Through a reconstruction of the murky process involved in the design of the Hajj Terminal at the King Abdulaziz International Airport in Jeddah, Saudi Arabia, by Skidmore, Owings & Merrill, Matthew Allen examines how computational expertise negotiated a new role for itself within the hierarchy of late twentieth-century corporate architecture. In The Genius of Bureaucracy: SOM’s Hajj Terminal and Geiger Berger Associates’ Form-Finding Software he explores how Henry-Russell Hitchcock’s rubric of “the architecture of genius and the architecture of bureaucracy” played out in a situation where the two ideals converged. Unusual among other large corporate firms, SOM not only invested in computation early on but also leveraged computer-generated images to promote its practice. Examination of SOM’s printouts from the Hajj Terminal project reveals the pervasive presence of the lower ranks of the corporate hierarchy—particularly the engineering subconsultants Geiger Berger Associates, who developed unique software for engineering tensile structures. In this contest between decisive SOM designers such as Gordon Bunshaft and Fazlur Khan and engineering paperwork, a new understanding of “the computer” emerged that congealed around the concept of form-finding, or “smart” digital modeling.

In the 1970s and 1980s, the growing subculture of architectural computation acquired new significance for large architecture firms. The leading firm of Skidmore, Owings & Merrill served as a key site for this development. Many projects by SOM from this era included a new cast of devices, technicians, and consultants within bureaucratic hierarchies that were both smaller and larger than the firm itself. Internally, SOM divided itself into teams, or “studios,” that came to include far-flung networks of external experts. Creative decision making sometimes happened deep within these ad hoc hierarchies rather than in the offices of the firm’s design partners. The blending of design talent with organizational acumen, or the productive tension between the genius and the bureaucrat, became not only a hallmark of architectural practice at SOM but also a defining feature of mid- to late twentieth-century architectural practice in general.

The Hajj Terminal at the King Abdulaziz International Airport in Jeddah, Saudi Arabia, designed by SOM from 1974 to 1981, exemplifies this new kind of bureaucratic architecture. The Hajj Terminal would not have turned out as it did without a strong bureaucracy to guide its development, and it probably could have been designed without any genius figure at all—even though, as we will see, several people competed for this part. This study examines the roles of experts, routines, software, and paperwork in the genesis of the Hajj Terminal, and in particular the emergence of a new type of “smart” modeling software—often called form-finding software—developed by a firm that served as a subconsultant on the project, Geiger Berger Associates.

In the years after World War II, SOM epitomized the rise of a new type of bureaucratic architecture firm. Henry-Russell Hitchcock ended his widely read 1947 article on genius and bureaucracy with a discussion of Oak Ridge National Laboratory, for which SOM had developed the plan.1 Hitchcock conceded that some firms exhibit a rare “genius for organization,” and as planning was becoming a field of its own, architects with “broad imaginative capacity” ought to shift into less glamorous but more necessary jobs in the postwar reconstruction effort. SOM did just that—although “reconstruction” here must be broadly defined to encompass establishing the global presence of American corporations and institutions through the design and construction of innumerable office buildings. By the late 1980s, SOM’s authoritative professional expertise in “corporate architecture” enabled the firm “set the standards” in this realm.2

Despite SOM’s reputation for designing architecture for a new breed of “organizational man,” it could hardly be characterized as an unimaginative firm that generated anonymous buildings.3 Although as a rule corporate architecture avoided the flamboyant forms of Frank Lloyd Wright (Hitchcock’s archetypal genius figure), its design standards were still high. At SOM, Gordon Bunshaft (at SOM 1937–79) was the best-known partner who focused on design, but other figures, including Myron Goldsmith (1946–53), Walter Netsch (1947–79), and Bruce Graham (1951–89), also attached their names to iconic works. The firm even produced a star engineer, Fazlur Khan (1955–82). SOM’s internal structure afforded the principals extensive control over projects: any principal could seek out their own commissions, pair up with an “administrative partner” within the firm, and assemble a hierarchy of architects, draftsmen, engineers, and various others to create what was essentially an ad hoc firm-within-a-firm.4 SOM restructured as a set of such “design studios” in 1972, and by 1980 fourteen of these studios had been established.5 The period when SOM employed its strongest designer, Bunshaft, arguably also represented the apex of the firm’s corporate image.6 More than any other design principal, Bunshaft had a reputation as a blunt and demanding decision maker, as the incarnation of the “modern master” archetype on which he modeled himself.7

