Christopher Alexander

A city is not a tree

Notes on the Synthesis of Form (Harvard University Press, 1964) starts from an observation about villages that nobody planned: their forms fit the lives lived in them because they changed a little at a time, each repair answering a particular misfit. A modern designer, working all at once from categories he brings with him, tends to produce arbitrary shapes. Alexander’s remedy was to state the problem, set aside inherited concepts, build new ones from the problem’s own structure, treat each resulting subsystem as a subproblem and turn each into form. Software people read it early; it was being cited in computer science from the late 1960s. Alexander did not extend the formal machinery himself. Commentators point to its mathematical difficulty and to Horst Rittel’s argument that design problems are “wicked”; that is their explanation, not his.

“A City Is Not a Tree,” printed in Architectural Forum in April and May 1965, turned the same instinct on whole cities. Alexander set natural cities that had grown (Siena, Liverpool, Kyoto, Manhattan) against planned ones (Levittown, Chandigarh, the British new towns) and gave the difference a shape. In a tree, any two groups of things are either nested or wholly separate. In a semilattice, groups overlap. His example is a corner in Berkeley, Hearst and Euclid: a drugstore, a newsrack, a traffic light and a stretch of pavement. People waiting at the light read the papers, so the newsrack belongs both to the light-and-sidewalk unit and to the drugstore’s. He counted what the tree gives up: on twenty elements it allows at most nineteen further groupings, where a semilattice allows over a million. He read Columbia in Maryland, Greenbelt, Le Corbusier’s Chandigarh, Lucio Costa’s Brasília and Kenzo Tange’s Tokyo plan as trees, suggested planners draw them because minds find overlap hard to hold, and ended by calling the city “a receptacle for life.”

In 1967 he founded the Center for Environmental Structure, the small office that would write his books and, later, build his buildings.

How to write a pattern

A Pattern Language (Oxford University Press, 1977) is credited to Alexander, Sara Ishikawa and Murray Silverstein of the Center, with Max Jacobson, Ingrid Fiksdahl-King and Shlomo Angel. Its 253 patterns run from regions and towns down to the way a window sits in a wall, and each is written to the same form. Take number 159, Light on Two Sides of Every Room. It opens by saying where it falls in the sequence: the major rooms are placed and the building’s actual shape is being fixed, which completes Wings of Light (107) after Positive Outdoor Space (106) and Long Thin House (109). The problem is an observation: “When they have a choice, people will always gravitate to those rooms which have light on two sides.” The evidence is informal. The authors say they checked buildings and asked friends in offices and homes, and the mechanism they offer is glare, which in a one-sided room hangs around people’s faces and makes them harder to read. They name Le Corbusier’s Marseilles block as a building that ignores the pattern. Then comes the solution, after “Therefore:” — “Locate each room so that it has outdoor space outside it on at least two sides” — with windows in those walls so that daylight arrives from more than one direction. A last paragraph hands the designer on to smaller patterns: Window Place (180), Deep Reveals (223), Filtered Light (238). The authors present each pattern as a hypothesis for later builders to test and revise, not a rule.

The same office produced The Oregon Experiment (1975; Alexander with Murray Silverstein, Shlomo Angel, Sara Ishikawa and Denny Abrams), a master plan for the University of Oregon built on six principles: organic order, participation, piecemeal growth, patterns, diagnosis and coordination. The Timeless Way of Building (1979) is the philosophical companion, arguing for a “quality without a name” in places people love. A pattern is not a style. The authors compared the language to English, which, in the words Andrea Torreblanca quotes from the introduction, “can be a medium for prose, or a medium for poetry.”

Five families, a builders’ yard and a block press

In 1975 Julio Martínez, director of public works for the state of Baja California, invited Alexander to Mexicali to talk about the Center. By Andrea Torreblanca’s account Alexander proposed building instead, and the state agreed. Contracts followed in the Center’s archive: professional services with the Universidad Autónoma de Baja California on 30 August 1975, a general agreement on 13 September among the Center, the state social-security institute ISSSTECALI, the state real-estate agency, the state government and the university, and a land loan on 9 February 1976. Martínez reappears as a co-author, with Howard Davis and Donald Corner, of the book that records it, The Production of Houses (1985). The plan was thirty low-cost houses at about $3,500 each. Five were built, for families who had answered an ISSSTECALI announcement: government employees earning no more than 5,000 pesos a month, married, with at least two children, owning no house, and willing to give their free time and labour.

