Skip to content

Method & sources

What the entries try to do

Each entry aims to be read in ten minutes, with enough information to:

  1. Locate the project in time, space, and the career of its architect.
  2. Describe what the project was meant to be, including approximate scale.
  3. Explain why it did not get built, or — in the few partially built cases — why it stopped.
  4. Offer a short counterfactual: what would it have meant, for its city and for the discipline, had the building gone up.

The fourth section is interpretive and is marked as such in the prose. The first three stay close to the shared historical record.

Sources

The catalog draws on the standard secondary literature. For a single source that covers most of the cases here, the best starting point is:

  • Wilkinson, Philip. Phantom Architecture. Simon & Schuster, 2017.

For individual projects, the following were consulted either directly or via the references cited in them:

  • Boullée — Jean-Marie Pérouse de Montclos, Étienne-Louis Boullée, 1969; and Boullée's own Architecture, essai sur l'art (posthumous).
  • Ledoux — Anthony Vidler, Claude-Nicolas Ledoux, 1990.
  • Tatlin — Camilla Gray, The Russian Experiment in Art, 1971; and the Christina Lodder literature on Constructivism.
  • Le Corbusier — Urbanisme (1925) for the primary-source presentation of Plan Voisin.
  • Wright — Neil Levine, The Architecture of Frank Lloyd Wright, 1996, for Broadacre and Mile-High.
  • Iofan & Speer — Sudjic and Jones, Architecture and Democracy (2001); Alex Scobie, Hitler's State Architecture (1990).
  • Archigram — Peter Cook (ed.), Archigram, 1972, reissued 1999.
  • Superstudio — Peter Lang and William Menking, Superstudio: Life Without Objects, 2003.
  • Hadid — Aaron Betsky and Patrik Schumacher, Zaha Hadid: The Complete Works, 2009.

Where dimensions are cited ("450 m," "2,004 m," "528 floors"), the figures are the ones the architects or their offices gave in their publications and competition submissions. They are usually the figures that appear on the drawings themselves. In a few cases the literature gives a range; where this happens the entry picks a widely-cited number and does not hedge.

What the catalog does not do

  • It does not reconstruct the buildings in three dimensions. Direction #4 in the project ideas page sketches what that would look like.
  • It does not attempt a complete bibliography per entry. The full references for any individual project run to dozens of items; that is a librarian's task.
  • It does not include unbuilt infrastructure (bridges, dams, tunnels) or unbuilt urban plans without architectural content (traffic schemes, zoning maps). The subject is the building as phantom.

Illustrations

The SVG illustrations are generated programmatically from scripts/generate_illustrations.py. Each is a silhouette in a unified blueprint style: warm cream paper, near-black ink, a ground line, a 1.8-metre human figure for scale where one is useful. The style is inspired by the line-drawings of Robert Venturi and Denise Scott Brown, and by the plates in Ackerman's Palladio — clean, diagrammatic, willing to sacrifice accuracy for legibility.

Where the original architect's drawings are famous in their own right (Bruegel's painting, Boullée's plate, Ferriss's charcoals, Hadid's paintings), the reader should seek out those sources. Wikipedia's Commons has public-domain reproductions of most.

Attribution

All prose is original to this site. Historical facts and direct quotations are drawn from the sources above. The illustrations are original. The SVG generator is MIT-licensed (see LICENSE).

If you spot an error — a wrong date, a misattribution, a false fact — open an issue at github.com/5x5x5x5/phantom-architecture.