Who Do We Think We Are?   /   Spring 2021   /    Notes & Comments

The Unbearable Burden of Invention

A corollary of giving priority to invention is that imitation, once the foundation of creativity in architecture, is banished.

Witold Rybczynski

The Stedelijk Museum in Amsterdam; photograph by Jannes Linders.

Buildings’ nicknames are the public’s attempt to make sense of the incomprehensible. Several odd-looking London skyscrapers have cheekily illustrative monikers: the Gherkin, the Cheesegrater, the Walkie-Talkie. Angelenos call the mammoth Pacific Design Center the Blue Whale. Beijingites offhandedly refer to the headquarters of China Central Television as Big Underpants. A Shanghai skyscraper with an aperture at the top is the Bottle Opener, and Bilbao has the Artichoke, Frank Gehry’s titanium Guggenheim museum. My favorite is the nickname of an addition to the Stedelijk Museum in Amsterdam—the Bathtub.

The original Stedelijk Museum, or city museum, was built in 1895 in the style of the sixteenth-century Dutch Renaissance. The gingerbread red-brick building with pale stone stripes is pretty as a picture. The 2012 modern addition, which doubled the size of the museum, is the work of the Amsterdam architectural firm Benthem Crouwel. The competition-winning design ignores its neighbor and obviously aspires to be the Dutch equivalent of the Bilbao Guggenheim, an in-your-face architectural icon. From certain angles, the windowless white form, raised in the air and covered in a reinforced synthetic fiber finished in glossy white paint, really does resemble a giant hot tub. Michael Kimmelman of the New York Times observed that “entering an oversize plumbing fixture to commune with classic modern art is like hearing Bach played by a man wearing a clown suit.” Not good.

“Good architecture can be startling, or at least might not look like what we are used to,” writes the critic Aaron Betsky in Architect magazine. “Experimentation can sometimes look weird at first, but it is a necessary part of figuring out how to make our human-built world better.” Now so used to buildings that break the bounds of convention, we find the suggestion that experimentation is an essential part of good architecture unremarkable, even banal. But is it true?

Berlin’s Altes Museum, built in 1822, doesn’t look like a plumbing fixture. Its architect, Karl Friedrich Schinkel, modeled the 300-foot façade of giant Ionic columns on an ancient Greek stoa (a covered walkway or portico). Inside, he based a two-story-high rotunda on the Roman Pantheon. Schinkel was one of the most inventive architects of the nineteenth century—the plan of the museum, with its circuit of long, narrow galleries, was without precedent, and the severe side and rear elevations, which would inspire later modernists such as Mies van der Rohe, were almost shockingly plain. Yet like so many architects before him, Schinkel kept one eye on the past. That meant imitation as well as invention.

Imitation was at the heart of the Italian Renaissance. Starting with Filippo Brunelleschi, architects sketched and measured Roman ruins and incorporated the capitals, friezes, and moldings into their own work. Although the functions of the buildings they designed, such as hospitals, palazzos, and country villas, were new, the elements of their architecture—its language—were old. Renaissance architects also copied from each other. Andrea Palladio copied the so-called Palladian window, an arch-and-columns motif, from the Library of Saint Mark in Venice, whose architect, Jacopo Sansovino, had copied it from Donato Bramante, who used it first in the choir of Santa Maria Del Popolo in Rome.

Bramante was responsible for another architectural invention. When he designed the Tempietto, a commemorative monument on the site of St. Peter’s crucifixion in San Pietro di Montorio in Rome, he modeled the tiny chapel on the circular Temple of Hercules Victor, the oldest surviving marble temple in Rome, and he incorporated Roman spolia (repurposed building materials) in the form of reused Tuscan columns. But he also added something novel: a tall drum surmounted by a dome projected above the circular colonnade. This combination of new and old struck his contemporaries as a stroke of genius. Bramante’s influence is apparent in Michelangelo’s great dome of St. Peter’s Basilica, as well as in domed buildings such as St. Paul’s Cathedral in London, the Panthéon in Paris, and the US Capitol in Washington, DC.

Invention, always a part of architecture, was usually restricted to a few gifted individuals—the rest followed. “Imitation is the sincerest form of flattery that mediocrity can pay to greatness,” wrote Oscar Wilde. Yet imitation not only allowed lesser talents to learn from the masters, and in the process raised the level of workaday buildings, it also permitted great architects such as Michelangelo and Schinkel to build on the achievements of their predecessors.