The method was the book’s, performed on a plot of dirt. Starting from two patterns, Common Land and House Cluster, the families and the Center’s staff laid out the cluster themselves with stakes, blocks and chalk lines, and each house, about 60 square metres, was designed with its family while it stood marked out on the ground. Alexander worked as what he called an architect-builder: the Center managed construction directly, made design decisions on site, and held the budget and the subcontractors, so that its archive compares estimated, extra and actual costs for five houses across 24 operations. A builders’ yard across the street held tools, prototypes and the communal meals. The walls were earth-cement interlocking blocks pressed on site in four shapes with an imported Italian machine, the roofs basket vaults of wooden strips and chicken wire under pumice concrete, and the colour took two weeks of trials until the whitewash, faintly tinged with green, looked right.

Accounts of the ending disagree on dates. The houses were finished in 1976 according to Torreblanca and in 1977 according to the Center’s archive; the Center’s own project page says 1987, which the contracts contradict. They agree that the other twenty-five were never built. Torreblanca writes that the authorities found the process slow and the handmade houses not “new”, and abandoned it. Later critics read the site harder. Felipe Orensanz and Alejandro Peimbert report that the common land Alexander’s team had “played a major role in defining” was split into five private backyards, that the vaults made adding a storey hard (the Reyes family built a metal frame around their house), and that the experimental blocks did not match the materials sold locally. They call the eligibility rules “nonnegotiable”, argue that Alexander’s alternative to state and market housing was itself a hierarchical “new social system”, and say he treated Mexicali as a wasteland instead of a place with its own building knowledge. Reyner Banham’s 1986 review suspected “some hidden agenda over and above the provision of better housing for urban Mexicans”, and found it in “the care for beauty.” The builders’ yard, El Sitio, still stands, and Torreblanca (2023) says it is now a clinic for the university’s nursing faculty.

Eishin: a Tokyo campus and the fight over who builds

The largest job the Center ever took was a school. Hisae Hosoi, managing director of the Eishin Gakuen foundation (Higashino High School), commissioned a high school and college campus near Tokyo. The Center’s archive dates the project 1982–1995, with the first and largest part finished in 1985; Alexander’s book says building began in 1981 and was largely done by 1990, with seven buildings never built. Counts differ too, 35 buildings in the archive and 36 in a published review of the book. Alexander was chief architect, Hajo Neis executive architect, Gary Black structural engineer, with Ingrid Fiksdahl-King and Artemis Anninou. The patterns came from interviews, about 1,200 hours with teachers and staff by the archive’s count, condensed into a written language of 110 patterns. The buildings were then placed on the ground with stakes and ropes (six-foot flags, in the book’s telling) and checked against a large site model that users could review.

Decisions on site were the point. The Center kept a full-scale office on the campus, built full-size mock-ups of key materials and details, and changed buildings as they rose; its archive says that “because of our method of cost accounting, these changes were made without increasing the cost of the finished project.” One reviewer, David Creelman, describes the practice as widening a planned two-metre entrance, or ornamenting a wall that did not feel right, because the architect is also the builder.

The contractor was Fujita, which the archive calls the sixth-largest construction company in Japan, and the relationship left a paper trail of grievance. The Center’s files include a 1984 letter complaining of “unexpected and harsh behavior” and threats, a 1985 note that the situation “could require legal intervention”, an attempt to cancel the Great Hall work over money, and requests for unpaid architectural fees running from 1985 to 1998. A 1985 Nikkei Architecture article by Toshihiko Shiraishi bore the title “Large Deviation from the Reality of Japan Revealing the Difficulty of the Directly Management Method.” In 1990 Alexander objected that a documentary film called him a “philosophical advisor” instead of the architect. His own account, The Battle for the Life and Beauty of the Earth (2012, with Hajo Neis and Maggie Moore Alexander), frames the contest as System A, building for life, against System B, which prizes “size, speed, profit, efficiency”. As the reviewer David Seamon summarises it, the book tells of lawsuits, under-the-table payments, betrayals by colleagues, threats to the families of faculty, a physical beating of the main client by thugs hired by the project’s opponents, and an arm-wrestling bout with Fujita’s foreman. We found no independent account confirming those episodes. Seamon, who found the campus vibrant, noted that the story comes only from Alexander and that the book’s warlike imagery sits oddly with its gentler philosophy. A 2009 Nikkei Architecture piece, he reports, called the campus a “great work.”