Even a cursory glance at SOM illustrates the inadequacy of Hitchcock’s simple dichotomy to describe the complex internal organization of large corporate firms. Genius coexisted with bureaucracy in various combinations and in differing degrees of tension. As we will see, SOM wrestled with the challenge of maintaining the appearance of an innovative and creative design firm as it became increasingly bureaucratic. Other firms followed similar trajectories: historians often trace the origins of corporate architecture back to the burgeoning firms of late nineteenth-century Chicago (Burnham & Root, founded 1873) and New York (McKim, Mead & White, founded 1872).8 After World War II, the largest firms grew even larger and claimed a greater proportion of work in the building industry.9 Similar developments occurred in other industries as well, with factors including new management styles, new technologies, a changing political and regulatory environment, and evolving professional norms and ideals.10 In a self-reinforcing cycle, large corporations across various industries sought out a different type of architecture to express their corporate identities and cultures, and they turned to corporate architecture firms to meet these goals.11 As a result, some architecture firms became more anonymous: for example, while renowned architects Cesar Pelli and Anthony Lumsden had directed Daniel, Mann, Johnson & Mendenhall (DMJM) in the late 1960s, the firm reorganized as the global giant AECOM (Architecture, Engineering, Construction, Operations, and Management) in 1990. Other large firms sought in various ways to highlight the individual creativity of their designers. Some relied on a single prominent name, as with Philip Johnson at Johnson/Burgee. Some, like SOM, organized into teams or studios. Sometimes these teams formed around just two partners who could thereby take on projects with creative autonomy (e.g., Harrison & Abramovitz). As time went on, a changing cast of partners in a growing number of cities often supplanted the original namesake partners (e.g., SOM and Hellmuth, Obata + Kassabaum, or HOK). These firms—detached from their founders—negotiated a relatively sustainable balance between organizational infrastructure and creative control, with teams forming in response to projects as they came in, or individual partners leading stable studios.

Yet genius did not disappear even as bureaucracy proliferated. It was still important for marketing, if nothing else, although perhaps the desire for creativity among clients shifted from the notion of artistry to the idea of innovation. SOM stands as the exemplar of corporate architecture because it developed a successful formula, with a flexible cast of designers leading teams of anonymous experts, that managed to satisfy the demands for both the traditional ideals of genius and the new ideals of bureaucratic architecture. This idea sold well, not only internally to the firm itself but also externally to clients, contemporary critics, and historians.

Part of the idea at SOM was that designers would be in control of design, but this was sometimes more fiction than reality. When we examine SOM’s output from this era, we find substantial evidence indicating significant input from the lower ranks of the corporate hierarchy. Technicians and their devices left their own indelible marks. For instance, a 1980 Architectural Record article titled “SOM’s Computer Approach” showcased a series of connections for a space-frame structure, a distinctive roof system, and the overall massing of a skyscraper complex, all apparently the work of architect-programmers. While the article explained that this work merely set the stage for “architects working as architects” (the promotional text reassured the reader that the firm’s “computer group … treats design studios as if they were 14 customers”), was this really the case?12 Is it not possible that these simple but well-detailed forms came, instead, from anonymous engineers and their finite element analysis software? Unsurprisingly, the article did not answer such questions about what happened backstage. Readers learned that SOM owned two PDP-11/70 computers and a new VAX-11/780 as well as two expensive plotters, and that the firm wrote its own software. A brochure the firm published in the same year—titled SOM’s Computer Capability—expanded on similar themes.13 But such factoids served less to enlighten readers regarding the firm’s design process than to gesture toward its overall corporate prowess. At the same time, they also highlighted a fundamental contradiction: How could computational tools be both “generic” and tailored to “specific problems,” as stated by the article? How could a hands-off “support staff” also be “integrated” into a design studio?14 These texts made no attempt to address these questions. Instead, they treated both “design” and “computation” as conceptual black boxes.15

This approach poses a problem for historians. If SOM successfully packaged a project as “design” by suppressing evidence of input from technicians in the lower ranks of the corporate hierarchy, on what basis can we describe the genesis of the project? How can we use an absence of evidence as evidence? My aim in this study is to catch obscuration in the act.16 This essay focuses on a moment when an element from deep within the hidden realm of backstage expertise assumed center stage in an architectural project—and then quickly withdrew offstage again to make room for the figure of the genius designer. This case provides an opportunity to investigate how the labor hierarchy at SOM took form around acts of rendering some things visible and others invisible at opportune moments. More generally, it reveals a typical pattern of interaction between architects and technicians in large firms and demonstrates how architectural culture sought to accommodate designers and computer programmers in the 1970s and 1980s.

