/attestations-lit.json still hold the flat record-level layers if you want them.

Images crowned more than once. coronation_dates is an array, and where it holds more than one date the image really was crowned again — typically a Vatican Chapter crowning later renewed by a Pope. 66 images in the catalogue were crowned more than once, among them:

Image	Crowned
Madonna della Febbre — Rome, Vatican sacristy	1631-08-27, 1697-08-15
Salus Populi Romani — Rome, Santa Maria Maggiore	1597, 1838-08-15, 1954-11-01
Madonna del Rosario di Fontanellato — Fontanellato, Italy	1660, 1925
Nostra Signora della Misericordia — Savona, Italy	1770, 1815-05-10, 1915-04-19
Nostra Signora dei Miracoli — Cicagna, Italy	1790-09-14, 1814-09-14
Vergine del Sacro Monte — Varallo, Italy	1857-08-20, 1862-08-17
Nostra Signora di Bonaria — Cagliari (Calaris), Italy	1870-04-24, 1926-04-24
Nuestra Señora de Guadalupe — Mexicopolis (Mexico City), Mexico	1895-10-12, 1945-10-12, 1979-01-27
Notre-Dame du Cap — Cap-de-la-Madeleine, Canada	1904-10-12, 1954-08-12
Nuestra Señora de Coromoto — Guanare, Venezuela	1952-09-12, 1985-01-27

Dates are collapsed by precision first: one source giving 1954 and another 1954-08-29 is a single crowning recorded at two levels of detail, not two, so only the fuller form is kept (and where two full dates fall in one year, the one from a source read at first hand). A bare year from a register that is one year off a full date from a dossier is treated the same way — a register often gives the year of the decree, the dossier the day of the ceremony (Lucca: decree 1689, crowned 30 April 1690). Distinct years otherwise stay distinct — which is what makes a multi-entry list mean something. A crown for the Child of an image already crowned (Briccolani lists twelve such Bambino Gesù concessions) is kept as a source on the image, linked to it by parent_act, but is not counted as a re-crowning. Some of the 66 are source disagreements rather than two ceremonies — the 43 where a Basilici–Bigliazzi date differs from another source's are listed in docs/BASILICI-BIGLIAZZI.md, and five older ones are: Sant'Agostino in Rome (Briccolani 1641, Balzamo 1643), Jarosław (Briccolani 1732, Vrabelová 1755-09-08), two Roman images where Bombelli's notice contradicts his own plate — Santa Maria del Sole (notice 18 May 1669, plate and Briccolani 1665) and the Rosario of the Minerva (notice 28 August 1644, plate and Briccolani 1640) — and Lima, where the act of 1925 prints die XXIV mensis septembris anno MDCCXXI and Zander/Magister, citing the dossier in vol. 25, give 24 September 1921 — the same day a century apart, and the dossier's century is the likelier. They are left as the sources give them.

How images are identified. By place first, then title. Marian titles repeat all over the world — there are seven distinct Nuestra Señora de Guadalupe here, in Mexico, Spain, Venezuela and El Salvador — so a shared title is never on its own evidence of a shared image. Two records are the same image when they name the same locality and carry compatible titles. An earlier build also merged on a "distinctive" shared title across different localities; it was removed after it fused Notre-Dame des Miracles at Rennes with Mauriac, the Carmine la Bruna at Naples with Matera, and the four separate images John Paul II crowned together at Jasna Góra. A false merge destroys a distinct image; a false split merely leaves a duplicate — so the matching errs toward splitting. Titles are also stripped of descriptive words before comparison. Latin acts praise an image as much as they name it, and adjectives like perinsigne, thaumaturga or veneranda are not identity. Leaving them in kept Notre-Dame du Cap as two records — one act called it Simulacrum B. M. V. a Rosario, another perinsigne Virginis simulacrum — for a single shrine crowned in 1904 and again in 1954. Where an act's title reduces to nothing but praise, the shared locality and subject carry the identification, and a record is merged only if exactly one cluster at that place fits. Subjects never cross: a St Joseph is never merged into a Marian image.