The architectural Modern Movement of the early twentieth century put a stop to this practice. The credo of the movement was that the modern age required its own distinctive architecture. As J.J.P. Oud (1890–1963), a prominent Dutch modernist, put it, “All in all it follows that an architecture rationally based on the circumstances of life today would be in every sense opposed to the sort of architecture that has existed up till now.” In the 1920s, opposing the past meant flat roofs without eaves or cornices, horizontal strip windows without frames, buildings raised up on stilts instead of sitting on the ground, and white walls bereft of decoration. Henceforth, history was canceled—no more looking back, no more learning from earlier trial and error.

Repudiating tradition opened a Pandora’s box. For a brief period, the stark International Style reigned supreme, but the creativity of architects—as well as the demands of clients—was irrepressible. Having banished the historical canon, all that architects had was their own invention. Le Corbusier was one of the first to exploit this new freedom, designing Notre-Dame du Haut, a pilgrimage chapel in Ronchamp, France, that resembled a nun’s coif. It was followed, at what is now John F. Kennedy International Airport in New York, by Eero Saarinen’s TWA terminal, which looked like a bird in flight, and the billowing forms of Jørn Utzon’s Sydney Opera House, which reminded some of yacht sails and others of the overlapping plates of an armadillo’s shell. To be called a “form-giver” became the highest praise an architect could receive.

One of early modernism’s dictums, first voiced by the Chicago architect Louis Sullivan, was “Form follows function.” But the forms of Le Corbusier’s chapel, Saarinen’s terminal, and Utzon’s opera house had nothing to do with what went on inside; in fact, the unorthodox shell shapes of Utzon’s building severely constrained the design of the performance halls within. The effect of invention on architecture can be judged by comparing two buildings that were designed within a decade of each other: the National Gallery of Art, in Washington, DC, and the Solomon R. Guggenheim Museum, in New York. John Russell Pope designed the former according to the established method of imitating and modifying old forms; in the latter, Frank Lloyd Wright produced a building that is a textbook example of unrestrained invention.

The centerpiece of Pope’s design is a Pantheon-like rotunda echoing the one in Schinkel’s Altes Museum, although larger in scale. The rotunda, like the Ionic porticos that signal the entrances to the building, is there to set the mood of the museum—a secular temple of art. In the galleries, Pope followed established convention, providing a variety of tall sky-lit rooms with glass ceilings to create a variety of settings for the paintings. At the same time, he designed two new kinds of public spaces: a long, airy, and monumental sculpture hall that provides a sense of orientation in the labyrinthine museum and, at either end of the hall, glass-roofed atria, not exhibition rooms but plant-filled conservatories where the weary visitor can relax. The result is one of the great museum buildings of the twentieth century.

Wright likewise took a rotunda as his starting point—he once called the Guggenheim “my Pantheon”—but he placed the entire museum within the round space. Instead of discrete galleries, he designed a continuous helical ramp. The visitor takes an elevator to the top of the ramp, then walks down, experiencing the art along the way. It is an ingenious invention, yet one with serious practical drawbacks: The display spaces are identical, the natural lighting of the paintings is not great, the sloping floor is a distraction, and the exciting views across the rotunda compete with the art. Nor does the idea of a continuous path really work; there are no shortcuts if you want to make a quick visit. Once you are on the ramp, you are Wright’s captive.

A corollary of giving priority to invention is that imitation, once the foundation of creativity in architecture, is banished, and copying is considered the mark of a lack of imagination, or worse, plagiarism. This is evident in the notorious case of the Kimbell Art Museum, in Fort Worth, Texas. The museum, designed by Louis Kahn in the 1960s, is celebrated both for its architecture and its success as a setting for art. In 1989, Romaldo Giurgola, a friend and colleague of the now-deceased Kahn, was commissioned to expand the museum, and his modest proposal replicated the original building’s modular plan and sky-lit vaults, in much the same way that buildings in the past were extended and added on to. Giurgola’s proposal caused an outcry among architects and critics, and he was accused of “vulgar mimicry.” The chastened museum shelved the plan, and twenty-five years later, when Renzo Piano designed an addition, he made sure that it was entirely separate—and different.

