“A significant portion of the intellect and capital invested in architectural education over the last 10 years has gone into digital fabrication, as schools have acquired new computer numerical control (CNC) machines and faculty and students have experimented with them. The relevant arguments made by Bernard Cache, Greg Lynn, and Mario Carpo that inspired much of this work have become implicit for many instructors and some students. Always practice driven, these ideas have seeped into the profession, enabling an expanding array of pavilion projects and fabrication competitions. . . . ‘”
*Image courtesy the author. Yale School of Architecture, Assembly One Pavilion. Diagram.
images, clockwise from top left: Yale School of Architecture, Assembly One Pavilion. Exterior elevation; interior; detail; exterior toward entry. Photos by Chris Morgan Photography.
A significant portion of the intellect and capital invested in architectural education over the last 10 years has gone into digital fabrication, as schools have acquired new computer numerical control (CNC) machines and faculty and students have experimented with them. The relevant arguments made by Bernard Cache, Greg Lynn, and Mario Carpo that inspired much of this work have become implicit for many instructors and some students. Always practice driven, these ideas have seeped into the profession, enabling an expanding array of pavilion projects and fabrication competitions. In fact, CNC processes continue to revolutionize the building industry at all scales, but they seem to have run up against something of an “acrylic ceiling” in academia, isolated at the periphery as electives and rarely playing a significant role in design studios.
Clearly, there is always value for students in working with current technology, at full scale and with real materials, but is there still a relevant project to be found in teaching fabrication beyond the general benefits of craft? Is there a new argument to be made, one following up on those about mass customization and consumer culture from Lynn, or aesthetic notions of sameness and repetition from Carpo? Can or should fabrication play a more central role in design education?
Together with a group of post-professional students, I recently initiated a new project at the Yale School of Architecture with the open-ended title “Assembly.” We thought of it as the younger, more carefree sister to the Yale Building Project—the 40-plus-year tradition in which Yale’s first-year graduate students design and build a house. Unlike the Building Project, which focuses on the core urban and programmatic questions of housing, our Assembly project was geared toward exploiting Yale’s extensive CNC technology, and that focus on technology not only transformed the students’ means of production, but also impacted how they approached the project from conception. A different way of realizing the project forced the students to rethink their roles as designers and uncovered a clear inversion of some basic assumptions about working digitally.
Almost since the introduction of digital techniques to architecture, they have been associated with Peter Eisenman’s project of explicit process and indexical form. The autonomous programming languages that underlie software evoke Eisenman’s vision of an underlying grammar for architecture’s own internalized language. His strategic use of explicit, often sequential formal manipulations lent itself to the distinct and numeric nature of digital transformations such as translation, rotation, and scaling. As they have proliferated, digital techniques have been read by countless others to foreground process, indexicality, and the apparent “difficulty” of design. What our Assembly course revealed and made clear to the participating students, however, was the opposite: that the integral nature of the digital model absorbs individual design decisions and specific manipulations, rendering them indistinguishable. The integration of fabrication into the project forced the group to work systematically rather than sequentially. Instead of moving from site to massing, to structure, material, and detail, responses were developed to each constraint simultaneously and incorporated into a single model. The result was a completed project that cannot be easily read as indexical—that is, as a record of a process or series of events. But if the link between the architectural index and digital technique can indeed be broken, Eisenman’s deeper interest in mediated authorship might still be preserved.
Explicit process gave Eisenman an alternative to architecture’s humanist focus, dominant since the Renaissance, and a way to challenge his own intuitive authorship. Alejandro Zaera-Polo, in an introduction to Eisenman’s work in a 1997 issue of the Spanish magazine El Croquis, describes this critical tactic: “By replacing the origins, the presence and the author by arbitrariness, absence and machinic behavior, he has found the recipe for a non-conservative resistance.”1 Zaera-Polo cites Eisenman’s Aronoff Center for the Arts in Cincinnati, Ohio, as the best example to date of this machinic process. Zaera-Polo’s extensive description of each successive formal manipulation—sequential “displacements,” “re-orientations,” “asymptotic tilts,” and “exponential overlaps”—is supplemented by a flow chart placing each move in a rationalized, if still arbitrary sequence.2
At the time, it appeared that emerging software would allow Eisenman to extend this trajectory, rendering each step in the sequence even more explicit and partitioned from the vagaries of intuition. However, as Patrik Schumacher maintains, the digital model can now easily become so info rich that it becomes circular, looping back to simultaneously incorporate ever more constraints. In fact, the digital design model may open up an alternate model of mediated authorship, one that “produces . . . results far beyond the architect’s ‘natural’ . . . range.”3
At a small scale, the Yale Assembly project cast in stark contrast the differences between a project developed in models and drawings and one developed for fabrication. Both Assembly and the Yale Building Project entail an elaborate design, mobilization, and construction process that involves both collaboration and delegation. For the Building Project, stick frame construction and some form of contextual deference are assumed, leaving the students to work out the massing and interior organization first, before developing strategies for windows and doors, materials, and the landscape. Later, once a winning design is chosen, the class tackles the specifics of structure, detailing, furnishing, and material sourcing. The entire sequence is a cascade of development that generally moves from the large scale to the small—from the apparently important concerns of site and program to the less consequential questions of character and environment.
The group of 13 students designing the Assembly pavilion was initially drawn to a similar sequence, at first diving into the site and potential massing shapes before realizing that the potentials of material, detail, and structure were actually the central questions to consider. Their process jumped between considering the size and shape of the project on the green, the rigidity of multiple materials in different configurations, the visual and environmental effects of those forms, and the limits of the project’s budget. Clearly any design project, whatever the medium of its rendering, will incorporate at least this many decisions, but what became clear during Assembly was a complete loss of both scalar and temporal sequence. The massing of the project was reinvented countless times as the material, detailing, or even the paint job changed.
This state of unstable interdependence was mandated by the design’s digital model, which consisted of a two-dimensional structural pattern, a single point toward which that pattern was extruded, and an inner and outer envelope used to trim away the extruded surfaces. Sketching or imagining any of the three in isolation was meaningless. What followed was a constant game of adaptation that took the students far from what they initially imagined. This synthetic structure affects the design process in several specific ways. First, the moment of inspiration is drawn out. Design conception no longer has the purity or immediacy of a momentary idea or quick sketch, but instead emerges in unexpected ways over the entire course of the project. Second, the hierarchy of constraints is leveled. Fabrication projects privilege a different set of questions than representation-based building projects. Program and urban or site constraints are generally simplified in favor of material properties and perceptual effects, raising the elements of the physical environment to the same status as site and program. Allowing this alternate structure for design to invade design studio might raise alternatives to the still prevalent design sequence that begins with site analysis and massing sketch and ends with material, detail, and finish.
Eisenman posed mediated authorship as a way to free himself from his own intuition, but he also hoped to escape the constraints of dominant modes of production. The arbitrariness of the design process allowed him to temporarily ignore and potentially reinvent the way his own buildings are built. “Assembly” suggested the reverse: that the imposed structure of digital fabrication enables its own form of mediated authorship. An expanded set of production techniques can allow architects and students to reinvent the way they design. Even without rendering the design process explicit, digital modes of design and production may help students not only transcend the assumptions and brackets they bring to their work, but also reframe the way they make architecture.