Three further rules came in with the registers, which name a hundred and more images in one city:

• A place-word inside a title is not identity. Santa Maria della Neve di Frosinone is identified by Neve; the locality already carries Frosinone, and matching on it fused Acquapendente's Madonna delle Grazie (1742) with its cathedral's Concezione (1778). Words for kinds of building (cattedrale, chiesa, monastero) are dropped too, so Santa Maria nella Cattedrale di Verona has no title at all and the place and the year must carry it — and then a year-only register never merges with another register's entry at the same place on the year alone (Briccolani's Madonna delle Grazie of Benevento, 1723, is not the Camaldolese Incoronata of the same year, which Balzamo and he both set 'at Benevento' and which in fact stands at Sant'Angelo a Scala), unless one of them gives the day.
• Rome is not a place. Over a hundred distinct images were crowned there, so two Roman records match only when their church signatures agree — Rome, Sant'Agostino with Rome, S. Agostino, never San Giovanni in Fonte with San Giovanni dei Fiorentini — and a Roman record never matches one from anywhere else, however many saints' names the two churches share (San Giovanni in Fonte is not San Giovanni Valdarno). A record that says only Rome reaches a church through its title (Madonna del Nome di Maria → Rome, Santissimo Nome di Maria); a record that names no place at all never reaches a Roman church on a title word alone (the Madonna del Rimedio of Arborea, in Sardinia, is not the one in San Dionigi alle Quattro Fontane) — a Roman act says Romae.

And four that came in with the Basilici–Bigliazzi database, whose localities are modern and administrative (Scandriglia, frazione di Ponticelli Sabino):

• Administrative words are not places. Frazione, quartiere, villaggio are dropped from the place signature — every hamlet in Italy is somebody's frazione — and so is a suffix like de Juárez (Acapulco de Juárez is not Oaxaca de Juárez). A saint's name shared between two multi-word localities is not a shared place either: San Giovanni Rotondo is not San Giovanni Valdarno, though San Sebastián alone still matches San Sebastian.
• A register lists distinct images. Two rows of the same register crowned at the same place on the same day are two images, not one: at Caldarola on 17 May 1814 Pius VII crowned the Madonna della Pietà of Pievefavera and the Madonna del Monte. The same-day rule that lets a Latin act meet an Italian register never joins two entries of one work.
• Their exonyms are aliases. Cracovia, Siviglia, Ginevra, Città del Messico are mapped to the place the other sources name, and so is a sanctuary's own village against the city or diocese an act names: Bologhine and the gazette's Icosium are both Algiers, Chaudfontaine is the commune of Chèvremont, Ourém the municipality of Fátima, Orselina the hill of Locarno's Madonna del Sasso. A placeless gazette act reaches an image known only from their database when a distinctive title word and the year agree (the Caysasay act of 1954 finds their Taal).
• Rome by alignment. Their Roman churches are described at length and in their own words, so the 133 Roman schede are matched to the catalogue's church strings by a hand-checked table (data/basilici-rome-alignment.json) rather than by the signature rule, and their twelve separate crowns for the Child are linked to the image by parent_act like Briccolani's.
• Two full dates that differ are two crownings. Genoa's Madonnetta (27 June 1920) and Nostra Signora delle Vigne (21 November 1920) are not one image because both were crowned in 1920.

One earlier merge was undone by these rules and is left undone on purpose: Julius II's crowning of the Loreto statue in 1500 (Balzamo) and the 1922 crowning at Loreto are now two records — the acts of 1922 crown a nova Virginis effigies, the statue that replaced the one burned in 1921.

Residual duplicates are possible where a source names no locality at all, or where two acts describe one image in words too different to match. One such pair was left open for a while: Piekary Śląskie (issue #1), where AAS 55 (1963), 225 records a crowning Pii Pp. XI permissu and AAS 111 (2019), 1343 records August Hlond crowning the Piekary image as Apostolic Administrator, neither act giving a date or naming the other. Basilici–Bigliazzi (after Anselmo 1933) record exactly one crowning at Piekary, on 15 August 1925 by papal brief, with Pius XI's crowns and the Nuncio Lorenzo Lauri officiating — inside both acts' windows — so the two are now held to be one image; neither act states the date, so it reaches coronation_dates only through their own row, as a low source. The word that had kept them apart, mirifica, was praise leaking into a title, and is now on the descriptive stop-list.

Primary and secondary sources are kept apart

Every source carries a series: ASS and AAS (the gazettes) and ACSP (the Chapter's dossiers) were read at first hand for this catalogue. LIT is different: it marks a coronation reported by a scholarly work, a printed register or a modern compilation and not read here. Six works:

