Architecture of Repair

Although Christopher Alexander, who died this year on March 17, was officially an architect, the significance of his life lay in the challenge he posed to architecture. In a sense, he did not believe that architects were necessary. Put a small group of people on a building site, give them materials and colors, and let them work through the endless factors—spatial, visual, tactile, olfactory—that go into the design of a house, and they will invariably produce something more attuned to their human needs than any architect sitting at his drafting table ever could.

These ideas were codified in a series of books, of which A Pattern Language (1977) was the most influential and the most extraordinary. Although compiled by a team of authors, A Pattern Language is essentially Alexander’s book, and yet in a sense it is scarcely a book at all. It comprises some 253 “patterns,” each forming a case study of “a problem which occurs over and over again in our environment,” although they are almost never recognized as the same generic problem.

For example, both a bus stop and the waiting room in a pediatric clinic are “A Place to Wait” (pattern no. 150), and waiting is an activity that always causes our spirits to sag. To counter this, ­Alexander proposed, people should be able to spend their waiting hours as fully as the rest of their day, which means that waiting should be fused with some other positive activity. The place to wait should also offer a protected quiet area, to offset the anxiety of waiting.

Alexander illustrated each pattern with a diagrammatic sketch that “describes the core of the solution to that problem, in such a way that you can use this solution a million times over.” His patterns began with the largest elements of planning, such as “the distribution of towns” (no. 2) and “parallel roads” (no. 5), and progressed step by step to cover the minutest aspects of house design, such as the “interior window” (no. 94) and the “half open wall” (no. 193). Each sketch was accompanied by several pages of discussion, informed by scientific research and Alexander’s own analysis, which could be charmingly opinionated. For example, he insisted on the practice of “Sleeping to the East” (no. 138), and for dissenters who objected to being woken by the sun in their eyes, he had no patience: “We believe there may be fundamental biological matters at stake here and that no one who once understands them will want to ignore them.”

And so it went, on and on, for almost a quarter of a million words. Few readers, even Alexander’s most stalwart defenders, will have read the whole book, and those who have will have done so only in installments. One is about as likely to read the Joy of Cooking straight through. Yet A Pattern ­Language became an unlikely cult book for two entirely different audiences. Architecture students embraced it for its spirit of dissent, its je m’en fous cheekiness, and computer scientists recognized its potential for the new field of software design. All this conferred a curious prophet-pariah ­status on Alexander, an odd fate for a quiet introvert who taught architecture and had a modest private ­practice.

Christopher Wolfgang Alexander was born in Vienna in 1936, the only child of two classical archaeologists. His mother was Jewish, a fact that forced the family to flee in 1938, the year of the Anschluss. They made their way to England, where out of necessity (the excavation sites of classical antiquity now barred to them) they reinvented themselves as teachers. Learning English from the age of two, Alexander was in every respect an Englishman, and yet in some fundamental sense he was an outsider. In English society he would be what anthropologists call an observer-participant, a functioning member with the inner detachment of an outsider. This position, along with his polyglot background, allowed him to move easily between cultures.

Alexander was a mathematical prodigy and was encouraged to cultivate his gifts. He took a mathematics degree at Cambridge (to please his father, he claimed), followed by a second degree in architecture. In 1958 he came to the United States to study at Harvard, where he received the first PhD in architecture it offered. After graduating in 1962, he went to India and built a small school, an ­unconventional introduction to the practice of architecture and the first sign that Alexander would not be a conventional architect.

Alexander’s adviser at Harvard was Serge ­Chermayeff, the Russian-born modern architect. Together they published Community and Privacy (1963), an ambitious attack on the American suburb and its false promise of a private house in the country (“The view from the picture window is the other man’s picture window”). This book brought Alexander to the attention of William Wurster, the California architect who had recently established a College of Environmental Design at Berkeley. Wurster believed in hiring interesting minds and he recruited Alexander, who would spend his entire professional career at Berkeley.

Alexander’s Harvard thesis was published in 1964 as Notes on the Synthesis of Form, and it contained in embryonic form the idea of pattern language. It was a curious book, which attempted to explain the process of design entirely in the abstract language of mathematics. To design even the simplest of objects entailed reconciling conflicting ­criteria; for example, one desired the best materials, but these were also the most expensive. The more complex the object to be designed, the greater the number of criteria to be satisfied, and the objects required by modern life—moon bases, villages, tea kettles (a characteristically ­Alexandrian list)—presented “insoluble levels of complexity” that could not be resolved by any intuitive method.

