The Verification Venue · the plague everyone gets wrong
The Dancing Was Real. The Death Toll Came Later.
In the summer of 1518, in Strasbourg, people began to dance and could not stop. That much is documented, in more than one contemporary hand. The part everyone repeats — that fifteen a day danced themselves to death, a hundred in the end — is in none of them.
Here is the whole popular story of the dancing plague of 1518, broken into the claims it is actually made of. Each one is stamped with the earliest source that says it. Drag the control below — how recent a source will you accept? — from the contemporary Strasbourg record outward. In under ten seconds you watch the lurid parts blink out. The dancing survives. The death toll does not.
Left = only documents written in Strasbourg during the event, 1518. Drag right to admit later chronicles, then modern reconstruction.
The only number behind "into the hundreds" is one assumed rate. Pick the days and the toll is whatever you chose — because nobody counted bodies.
Accepting the contemporary 1518 record only: 4 of 8 claims stand.
The dancing, the crowd, the encourage-then-ban response, and the pilgrimage all survive. The death toll is not in the record.
Now the number the whole legend hangs on. The "hundreds dead" was never a count. It is a single rate — fifteen a day, which John Waller offered conditionally ("if such a rate were true") — multiplied by a guessed number of days. Drag the second slider and watch the "toll" become whatever you assume:
At fifteen a day you reach "100 dead" in just 7 days, and the often-quoted 400 in 27. The total is an arithmetic product of one number — not an observation. The contemporary Strasbourg sources record the dancing, the city's response, and the pilgrimage in some detail, and mention no deaths at all.
The check — every number stamped with the source that first says it
The green column is the cross-check: is this claim in the contemporary 1518 Strasbourg record, or only in a later source? Nothing here is asserted — each stamp is the earliest attesting document named in the sources below.
| the claim | earliest source | in 1518 record? |
|---|---|---|
| People danced compulsively, mid-July–Sept 1518 | contemporary: physicians, sermons, chronicles, council | yes |
| > 30 dancing within days | contemporary chronicle / council | yes |
| Authorities encouraged it, then banned it | 1518 council orders (Brant); Specklin, 16th c. | yes |
| Sufferers sent to St. Vitus, Saverne (red shoes) | contemporary council record | yes |
| Ringleader named "Frau Troffea," July 14 | Paracelsus tradition, ~1526 (hostile; disputed) | no |
| As many as 400 danced | Imlin family chronicle, 16th c. | no |
| "Many hundreds" danced | Goldmeyer 1636; poem via Schilter, 17th c. | no |
| ~15 died per day, into the hundreds | Waller, 2008–2009 (stated conditionally) | no |
The death-toll arithmetic, recomputed: 15/day × 7 = 105 ("100 dead" needs one week); 15/day × 27 = 405 ("400" needs four). Deaths in the contemporary record: 0. Outbreak span mid-July to early September ≈ 55 days (~8 weeks). An earlier outbreak struck Aachen, 1374.
Run it yourself: node research/the-dancing-plague/verify.mjs.
So what caused it? Test each answer against the evidence.
Correcting the death toll is the easy half. The hard half is the cause — and here the honest move is not to swap one confident story ("ergot!") for another ("mass hysteria, obviously"). Pick a candidate. Watch which rows of the evidence it fails to meet. Every one leaves at least a row unanswered.
| does it explain… | Ergot poisoning | Mass psychogenic illness | A saint's curse |
|---|
Tap a cause to read, row by row, where it holds and where it breaks.
Ergot fails every row — it can't drive days of coordinated dancing, can't produce a uniform mass response, predicts convulsions not dancing, and (Waller's decisive point) the outbreaks track the Rhine–Moselle belief corridor, not the rye harvest. Weakest fit — but proponents of convulsive ergotism still exist, so: not disproven, not crowned. Mass psychogenic illness meets almost every row — but one stays amber, and it is the important one: it explains the stress-trance and the contagion, yet not, on its own, why the symptom was dancing.
The one operable idea that closes the gap — without crowning a winner.
