World’s FairsFound, not built?
ResearchDoes It Add Up?
Does It Add Up?

Photo: Mclib · CC BY 2.0 · via Wikimedia Commons

Research

Does It Add Up?

Also called “Too Big for the Town”

Buildings set against the towns, workforces and timelines the record gives them, scored from sourced facts only.

The hypothesis

Where were the skilled hands? Emerging

Where the record does name the workforce, it keeps saying the same thing: the skills had to be imported (Scottish granite cutters, German and Irish masons) or forced (convicts). Yet the same decades are credited with thousands of carved stone buildings across small towns. Common sense: if Texas had to send to Scotland for 86 stonecutters, who was cutting stone for every county courthouse and capitol at the same time? One possibility is that far fewer buildings were cut new than the record implies, and many crews were finishing, repairing or refacing existing stonework.

  • State counts are now in. What were the town and county counts of masons, stonecutters and carvers in the build years (IPUMS full-count data)?
  • Where were the training yards for the carvers of 250 City Hall sculptures?

See the evidence › · 1 more hypothesis use this section

Hidden history

What the record shows

Some buildings are far bigger, finer or faster than the record says the place could manage at the time: a hospital for hundreds with room for thousands, a capitol in a frontier town, a downtown rebuilt in stone within months of a fire. Each entry here is scored from facts in the linked sources, factor by factor, so you can see exactly where the numbers stop adding up.

  1. Seven factors

    Population vs size, masons, artisans, materials and transport, build time, gaps in the records, and physical clues. Each factor scores 0 to 10 from sourced inputs only.

  2. Out of 100

    The score is ten times the average of the factors that have sourced inputs. Factors with nothing on record are left out, never guessed.

  3. From N factors

    The factor count shows how much the score rests on. One factor is a hint; five is a pattern.

Questions worth asking

  • Who was each building really sized for?
  • Where did the skilled hands come from, and where are their names?
  • Why are the records thinnest exactly where the work was biggest?

Ranked by score 52

Highest score first. Tap a gauge to see every factor and the sourced facts behind each point; the factor count shows how much each score rests on.

#37Augusta · 1832Maine State HouseAbout 3,980 people in 1830, and a building costing $80,000,. Who in a town that size was paying for this, and who was building it?The building was completed in 1832, one year after Augusta became the capital of Maine. The cost of the building was originally estimated to be $80,000, but when completed, expenditures of the building, furnishings and grounds amounted to $135,000, of which $11,500 was contributed by the city of Augusta. Construction was of granite from Hallowell quarries and took three years to complete.
#47Carson City · 1871Nevada State CapitolAbout 3,042 people in 1870, and a building costing $84,000. Who in a town that size was paying for this, and who was building it?The building was constructed in the Neoclassical Italianate style between 1869 and 1871. The Board of Capitol Commissioners received bids of $84,000 to $160,000 for construction and they chose the lowest bid, submitted by Peter Cavanaugh and Son of Carson City. To reduce costs, the building sandstone was obtained free of charge from the Nevada State Prison quarry, just outside Carson City.

How this score works

A transparent count of the gaps between what the record says was built and what the record says was available to build it. It is not a probability and not a verdict. Every point comes from a sourced input listed under the factor; a factor with no sourced inputs is left null and does not count.

Each factor scores 0 to 10 (0 = the record adds up, 10 = a large, documented gap).

overall = round(10 × average of the non-null factor scores), 0–100.

confidence = number of non-null factors that went into the score (1–7). A score from one factor is a hint; a score from five is a pattern.

Population vs size
Floor area per intended occupant (sq ft ÷ design capacity): under 300 = 1; 300–700 = 3; 700–1,500 = 6; over 1,500 = 8. Or, if the record gives opening occupants and design capacity: capacity ÷ opening occupants under 2 = 1; 2–5 = 4; 5–10 = 6; over 10 = 8. Add 2 (max 10) if the building's capacity is over 10% of the town's recorded population at the time. Also, only when the record gives the numbers: (a) Asylums: peak or design patients ÷ the host town's census population nearest the opening year: under 5% = 1; 5–15% = 3; 15–30% = 5; 30–60% = 7; over 60% = 9. (b) Civic buildings (capitols, courthouses, cathedrals, opera houses): recorded construction cost ÷ the town's census population nearest completion = dollars per resident (not adjusted for inflation): under $10 = 1; $10–50 = 3; $50–150 = 5; $150–500 = 7; over $500 = 9. Planned, estimated or 'expected' costs are not used.
Masons needed vs available
+4 if masons or stonecutters are recorded as brought in from another country; +3 if convict or forced labour is recorded; floor area per recorded mason: 1,500–3,000 sq ft = +3, over 3,000 = +5. Max 10. 0 if the record shows an adequate local workforce.
Artisans needed vs available
+4 if carvers/sculptors/designers are recorded as imported; +3 if fine detail is recorded but no carvers are named; +3 if the detail designer is recorded as working under someone else's name. Max 10.
Materials and transport
+2 if a rail spur or special vessel had to be built to bring stone in; +3 if a single block over 100 tonnes was moved; +3 if stone came from over 300 km away; +2 if over 1,000 tonnes of material per year of build is implied. Max 10. 0 if local stone and existing rail are recorded. Quarry distance is a straight line between the building's and the named quarry town's coordinates, and both sources are listed.
Build time vs scale
Floor area per build year: under 10,000 sq ft = 1; 10,000–30,000 = 3; 30,000–60,000 = 5; 60,000–120,000 = 7; over 120,000 = 9. For a district rebuild: masonry buildings finished within two years of a fire: under 10 = 2; 10–30 = 5; over 30 = 7. For a single heavy lift: +4 if a monolith over 500 t was raised in under a day.
Gaps in the records
+3 builder or contractor not recorded; +3 credited architect disputed or designs filed under another name; +2 plans or drawings recorded as lost; +2 payroll or cost not recorded, or records recorded as destroyed by fire. Max 10.
Physical clues
+4 windows, doors or a floor now below grade are documented; +3 ground level around it documented as raised 1 m or more; +3 documented as buried and dug out. Max 10.

