Ten thousand texts, and most saints hold one

Claim (verbatim)

The Bibliotheca Hagiographica Latina numbers roughly nine thousand distinct Latin hagiographic texts, and their distribution across saints is one of the steepest inequalities in any literature. At the head sit the giants: Martin of Tours, whose dossier runs from Sulpicius Severus's Vita (BHL 5610) through the letters and dialogues, Gregory of Tours's four books of Virtutes, verse reworkings by Paulinus of Perigueux and Venantius Fortunatus, and on into dozens of medieval derivatives; Benedict, Nicholas, Mary Magdalene, Catherine - each a shelf. At the tail, the modal saint: one text, often one version of one passio, itself often surviving in a handful of copies - the founding masterpiece of the genre, the Passio Perpetuae (BHL 6633), survives in about nine Latin manuscripts. The mechanism is a market with increasing returns: a saint with texts got feasts, altars, and relic shares; feasts demanded lessons; lessons begot new versions; every new church under the patron's name was a new commissioning customer. Fame compounded like capital, and the poor stayed poor. The claim is that this is not an impression but a countable geometry with a heavy head and a one-text majority, legible in the repertorium's own index. Prediction restated: the top tenth of saints holds at least forty percent of all numbered texts while at least half of all saints hold exactly one.

Prediction clause (verbatim)

Prediction: tallying numbered items per saint-headword across the BHL index for the letters A through C (letter-suffixed numbers counted as distinct items, as the Novum Supplementum itself counts them; homonymous saints kept apart by the index's own headwords), at least 50% of saints will hold exactly one numbered item, and the top 10% of saints by item count will hold at least 40% of all items in the sample (primary clause: both inequality figures jointly; the verdict follows them); the test voids if the three letters yield fewer than 300 saint-headwords.

Kill-dataset (verbatim)

Kill: the BHL's index sanctorum (public-domain 1898-1901 volumes plus the 1986 Novum Supplementum), letters A-C - an items-per-headword histogram and top-share computation, runnable from the printed indices alone.

Novelty / leakage triage

anticipated in the literature — this exact test has never been run

The extreme inequality of dossier sizes is printed doctrine - Philippart's survey of the legendiers and the Philippart-Trigalet statistics count texts, saints, and the giant dossiers against the mass of one-text saints - but no items-per-headword histogram of the BHL index has ever been published. The printed totals (roughly nine to ten thousand numbered items over some thousands of saints) constrain the mean, not the median or the top-decile share, so the joint 50%-single-item / 40%-top-decile clause is unrun arithmetic on the printed index rather than a figure guaranteed by it.

• G. Philippart, Les legendiers latins et autres manuscrits hagiographiques (Typologie des sources du Moyen Age occidental 24-25, Turnhout, 1977)
• G. Philippart & M. Trigalet, 'Latin hagiography before the ninth century: a synoptic view', in The Long Morning of Medieval Europe (Aldershot, 2008)
• H. Fros (ed.), Bibliotheca Hagiographica Latina: Novum Supplementum (Subsidia Hagiographica 70, Brussels, 1986), as the counting instrument

Predictions

No prediction registered yet.