Random Acts of Architecture
In the industrial section south of San Francisco’s Market Street, the straight blocks of buildings unexpectedly give way to an intimate circle of homes and businesses surrounding a green oval. Conceived in the 1850s as a posh residential development, South Park has become a truly mixed-used neighborhood where architects’ offices and high-tech start-ups alternate with private residences in a stylistic hodgepodge that includes both neoclassical moldings and steel-and-glass cladding. But even in the midst of this design diversity, one of the newest homes stands out: a three-story modernist box with a sculptural metal web spanning the top two floors.
The oversize front door of blackened steel offers a clue to the building’s double identity. At 12 feet high, it’s tall enough to permit large artworks to pass through, which is ideal since the dwelling functions as a mini-museum for the owners’ 300-plus-piece art collection. Luke Ogrydziak and Zoë Prillinger, San Francisco architects who are known for their conceptual approach, created this half-curatorial, half-domestic environment called the Gallery House.
Particularly interesting are the structure’s open-ended experiments in technology, an indication that architecture may finally be moving beyond its infantile preoccupation with “blobitecture” and computer-generated forms. Sean Keller, an architectural historian and a professor at the Illinois Institute of Technology, describes Ogrydziak/Prillinger Architects as “one of the best examples of a second-generation practice. Ten years ago, everyone was dazzled by technology. The tools were so impressive, you could make forms that seemed impossible. That was the breakthrough. But this firm is continuing past that moment. They accept that the tools are there, but they have a deep understanding of the algorithms and what they do, instead of just taking the tools that they are given. They’re also tying the technology in a smart way back to architectural history and broader aesthetic issues.”
The most visible of these experiments is the Cor-Ten steel latticework draped over the front of the house. Like many modern artworks, it was fueled by an impish desire to thumb one’s nose at the establishment. “The San Francisco planning code encourages Victorian bay windows, but when you look at it closely, there’s nothing that actually states that you have to create a bay window,” says Ogrydziak, a 39-year-old partner in the firm. “It just describes this little chamfered envelope. In the zeitgeist that architects live in, where we’re always constrained by planning codes, we found ourselves emboldened by the realization that it could be crazy! So we deliberately misread it as a mathematical description, which allowed for infinite possibilities.”
The clients — she’s a professor at the University of California, San Francisco, and he’s a researcher and software developer — were able to appreciate the formal pleasures of the design process. They collect work strictly by female artists, such as Louise Bourgeois and Jenny Holzer. For the facade, they simply requested something that would provide a transition between the glass walls of their living room and master suite and the street. Playing with the ambiguity in the planning code, the architects designed an undulating shape in lieu of a projecting window. Ogrydziak wrote a computer program that randomly set down the points that would become the latticework’s vertices and had the intelligence to keep them from overlapping (“collision detection,” in programming terms) or obstructing sight lines. He then used a classic algorithm for modeling landscapes in order to “grow” the triangular mesh. The architects ran the sequence many times before selecting the one they found most appealing.
The clients say the final result, which is as much art as architecture, conveys the “idea of tree branches.” It acts as a bridge between the tree canopy of the park and the rigidly orthogonal building. Because the couple wanted the house to be a vehicle for displaying their collection, the architects created a rational structure with white walls, floor-to-ceiling sliding-glass doors, and skylights to let in profuse natural light. Presented with the narrow, 24-foot-wide infill lot, they modeled the building after a warehouse’s open, generic shell; all the internal walls are non–load-bearing. The first floor, which was designed as a code-compliant commercial space and has its own separate entrance, was inspired by the loft-like galleries of New York’s Chelsea neighborhood.
The house’s custom-programmed “bay window” qualifies as one of the most innovative interpretations of San Francisco vernacular architecture to date, but Keller thinks it signifies something more. By opening the design process to chance, Ogrydziak and Prillinger are exploring territory similar to what abstract painters investigated in the 1950s and ’60s. “Rather than paint a single work, artists like Sol LeWitt were setting up guidelines and environments in which things could happen,” he says. “Digital design tools have brought architects back to this question. They’re not placing every line themselves, which is a different way of working. Architects like to be in control of everything, so the idea that you would relax and see what happens is fairly revolutionary.”
Professional and personal partners, Ogrydziak and Prillinger met during architecture school at Princeton University, where their professors included the theoreticians Peter Eisenman, Anthony Vidler, and Mark Wigley. “Everybody was interested in psychoanalysis, and there was a lot of talk about how to release the repressed, if you will,” says Prillinger, who is also 39. “When design is truly exciting, you don’t know exactly what’s going to happen.” But that doesn’t mean turning things over entirely to the forces of chaos. “Of course you still want to guide the design, like training a plant,” Ogrydziak says. “When we set up the program to be totally random, there was no tension in it, so we had to modify it. But every time you press ‘Go,’ the program does something that you couldn’t have done by hand.”
After teaming up in 2000, the architects looked at ways to break out of what Prillinger describes as “the inheritance of classic modernism, a staunchly retrograde attachment to the Cartesian grid.” A series of well-received residential projects included Honighaus, a remodeled Edwardian home with a dramatically angled penthouse in San Francisco’s Lower Pacific Heights neighborhood. The architects make use of Rhino, ArchiCAD, and other 3-D-modeling tools in their practice but began developing their own design tools a couple of years ago. They used a visual-programming tool called Processing, designed specifically for artists. Through it, Ogrydziak was able to get comfortable enough with computer programming to write his own code.
The architects first introduced controlled randomness into their work with a CNC-milled concrete wall, a water feature at Honighaus. The program generated graceful, origamilike pleats for the water to flow over. “It’s scary at first to build something straight from the computer,” Ogrydziak says. “We wouldn’t have had the guts to do that with the facade here if we hadn’t already had that experience.”
Ogrydziak/Prillinger had the chance to try a little more randomness in the roof garden of the Gallery House. Dotted among the alien-looking succulents are more of what the architects call “mathematical-organic” objects. Pavers marked with the facade’s mesh pattern have irregular forms generated by a complementary algorithm. Three intricately faceted boulders, created by manipulating geometric solids, serve as benches. In this otherworldly sculpture garden, rocks have been swapped out for counterparts from a parallel universe. “We’re trying to see a world with many, many options and then selecting a few,” Prillinger says. “We’re trying to see a space for design as charged with possibility.”
What do the architects hope to do next? “We’d like to see how we can introduce that sense of discovery to an entire building,” Ogrydziak says. “It’s interesting to see what happens when there isn’t a neat resolution, that there’s some messiness to a project.”