What the census says

Counts of carpenters, masons, stonecutters, plasterers and iron workers, read from the printed US Census occupation tables. State counts cover the whole state, not the town.

Where it cuts against the hypothesis

  • Strong by count (31) once the extra asylum-and-town pairs from the second research batch are added, but kept at moderate: the census occupation counts show large state-wide trade pools, which weakens the 'no one to build it' part of the mismatch. The size-versus-town gap itself stands.
  • State census counts show sizeable trade pools in most pair states (e.g. Kansas 2,999 masons and stonecutters in 1880 while Topeka's capitol was going up; Michigan 4,412 in 1880 for Lansing and Traverse City). A state-funded building could draw on those pools, so a small town population on its own is not a labour gap.

Crews and the census

  • Texas State CapitolAustin
    • Texas, 1880: 1,668 brick and stone masons, marble and stone cutters in the whole state (1880 Table XXXIV, printed p. 847). 86 imported cutters = 5.2% of that count.
    • Texas, 1900: 426 marble and stone cutters statewide (1900 Table 32). 86 = 20.2% of the 1900 cutters, twelve years after the capitol was finished.

    Texas did have stoneworkers (1,668 masons and cutters in 1880), yet the builders still brought 86 cutters from Scotland. That fits a shortage of granite cutters specifically, not of masons in general.

  • Biltmore HouseAsheville, North Carolina
    • North Carolina, 1900: 1,756 masons (brick and stone) and 415 marble and stone cutters statewide (1900 Table 32). 60 stonemasons = 3.4% of the state's 1900 masons. (1890 North Carolina figures not sourced.)
    • Biltmore is Indiana limestone, not brick, so no brick-rate estimate is made.

    A crew of 60 is a small share of the state's masons.

  • Trans-Allegheny Lunatic AsylumWeston, West Virginia
    • West Virginia, 1880: 1,059 brick and stone masons, marble and stone cutters statewide; Weston itself had 1,516 people (1880 census, https://en.wikipedia.org/wiki/Weston,_West_Virginia). Weston town counts not sourced.
    • West Virginia, 1900: 1,626 masons and 499 marble and stone cutters statewide.

    No crew size is recorded, so no ratio can be made.

  • Trans-Mississippi and International ExpositionOmaha 1898
    • Omaha city, 1900 (males): 1,364 carpenters and joiners, 423 masons (brick and stone), 139 plasterers, 94 marble and stone cutters, 244 iron and steel workers (1900 Table 43, printed pp. 664 and 666).
    • Nebraska, 1900: 7,329 carpenters, 1,701 masons, 692 plasterers statewide.

    The fair palaces were staff (plaster and fibre) over frames, so carpenters and plasterers are the relevant trades. Without a recorded crew size there is nothing to divide.

31 census counts, by state and city
What the census volumes could not tell us

  • Town- or county-level counts for the small towns (Weston, Athens, Fergus Falls, Traverse City, Ovid/Willard, Napa, Salem OR, Milledgeville, Lansing, Austin, Jackson, Boise, Olympia, Helena, Frankfort, Jefferson City, Leadville, Asheville) in any year 1850–1900

    The printed census volumes only break occupations down by city for the largest cities (1870 and 1880: 50 principal cities; 1890: cities of 25,000+ in Table 118; 1900: cities of 25,000+ in Table 42). None of these towns was big enough. County counts would need IPUMS full-count microdata, which needs an account, so they are left blank.

  • 1850, 1860 and 1870 state counts

    The 1870 census.gov PDFs we downloaded (1870a-34 to 40) have no text layer and were not read page by page in this pass; 1850 and 1860 volumes were not downloaded. Left blank.

  • 1890 state and city counts

    The 1890 Population Part II volumes were downloaded, but their state occupation tables did not come through in the text layer and were not read from images in this pass. Table 118 (cities of 50,000+) is in section 5; of the pair towns only Omaha and Lincoln were that size in 1890, and the table was not read from images in this pass. Left blank.

  • Seattle 1900 city counts; Springfield 1900 carpenters

    Too faint in the census.gov scan to read every digit (see 'legibility').

  • Stone carvers and bricklayers as separate trades

    Neither is its own line in these tables: bricklayers fall inside 'Masons (brick and stone)', and carvers are not split out from 'Marble and stone cutters'. Left blank rather than guessed.

  • UK and Canada equivalents (e.g. Holloway Sanatorium, Colney Hatch, Ottawa, Victoria BC)

    Not sourced in this pass. Left blank.

Where it happened 42 places

Open the globe ›

Land: Natural Earth (public domain).

Patterns

The same story turns up in other threads.

Hypotheses20 patterns across all the threadsClues Others Missed110 details in the photosMud FloodBuried first floors, raised streets and what lies under the citiesAsylumsKirkbride palaces and why admissions soaredOrphan TrainsAbout 200,000 children sent west, 1854–1929Child Labor & RepopulationBreaker boys, mill girls and the 1850–1920 windowBurned and Rebuilt?Ruins, then palaces: swipe between the two photosSame Story, Different CityFire, temporary, rebuilt: the same beats in city after cityArchitects of RecordWho got the credit, how old they were, and how long it tookCatacombs & UndergroundCatacombs, ossuaries, underground cities and forgotten tunnels under the cities