Mass psychogenic illness has no fixed symptom. It borrows the shape of whatever a community already fears. Keep the mechanism fixed — stress plus social contagion, tipping into a dissociative trance — and change only the fear. The symptom changes with it. These are all documented episodes:
the live fear St. Vitus can curse you with uncontrollable dancing
the mechanism (fixed) stress + social contagion → dissociative trance
so the symptom takes the form of DANCING — Strasbourg & the Rhineland, 1518
That is why the symptom was dancing, and why the recurrence (Aachen 1374 onward) clustered in the Rhineland — the corridor where the St. Vitus belief lived, which is exactly the geography on which the ergot theory fails. The cause is cultural, not agricultural. This is the leading reading, John Waller's; it is not proof. No cause is settled.
Documented event. Contested magnitude. Unrecorded deaths. Unresolved cause — but a defensible sense of why it took the shape it did.
What the record supports, and where the popular story runs ahead of it
Exactly true (the contemporary 1518 record)
Multiple roughly-contemporary Strasbourg sources — the city physicians' consultation, cathedral sermons, local and regional chronicles, and notes of the city council (the Twenty-One) under chancellor Sebastian Brant — agree that many townspeople danced compulsively from mid-July into September 1518; that within days more than thirty were dancing; that the authorities first encouraged it (clearing guild halls, building stages, hiring musicians and strong dancers to "dance it out") and then reversed, banning dancing and music and sending sufferers to the shrine of St. Vitus above Saverne with red shoes, small crosses and holy water. An earlier outbreak struck Aachen in 1374 and spread along the Rhine.
Only in later sources (named, not contemporary)
The ringleader's name. The contemporary record does not name the first dancer. "Frau Troffea" surfaces in the later tradition, associated with Paracelsus's account after his 1526 visit (a hostile witness), and several historians note the name's authenticity is itself disputed. The exact start date "July 14" heads the 16th-century Imlin family chronicle's narrative; its precise-day provenance is soft — "mid-July 1518" is the safe claim.
"As many as 400." The upper figure rests on one later source, the 16th-century Imlin family chronicle. The honest range is 50–400; the contemporary record supports "a number" / dozens. "Many hundreds" is 17th-century (the lost poem quoted by the jurist Johann Schilter; the 1636 Goldmeyer chronicle).
The death toll — the biggest overclaim, and the reason this page exists
The contemporary Strasbourg sources do not mention deaths at all — not a number, not even whether there were any. The famous "fifteen dying each day / into the hundreds / ~100 dead" comes from later accounts plus John Waller's own conditional extrapolation: "the final toll is unknown but, if such a daily death rate was true, could have been into the hundreds." The page shows what that total really is — one assumed rate times a chosen number of days, a product, not a tally. Do not read this as "nobody died": some sufferers may well have collapsed from exertion or heat. Read it as: no toll is documented, and the specific numbers are a modern estimate, not a record.
The cause — unresolved; no candidate is crowned
Ergot (convulsive ergotism) is the weakest fit — but not disproven. Waller's objections: its vasoconstriction and convulsions cannot sustain days of coordinated dancing; it cannot produce a uniform mass response; and the outbreaks track the Rhine/Moselle belief-corridor, not rye harvests or damp climate. Convulsive-ergotism proponents still exist. Mass psychogenic illness is the most-accepted reading — not established fact, and is itself a loose, retrospective label; it explains the trance and contagion but not, unaided, the specific form. The shape-of-fears idea (Waller's) — that MPI takes the form of a community's live fear, so a St. Vitus-fearing culture danced — is the leading interpretation of why the symptom was dancing, presented here as a reading, not a proof. The comparison episodes (convent "possession" outbreaks; the 1962 Tanganyika laughter epidemic; modern "sick-building" fainting) are real documented cases of mass psychogenic illness taking different symptomatic shapes.
Free choices in the instrument
The five "acceptance tiers" are a simplification of a messier documentary record into its source-layers (contemporary 1518 · Paracelsus ~1526 · 16th-c. chronicles · 17th-c. chronicles · modern reconstruction); the timeline positions for the two undated chronicle tiers are indicative, and no precise year is asserted for them. The evidence grid's five tests, and the met/partial/fail calls, are a compact rendering of the arguments in the sources below — the direction of each (ergot weakest; MPI best-but-incomplete) is robust; the exact cell labels are ours. The "days" slider's rate (15/day) and the durations are the figures the sources actually use.