His answer was to treat the design process as if it were a complex mathematical equation, which one solves by breaking it into discrete parts. All the requirements that a design should fulfill could be considered as an abstract web of interlocking relationships, which could be expressed in terms of a diagram:

The idea of a diagram, or pattern, is very simple. It is an abstract pattern of physical relationships which resolves a small system of interacting and conflicting forces, and is independent of all other forces, and of all other possible diagrams.

Alexander illustrated his method with a case study of an agricultural village in India with a population of six hundred, to be reorganized to suit present and future demands. He drew up a list of all the variables the village had to fulfill, ranging from fixed strictures (no. 3, house doors may not face the south) to open-ended requirements (no. 28, provision for daily bath, segregated by sex, caste, and age). Having drawn up a list of 141 separate variables, he took each one in turn and systematically tabulated each of the other variables it interacted with. Step by step, he isolated clusters of variables that had a dynamic relationship with one another, but much weaker relationships with other variables. For example, the need to regulate animal traffic (no. 94) might come into conflict with the belief that cattle are sacred (no. 7).

A heady optimism pervades Notes on the Synthesis of Form, a belief that any design problem might be solved by the proper application of mathematics and modern technology. (Its last paragraph points out that Alexander’s mathematical model was programmed for the brand-new IBM 7090 computer.) But after publishing the book, Alexander recognized that he did not need the complicated mathematics model. Once it had been demonstrated that design problems broke themselves down into discrete parts or patterns, one could solve them piecemeal, simply by recognizing the pattern of reciprocal forces that was at play, and by drawing on one’s own experience with design, materials, and construction. Of course, if this was true, one scarcely needed architects at all. It is understandable why some of them bristled when the ramifications of Alexander’s ideas began to sink in.

Notes on the Synthesis of Form showed no great reverence for modern architecture, and it took passing swipes at such landmarks of modernism as Le Corbusier’s Villa Savoye and Mies van der Rohe’s flood-prone Farnsworth House. By 1965 Alexander was confronting modernism head-on in an extraordinary essay for the Architectural Forum with the arresting title “A City Is Not a Tree.” The tree in question was not botanical but mathematical, referring to an abstract diagram in which many units branch off from a main trunk but are not related to one another. This was his way of distinguishing modern city planning, with its neatly ordered subdivisions and separation of work and housing, from the traditional city, with its interconnected and overlapping functions (a “semi-lattice” rather than a tree).

This was much the same insight as Jane Jacobs’s staggeringly influential The Death and Life of Great American Cities (1961), which affected Alexander profoundly. Jacobs had shown that traditional cities have a rich and complex metabolism, and their decentralized decision-making generates its own system of order. For Alexander, as a practicing architect, the challenge was to create something like this system of order without waiting decades for it to emerge organically. In 1971 he was given the chance to do it on a large scale, when the University of Oregon engaged him to draw up a master plan.

Of course a master plan was the last thing ­Alexander wanted to make. Master plans, he felt, were inflexible and rigid, and they hobbled future growth like the wires and clamps of a bonsai plant. Instead, the campus should grow by accretion, with small-scale additions. All changes should be ­initiated by users of the buildings, never by the administration. The campus should be a palimpsest, its building “never torn down, never erased: instead they are always embellished, modified, reduced, enlarged, improved.” The goal was to create as successful and satisfying a campus as that of the University of Cambridge. “Who designed Cambridge?” he liked to ask. “No one, and that’s the point.”

“The Oregon Experiment,” as Alexander called it, brought him international attention, and further opportunities to experiment with user-designed architecture. Architecture, he believed, should not be an arcane priesthood into which one has to be initiated, and whose secrets must be jealously guarded. He believed the elements of design were logical, comprehensible, and self-evident, and that users could work them out on site and in conversation, if given proper materials. In 1975 the Mexican government engaged him to build thirty houses in the town of ­Mexicali, each to cost $3,500, a figure he met by letting the owners do much of the work themselves.

Alexander took particular delight in “unselfconscious building,” a term he applied to the vernacular architecture of traditional societies, but also to the forms that emerged spontaneously when the educated layman was given the chance to design. Having worked in India, he had respect for what the intelligent layman might do. But this could only be realized on site, and not from a remote drafting table. Clients who had commissioned a house in San Anselmo must have been surprised when he appeared on site with strips of butcher paper painted in various colors to see how they would respond to the ambient light and to work out the most pleasing ratio (which turned out to be 51 percent sea green, 35 percent red, and 14 percent yellow). His own buildings had something of this same rambling improvisational character, in a kind of playful vernacular.