Work	LIT records	What it is	Citation it carries
[Pietro Bombelli], Raccolta delle immagini della Beatissima Vergine ornate della corona d'oro dal R.mo Capitolo di S. Pietro, Rome 1792, tomi II–IV	81 (74 images and 7 crowns for the Child)	One notice per crowned Roman image, facing an engraving: the day of the crowning, often the decree, the canons deputed, the cost of the crowns; the plate gives the support, the size and the year	none — 'le memorie dell'archivio', the memoriali and atti capitolari, no folio; confidence low
Vincenzo Briccolani, Descrizione della sacrosanta Basilica Vaticana, Rome 1800, pp. 144–161	256	The earliest printed register: every Chapter coronation 1631–1791, year by year, Rome first and then all Italy and abroad	none — a year only; confidence low
Pietro Zander and Sara Magister, Full of Grace: Crowned Madonnas from the Vatican Basilica, New Haven 2011	107 (96 entries; a reported re-crowning is a second record)	Catalogue of the Fabbrica's painted copies of 89 crowned images, with the day of crowning	BAV, ACSP, Madonne Coronate, vol., cc. for 83 of them, and the Fabbrica's catalogo delle immagini (AFSP, Arm. 12, F, 11, nr. 10)
Dana Vrabelová, Imago gratiosa, Prague 2013, § XXI	30	Thirty Chapter coronations in Central Europe, 1717–1786	sv., fol. for most
Nicolas Balzamo, « Uniformisation ou distinction ? », RHR 97 (2023)	31	Roman and Italian coronations 1631–1798 named in the text	volume, folio where given; town-and-year lists without
Massimo Basilici and Rita Bigliazzi, Le Madonne Coronate, 4 vols, 2025, and their database at pereto.org	1,420 (of 1,737 schede: those they mark verificata: sì and a real crowning)	One scheda per crowning, 1631–1981, with locality, church, title, day of crowning, decree date, authorising body and officiant; compiled from Anselmo 1933, the printed repertories and the internet, the archive consulted for particular cases	none — no folio; confidence low; their authority, officiant and notes in notes

The Basilici–Bigliazzi rows are the fullest list of the practice there is and the only coverage here of the Chapter's and the briefs' coronations after 1800, but they are a compilation, not a reading of the archive: 909 images rest on them alone. docs/BASILICI-BIGLIAZZI.md says what they add, which of their schede corroborate an image known from another source (498), and where their dates disagree with it (43); the 317 schede they themselves mark uncertain, refused or unverified are left out.

Each row is cited to BAV, ACSP, Madonne coronate, volume, folio where the author gives one, and those citations are carried in register_refs, so a LIT row can be taken back to the primary dossier; until it is, its confidence is capped at medium (low where no folio is cited). The rows are in data/vrabelova-2013-table-xxi.json, data/balzamo-2023-crownings.json, data/zander-magister-2011-catalogue.json, data/briccolani-1800-serie.json and data/bombelli-1792-raccolta.json. Rows that name only a town and a year are kept standalone — they are not merged into any named image on place and year, only on a shared title word, because a town and a year do not identify one.

Why keep the distinction visible: where the literature can be checked against the dossiers read here, it is mostly right and sometimes not. Vrabelová dates Trsat to 21 March 1715; the instrument says 14 September, the letters say 8 September. Where Briccolani and a tomo IV dossier meet (twenty-seven images), his year is the dossier's in all but three, and those are one year off (Bertinoro and Otranto a year early, Frascati a year late); where Zander/Magister meet a dossier or a gazette act (Otranto, Cuglieri, Narni, Bonaria) the day agrees exactly. Bombelli gives the day for fifty of his seventy-four Roman images, and where a tomo IV dossier covers the same image the two fit: San Pantaleo's receipt is dated 25 March 1694 and so is his crowning; for San Marcello the dossier's decree is of 16 August 1694 and his ceremony of 17 April 1695, for San Lorenzo in Borgo decree 3 November and ceremony 6 December 1696. Against Briccolani his year agrees in 71 of 74; in the three that differ (the Vittoria, the Madonna del Sole, the Rosario of the Minerva) the caption engraved under his own plate sides with Briccolani against his notice. He and Zander/Magister differ on the Madonna della Colonna (7 February against 1 January 1645), he and Balzamo on the Altemps chapel (15 against 12 April 1673). Secondary reports are valuable, but they are not the same kind of evidence.

What "crowned under papal authority" means here

Coronation of an image is a papal act, but it reaches the gazettes in several shapes. Each record carries an evidence_type so you can filter to exactly the sense you need:

evidence_type	Meaning
papal_coronation_act	The act itself grants or permits the coronation — "pretioso diademate redimiri sinitur".
papal_legate_deputation	The Pope deputes a Cardinal or Bishop to crown the image in his name.
papal_personal_coronation	The act records that the Pope crowned the image himself.
retrospective_attestation	A different act (typically a Basilica Minor elevation or a patron declaration) states in passing that the image was crowned, often "ex decreto Capituli Vaticani".
chapter_decree	The Chapter of St Peter's own dossier under the Sforza Pallavicino legacy: petition, decree of concession, instrument of donation of the crowns, and the relazione of the ceremony. From Madonne coronate IV.
norms	General legislation on crowning images.
petition_not_conceded	A petition to the Chapter that was refused or never granted — the image was not crowned by it. One case so far: Żabbar, Malta, c. 1700.

