World’s FairsFound, not built?
HypothesesFire as the reset button

Hypothesis #6 of 20

Fire as the reset button

Strong53 supporting entriesUpdated (was strong 36)

Strength is how much of our gathered material lines up with the pattern. It is not the odds that the idea is true.

The working idea

City after city loses its old centre to one big fire, usually blamed on something homely like a pipe, a glue pot or a cow, and within one to three years a grander stone and brick centre stands in its place, often on a brand-new street plan that was ready almost at once. Common sense says rebuilding a downtown in stone takes money, masons and quarries a city doesn't have the morning after a disaster. One possibility is that some of these fires cleared old fabric and old ownership, and the rebuild was planned or supplied from outside.

Strongest examples

Great Chelsea FireReports say buildings of solid granite crumbled in a town fire.Four fires in one hourAtlanta 1917: the big fire was the fourth call in about an hour.Insurers saw it comingUnderwriters mapped Bangor's fire path in 1907; it burned that way in 1911.Charleston earthquake81% of the 1886 damage was to brick buildings: brick failed where wood stood.

The evidence

Distinct city fire, quake-fire and records-fire events in the dataset (now including the second research batch)

74 events; 53 of them fall between 1835 and 1922 (the 36 before plus 17 new). 1906 San Francisco is now counted here too.

53 entries

Events per decade (overlay 'fires' series)

1840s 5, 1860s 4, 1870s 8, 1880s 10, 1900s 10, 1910s 8, 1920s 4; every other decade 0–3

Same-season clusters

Washington 1889: Seattle 6 Jun, Ellensburg 4 Jul, Spokane 4 Aug (three fires in two months). 1904: Toronto 19 Apr, Ålesund 23 Jan, Baltimore 7–8 Feb. 1917: Atlanta 21 May, Thessaloniki. 1871: Chicago Oct, Paris May.

8 entries

Rebuild or new plan on record within about 2 years

9

9 entries

Homely causes on record

stable pipe (St. John's), glue pot (Seattle), cow and lantern (Chicago), cigar factory (Hamburg), mattress factory (Jacksonville), fireworks (Portland, Maine)

6 entries

Fires followed by a world's fair in the same city

Chicago 1871→1893 (22 yrs), San Francisco 1906→1915 (9), Seattle 1889→1909 (20), Paris 1871→1878 (7), St. Louis 1849→1904 (55), Atlanta burned 1864→1895 (31)

6 entries

New in the second batch: same-year clusters

1871: Chicago (8–10 Oct per its row), Peshtigo and the Great Michigan Fire (both dated 1871 in their rows), plus Paris in May. 1888: Sundsvall and Umeå (Sweden). 1889: Seattle, Spokane, Ellensburg, Bakersfield. 1906: San Francisco and Valparaíso.

7 entries

Open questions

  • Who owned the burned land before and after each fire?
  • How were new grid plans (Turku, Lisbon, Thessaloniki) ready within months?
  • Where did the brick, stone and masons for dozens of buildings in 18 months (Seattle) come from?

What would support it

  • Land-title records showing large ownership changes right after the fires.
  • Rebuild plans dated before the fire.
  • Quarry and brickyard output that can't cover the rebuild.

What would rule it out

  • Ordinary insurance payouts, bank loans and local brickyard records that fully cover each rebuild.
  • Fire-science reconstructions matching each spread (dense timber, wind, broken mains).
  • A full list of all city fires 1600–2000. Round 2 adds six fires dated 1600–1694 (Edo about 1600, Stockholm 1625, Meireki 1657, Northampton 1675, Trondheim 1681, Warwick 1694) and six from 1700–1834 (Tiverton 1731, Stockholm 1759, New Orleans 1788, 1805, Savannah 1820, Edinburgh 1824), so big city fires are not confined to 1835–1922; the cluster may simply be the era of dense wooden downtowns.

Selection bias

Medium: fires were gathered for the 'Burned and Rebuilt' and 'Same Story' sections, so 1835–1922 was searched hardest. Round 2 widened the search, and 22 of its 39 rows fall outside 1835–1922, have no year, or repeat events already counted.

The rating counts how much of our gathered material lines up with the pattern. It is not a probability that the hypothesis is true, and every count is limited by which buildings were picked for the app (see 'selectionBias' on each card).

Next hypothesis

Strong36 entries#7The names on the plaqueThe credited architects include men in their early twenties handed city halls and parliaments, offices running several capitols at the same time, a Gothic genius who had to file his designs under other men's names, and competitions won by the fourth- or third-place entry. Common sense says the name on the plaque may not be the person who drew the building. One possibility is that some credited architects were administrators or certifiers of designs that came from somewhere else.

← All hypotheses