Alexander’s anti-establishment idealism appealed to architecture students during the 1960s, then cheerfully engaged in dismantling the Modern Movement and toppling its utopian claims. But by the 1980s, the stir caused by A Pattern Language had subsided. Alexander’s views seemed less attractively incendiary than did those of a new generation of architect-theorists who drew their inspiration from the world of linguistics, semiotics, and French literary theory. The most self-­consciously theoretical of these was Peter Eisenman, the champion of architectural deconstructivism, and in 1982 he and Alexander squared off in a celebrated and shockingly freewheeling public debate at Harvard’s Graduate School of Design.

To read the transcript is to get a sense of two dangerous animals from entirely different worlds, like a polar bear and a tiger, mutually baffled, and circling one another warily, not even knowing how to strike. Eisenman feinted first, claiming that ­Alexander’s Notes on the Synthesis of Form “so ­infuriated me, that I was moved to do a PhD thesis myself . . . to dialectically refute the arguments made in his book.” But his attempt to bring the authority of the French post-structuralists faltered when it came up against Alexander’s total—and utterly disarming—ignorance of fashionable theory:

Alexander: I don’t know the people you are talking about.

Eisenman: I am talking about people like Roland Barthes, Michel Foucault, Jacques Derrida.

Alexander: What do they say?

Trying to find common ground, Alexander proposed that they consider a building they both regarded as great, Chartres Cathedral. Eisenman interrupted, calling Chartres boring and uninteresting: “Once you’ve seen one Gothic cathedral, you’ve seen them all.” Alexander was stunned, and responded heatedly: “It never occurred to me that someone could so explicitly reject the core ­experience of something like Chartres. . . . If this weren’t a public situation, I’d be tempted to get into this on a psychiatric level.”

For Alexander, architecture should aspire to a condition of wholeness, to that “order produced by centers or wholes which are reinforcing each other and creating each other.” Eisenman was having none of this. For him, “it is not wholeness that will evoke our truest feelings[. Rather,] it is precisely the wholeness of the anthropocentric world[, and] it might be the presence of absence, that is, the nonwhole, the fragment which might produce a condition that would more closely approximate our innate feelings today.” Alexander was appalled by the linguistic acrobatics, which he saw as “nothing but intellectualisms which have little to do with the core of architecture. This depends, as it always has, on feeling.” Seldom does a debate expose with such clarity the fault line between two mutually incompatible systems of thought.

This depends, as it always has, on feeling. It is curious that the word feeling looms so large in Alexander’s mature thought. It barely appears in his Notes on the Synthesis of Form, which sought to account for all design in terms of objective, formal relationships. There he was thinking as a mathematician, or perhaps as a physicist, looking for a unified field theory of design, so to speak, which would apply to everything from a small city to a bathroom towel rack. One might think that a rigorous-minded mathematician, turned loose on architecture, would produce mere elegant engineering, along the lines of ­Buckminster ­Fuller’s geodesic dome. But Alexander’s logical bent, far from reducing architecture to a sterile affair of abstract relationships, brought with it a thrilling intellectual freedom. By looking at architecture as a mathematician might, unsentimentally and ahistorically, he was liberated to look at any building, across time and across cultures, and give it respectful attention. His writings are filled with ­thoughtful references to the mud huts of the ­Musgum in Cameroon, the hogans of the Navajo, and the “black houses of the Outer Hebrides.”

Much of Alexander’s charm derives from his habit of delivering opinionated pronouncements on any topic, some of them simply wacky (for example, his prediction that the England of the future would split into five independent regions: Kent, Wessex, Mercia, Anglia, and Northumbria). And yet for all its personal digressions and indulgences, A Pattern Language retains its bracing freshness. At a moment when CAD, the technology of digital design, has removed the architect further than ever from the tactile world of site, climate, and materials, it is more timely than ever.

It is the paradox of Christopher Alexander that he spoke like a mathematician but wrote like a humanist. He lectured haltingly, thinking as he went along, and pausing with apparently no sense of elapsed time (a habit that worsened as he aged). But when he wrote, he did so with a poignant, even tragic dignity, as in his celebrated introduction to A Pattern Language: “When you build a thing you cannot merely build that thing in isolation, but must repair the world around it, and within it, so that the larger world at that one place becomes more coherent, and more whole.” As a rule, architectural epigrams are dreary affairs—form follows function or less is more—but here is one that speaks to the human heart. Is it asking too much to have it engraved on the threshold of our schools of architecture?

Michael J. Lewis is Faison-Pierson-Stoddard Professor of Art History at Williams College.

Image by Peter Morville via Creative Commons. Image cropped.