Important limits — please read before citing

1. This is not a global census. For most of the period covered, the crowning of an image was decreed by the Chapter of St Peter's Basilica (Capitulum Vaticanum) and never published in either gazette. What this dataset documents is what ASS and AAS attest — a well-defined and citable subset.

2. The Chapter's own archives are a separate source, only fractionally online. For three centuries the crowning of an image was granted by decree of the Chapter of St Peter's Basilica, under the 1636 Sforza Pallavicino legacy, and those decrees were never gazetted. The dedicated series is Madonne coronate, cited as BAV, ACSP, Madonne coronate, tomo I, foglio N (mecc. N). Exactly one volume of it is digitized in full — Arch.Cap.S.Pietro.Madonne.coron.4, 726 images, not OCR'd — and ten folios of vol. 19 (the Capocolonna dossier of 1892). Everything else must be consulted at the Vatican Library or at the Archivio Capitolare. That one volume has now been transcribed and swept (chapter_decree records, cited by folio); the other tomes remain the largest unexamined source, reached here only through the registers and catalogues that cite them. See docs/CHAPTER-ARCHIVES.md for the series layout, the citation form, the petition procedure and what each coronation deposited.

   A concrete case. Our Lady of Altagracia (Higüey, Dominican Republic), patroness of the country, is generally said to have been crowned on 15 August 1922 under Pius XI. AAS volumes 14 (1922), 15 (1923) and 16 (1924) do not mention Altagracia, Higüey or the Dominican Republic at all — nor does the 1959 Apostolic Constitution erecting the diocese of Higüey (AAS 51, 688) recite any crowning. That coronation is simply not in the gazette; its record lies in the Chapter's archives. What AAS does preserve is a later papal act: see below.

3. 1973 is a watershed. The Normae circa patronos constituendos et imagines B. M. Virginis coronandas of 25 March 1973 (AAS 65, 276) reshaped the practice; from 1974 the acts appear as Litterae Apostolicae granting the faculty to crown, and they are numerous.

4. A rule worth knowing. ASS 41 (1908), 621 records Pius X's ruling that images of the Sacred Heart are not to be crowned — a crown may only be laid at the statue's feet. That explains the near-absence of Sacred Heart coronations in the record.

5. The sources are OCR scans, and ASS is markedly rougher than AAS. Place names are frequently mangled. Every record keeps the verbatim Latin rubric and incipit and carries a confidence value. Anything marked low, or carrying a notes field, should be checked against the printed page before being treated as authoritative.

6. Several ASS volumes span two calendar years; year is the first year of the volume.

A note on borderline cases: San Giovanni Rotondo

The inclusion bar for this catalogue is that an image was crowned under papal authority — by a papal act, by a Legate deputed for it, by the Pope in person, or by decree of the Vatican Chapter acting on a papal indult. One case sits just outside that line and is worth recording explicitly, because it shows both what the bar excludes and what the gazettes incidentally preserve.

ASS 36 (1903–04), 226–231 — Sypontina, Iurium, a case decided by the Sacred Congregation of the Council on 18 July 1903. It is a property dispute at San Giovanni Rotondo, in the archdiocese of Siponto (Manfredonia), between Canon Michele Limongelli and the Confraternity of the Blessed Sacrament at the church of St Catherine, over who owned certain sacred furnishings — a silver monstrance, a silver crown, a chalice veil, a manuscript ledger, and a statue of St Agnes.

The crowning surfaces only as evidence about the crown's ownership. At p. 230 the confraternity argues that the canon had himself acknowledged the crown was bought with the faithful's offerings, saying so "dum functionem peragebat solemnem apponendi coronam fronti sacrae imagini" — while he was performing the solemn function of placing the crown on the forehead of the sacred image — and that he had called himself the faithful's mandatary "in apponenda hac corona in fronte B. M. Virginis".

So a solemn crowning of a Marian image at San Giovanni Rotondo is genuinely attested. But the record names no papal authority and no Chapter decree, gives no date for the crowning, and does not name or title the image; the coronation is incidental to a lawsuit about a piece of silver. It is therefore not in the dataset. It is noted here so that the omission is a recorded judgement rather than a silent gap — and because it is a reminder that the Congregational jurisprudence in ASS and AAS carries evidence of local crownings that a search for coronation decrees will never surface.