There is a final downside to invention. Buildings that look weird are one thing, but buildings that act weird are quite another. Invention, as Betsky admits, “sometimes stretches the technology of building to the point that it creates problems.” That is what happened in 1978 when I.M. Pei built the East Building, a modernist addition to Pope’s National Gallery. Although Pei matched the Tennessee marble skin of the older building, he did not match the way the addition was built. There is nothing particularly novel about using marble as a veneer—the Roman Colosseum was built that way. The original National Gallery, which was completed in 1941, has a marble skin that is four to eight inches thick, effectively a separate self-supporting wall. That wall has stood intact for eighty years. Large buildings like the National Gallery require expansion joints, which in Pope’s design are concealed behind columns and moldings. The minimalist East Building has no such features, and in order to eliminate the unsightly joints, Pei invented a new type of skin that required no expansion joints and consisted of three-inch-thick marble slabs independently suspended from the concrete structure by stainless steel clips and anchors. This technique had not been tried before; Pei called it a “breakthrough.” Less than thirty years after the East Building was completed, the slabs started to bulge and crack. In 2011, in a process that lasted three years, the entire marble skin had to be taken down and rehung.

Another celebrated failure is Lever House, on Park Avenue in New York, designed by Skidmore, Owings & Merrill in 1952 and hailed as one of the first high-rise office buildings in the city to have a so-called curtain wall. Previous skyscrapers, up to and including the Empire State Building and Rockefeller Center’s RCA Building, had steel frames surrounded by thick masonry walls made of stone or brick, well known materials that architects had been using for centuries. The curtain wall replaced the heavy masonry with a lightweight grid of steel and glass that hung—like a curtain—from the structural frame. After only thirty years, the blue-green glass skin of Lever House showed signs of deterioration, with many of the glass spandrel panels requiring replacement due to cracks. Sixteen years later, the entire curtain wall was removed and rebuilt from scratch.

Failures such as the East Building and Lever House stand out for their visibility—and the resulting expense of the repairs—but it is not unusual for modernist buildings to require extensive renovation after a relatively short time. A participant in a 2013 Getty Center colloquium on conserving modern architecture casually observed that conventional buildings traditionally lasted about 120 years before major repairs were required, whereas for modernist buildings it is only half that time—sixty years. Only sixty years! Yale University’s architectural masterpieces of the 1960s lasted even less time than that. Louis Kahn’s Art Gallery, Paul Rudolph’s Art and Architecture Building, and Eero Saarinen’s Morse and Stiles Colleges all underwent major renovations in the early 2000s at a cost far exceeding that of the original construction. Part of the reason was that their architects had been guided by invention rather than convention. Kevin Roche, who was Saarinen’s associate at the time Morse and Stiles were designed and built, confessed in a filmed interview, “You’re sort of lunging out into the future so you do things that in retrospect may or may not work. That’s the nature of any experimental architecture.”

According to the first-century Roman architect Vitruvius Pollio, the three essential qualities of good architecture are firmitas, utilitas, and venustas: firmness, utility, and beauty. He did not include experimentum. For a long time, firmness, that is, durability, could be taken for granted. A building might be clad in marble, brick, or stucco, but with regular maintenance—cleaning, repointing, plastering, and painting—it could be expected to last. “Experimental architecture” changed that. Reinforced concrete, for example, seemed almost magical; not only was it inexpensive, it allowed dramatic cantilevers, shell-thin vaults, skinny columns. Reinforced concrete proved to be useful for the structure of a building—columns and floors—but because it was porous and weathered poorly, it was a poor substitute for stone or brick as an external cladding. It took several decades to discover that steel and concrete are precarious partners—concrete cracks, steel rusts, and spalling follows. By then, Brutalism had come and gone, leaving a trail of rusting, discolored, and flaking buildings in its wake.

Evidently, experimentation and invention can be dangerous, practically as well as aesthetically. Traditionally, learning from the past ensured continuity, consistency, and material solidity. Looking back meant learning from inventive predecessors, the way Michelangelo learned from Bramante, and Christopher Wren learned from Michelangelo. “Architects have always looked back in order to move forward,” observed the British master James Stirling. But modernism has removed the rearview mirror. Now architects look only in one direction—ahead. Looking ahead, not learning from the past, inventing and not copying, means that architects are in the position of constantly starting from scratch. This can be exciting—when it works. But creative genius is rare, and the inevitable result is a small number of remarkable works and a very large number of failed attempts, not to mention many weird buildings. Cue the Bathtub.