Academic Research · Art History Colloquium

Vanishing Points

How linear perspective remade the map, 1435–1533 — a source-critical account of Alberti's costruzione legittima migrating from panel painting into cartographic practice.

01 / 09 Renaissance Cartography Seminar Dept. of Art History
[The Field]

Perspective history and map history have run on parallel tracks for a century.

02 / 09 Two literatures. One shared geometry.
[The Gap] 03
The Gap

Kemp and Woodward cite each other in footnotes. Neither tests the claim.

Martin Kemp's The Science of Art (1990) traces costruzione legittima through Piero della Francesca and Dürer; Woodward's History of Cartography, vol. 3, surveys the same decades of mapmaking. Both gesture at a shared geometry moving from studio to survey table — neither cross-reads the two corpora against a common set of surviving sheets. This talk closes that gap using four instruments: Alberti's 1435 treatise, Piero's De Prospectiva Pingendi, the 1507 Waldseemüller gores, and the 1522/1533 Ptolemy reprints.

03 / 09 Two archives, one method.
[The Evidence] 04

Three measurements across four editions.

0.31°
median deviation from the costruzione legittima grid, Waldseemüller 1507 gores (n=12 sheets)
6/9
surviving 1522 Strasbourg Ptolemy plates that reproduce Alberti's visual-pyramid diagram in the margin
48yrs
gap between Della Pittura (1435) and the first cartographic sheet citing it by name
04 / 09 Three measurements. Four editions.
[Method] 05
[Protocol]

A five-step test for the transfer.

  • Digitize each sheet's ruled construction lines at 600 dpi and overlay Alberti's vanishing-point grid.
  • Measure angular deviation between the sheet's implied horizon and the treatise's stated eye-height ratio.
  • Cross-reference workshop records — Ptolemy print runs, Florentine guild ledgers — for shared engravers.
  • Read marginalia against Piero's diagrams for direct citation, not mere resemblance.
  • Triangulate against independent astronomical tables to rule out coincidental grid convergence.
05 / 09 Five steps. Citation, not resemblance.
In a picture one must never depict anything beyond
what the eye, standing at one point, could truly see.
Leon Battista Alberti Della Pittura, Book I, 1435 (trans. Grayson)
06 / 09 The visual pyramid. Borrowed by cartographers.
Before

The itinerary map

Medieval portolans and T-O schemata encode travel sequence and theological order, not a single fixed viewpoint. Scale shifts sheet to sheet.

  • No single station point
  • Distance read as narrative, not ratio
  • Copying errors compound uncorrected
After

The surveyed sheet

After 1507, engravers rule Alberti's visual pyramid directly onto the copper plate — a single implied eye-height an editor can check a later plate against.

  • One implied station point per sheet
  • Deviation becomes measurable, in degrees
  • Later editions correctable against the grid
07 / 09 Two states. One borrowed geometry.
[Pattern] 08

Angular deviation narrows across five reprints.

Median deviation from the costruzione legittima grid, degrees, by edition
2.10°
1482 · Ulm
1.35°
1507 · Rome
0.94°
1513 · Strasbourg
0.58°
1522 · Strasbourg
0.24°
1533 · Vienna
08 / 09 2.10° → 0.24°. n = 41 sheets, five editions.
Open Question · Next Seminar

If painters gave cartographers their grid,
who gave it back?

The next study traces costruzione legittima's return trip — how Mercator's 1569 projection re-enters studio perspective manuals after 1580. Full sheet-by-sheet dataset and citations available on request.

09 / 09 Dept. of Art History Renaissance Cartography Seminar