Record shape

{
  "id": "nossa-senhora-da-conceicao-aparecida-aparecida",
  "image_title_vernacular": "Nossa Senhora da Conceição Aparecida",
  "image_subject": "Blessed Virgin Mary",
  "locality": "Aparecida",
  "country": "Brazil",
  "coronation_dates": ["1904"],
  "first_coronation": "1904",
  "strongest_evidence": "retrospective_attestation",
  "confidence": "high",
  "source_count": 3,
  "sources": [
    {
      "series": "AAS", "volume": 23, "year": 1931, "page": 7,
      "citation": "AAS 23 (1931), p. 7",
      "evidence_type": "retrospective_attestation",
      "pope": "Pius PP. XI",
      "act_date": "1930-07-16",
      "coronation_date": "1904",
      "rubric_latin": "...", "incipit_latin": "...",
      "confidence": "medium",
      "notes": "Patron-declaration act; states the simulacrum was crowned 'decreto Capituli sacrosanctae Patriarchalis Vaticanae Basilicae'..."
    }
  ]
}

Recall check

An independent audit swept the corpus for the unambiguous act-granting formulas (redimiri sinitur, coronam apponendam, facultas conceditur … coronandi, coronasse imaginem, redimiendi serto, and others) and collected every page on which one occurs. Excluding back-of-volume index pages, 47 of 48 such act pages are represented; the one exception is an index entry whose underlying act (Our Lady of Fátima, AAS 38, 376) is included.

Three systematic traps were found and fixed during the build, and are worth knowing if you extend this work.

1. The act rubric switches from ALL-CAPS to lowercase italic around 1970, so a caps-based parser goes blind on everything after it.
2. The OCR hyphenates words across line breaks (pretioso dia-\ndemate redimiat), hiding the key terms from a naive search. De-hyphenating the corpus recovered 96 passages.
3. Requiring a word for "image" near the coronation word. This looks like a safe precision filter and is not: a papal act often names the devotion by its title alone. It dropped AAS 71 (1979), 158, where John Paul II, preaching at Santo Domingo on 25 January 1979, says of Nuestra Señora de la Altagracia that "el Papa quiere dejar como homenaje de devoción una diadema" — no word for image anywhere near it. Re-triaging the 348 passages that filter had rejected recovered that record and several more, among them the diadem presented at Guadalupe two days later (AAS 71, 177), the Virgen del Carmen de Maipú (AAS 66, 727), and the Virgen de la Caridad del Cobre, of which the Pope later recalled "el momento en que le ceñí la corona que sus hijos le ofrecieron" (AAS 91, 105).

Reproducing

Script	Does
scripts/fetch_aas.sh	Downloads all AAS PDFs and verifies each against its Content-Length.
scripts/fetch_ass.sh	The same for the 41 ASS volumes.
scripts/extract_text.py	Extracts the text layer of every PDF, with page markers.
scripts/find_passages.py	De-hyphenates and sweeps for coronation vocabulary; emits candidate passages.
scripts/find_passages_vernacular.py	Second sweep for Italian, Spanish, French, Polish and German coronation terms.
scripts/merge.py	Merges and de-duplicates the per-batch extraction output by locality.
scripts/build_dataset.py	Assembles the act-level evidence layer, data/attestations.json.
scripts/build_acsp.py	Turns the Madonne coronate IV extraction into data/attestations-acsp.json.
scripts/transcribe_briccolani.py, scripts/transcribe_bombelli.py	The transcriptions of the two printed registers, as Python rows, written out to data/.
scripts/build_basilici_db.py, scripts/align_basilici_rome.py	Normalise the fetched Basilici–Bigliazzi schede, and align their Roman rows to the catalogue's church strings.
scripts/build_lit.py	Turns the secondary-literature table into data/attestations-lit.json.
scripts/build_images.py	Consolidates acts into one record per image, with sources.
scripts/build_registry.py	Renders the two registry tables from the catalogue.
scripts/report_basilici.py	Writes docs/BASILICI-BIGLIAZZI.md and data/basilici-in-catalogue.csv from the built catalogue.

The reading-and-structuring step between the sweep and the merge was performed by parallel LLM agents against the specifications in docs/EXTRACTION_SPEC.md, docs/EXTRACTION_SPEC_ASS.md and, for the Chapter dossiers, docs/EXTRACTION_SPEC_ACSP.md.

Licence

The underlying ASS and AAS texts are © Libreria Editrice Vaticana. This compilation is offered for research use.