While historians have written some—though not much—about corporate architecture, bureaucratic architecture remains an underdeveloped field of study. According to classical sociological definitions, a “bureaucracy” is an organizational structure characterized by hierarchy, one that is governed by policies and rules that are beyond the control of the bureaucrats themselves.17 Another key feature might be added: bureaucracies operate by processing documents. In his celebrated description of bureaucratic social organization, Max Weber emphasizes this point: “Modern administration is based on two things. First, on documents that are preserved as original copies … and second, on a staff of officers and writers of all kinds.”18 In short, Weber concludes that officers working with equipment and documents in an office constitute a Büro.

It is worth noting that while bureaucracies flourished in the twentieth century, the organizational form of the bureaucracy emerged with the earliest civilizations: the use of cuneiform tablets for accounting set the Sumerians apart from their neighbors.19 David Wengrow argues that the split of “administrative” societies from “heroic” societies first occurred roughly ten millennia ago.20 However, the modern understanding of bureaucracy reflects other nuances as well. In 1791, a journalist coined the contemptuous term bureaucrat to describe those whose administrative incompetence stifled the dreams of the French Revolution.21 Imaginary accounts of the absurd and cruel behaviors supposedly running rampant inside mindless bureaucracies have accompanied us into the present.22

In this study I take SOM’s status as a bureaucracy for granted. In order to function, most large organizations are by necessity bureaucratic.23 However, bureaucratic architecture entails something more than sheer quantity of bureaucratic production. I argue that SOM’s use of hierarchy, rules, and paperwork was instrumental in producing many of the remarkable qualities of the firm’s architecture. The term bureaucratic architecture rightly refers to this sort of virtuosic performance of bureaucracy.

Prior to the official opening of the new Hajj Terminal designed by SOM, the architectural press published a number of renderings that highlighted the project’s unprecedented features. These included a bird’s-eye view of two enormous white roofscapes separated by a broad roadway, each consisting of a field of 105 fabric units of roughly pyramidal shape (Figure 1). Each of these units covered almost half an acre, and together they dwarfed the adjacent Boeing 747s. Although the journalists who compiled descriptions of the airport did not usually refer to the roofs as “tents,” they clearly identified parallels between the design for the terminal and the design of temporary pilgrim accommodations that “cluster[ed] on the hills” nearby, implying that a temporary nomadic dwelling type provided a key source of inspiration for SOM’s designers.24 Elaborating on this idea, an early feature article indicated that the project lead at SOM used the insights of structural engineering and climate science to reinterpret the classic tents of Saudi Arabia, thereby creating a minimal but monumental roof in the desert.25 The publication juxtaposed photographs of the undulating fabric bays against pictures of rows of temporary tents for pilgrims making the hajj (Figure 2). Although the article identified Fazlur Khan as the lead designer, it conjured up the ideals of Gordon Bunshaft: functional design, simple forms, and clean detailing.26 Photographs of elegant fabric roofs soaring over the heads of suitably ascetic pilgrims reinforced the point that the principles of modernism could be deployed to fit any circumstances around the globe. Such imagery advanced the idea that the terminal’s design consisted of the simplest of formal notions, a stroke of inspiration heroically enlarged and transplanted to the Middle East to create the secular equivalent of a religious experience (Figure 3).

Figure 1

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, 1974–81, rendering (Architectural Record, June 1978).

Figure 1

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, 1974–81, rendering (Architectural Record, June 1978).

Close modal
Figure 2

Temporary tents for the hajj, Mina, Saudi Arabia, 1975 (photo by Fazlur Khan; SOM, Chicago).

Figure 2

Temporary tents for the hajj, Mina, Saudi Arabia, 1975 (photo by Fazlur Khan; SOM, Chicago).

Close modal
Figure 3

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, 1974–81 (Progressive Architecture 63, no. 2 [Feb. 1982]).

Figure 3

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, 1974–81 (Progressive Architecture 63, no. 2 [Feb. 1982]).

Close modal

Intriguingly, the more technical articles offered an alternative way to understand the project by centering on a different type of image: the computer-generated wireframe view (Figure 4). Such images proposed a competing narrative regarding the project’s genesis, suggesting that an engineered form—a precise computer-generated shape—drove “the project” and motivated all other decisions, facilitated by a loophole in SOM’s elaborate bureaucratic hierarchy.27

Figure 4

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, screenshot of roof study in DRAW3D, ca. 1978 (L’Architecture d’Aujourd’hui 223 [Oct. 1982]).

Figure 4

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, screenshot of roof study in DRAW3D, ca. 1978 (L’Architecture d’Aujourd’hui 223 [Oct. 1982]).

Close modal

In the working documents that provide a backstage glimpse into how this process unfolded, we can find numerous contradictions of SOM’s careful public presentation. SOM typically sought to present itself as a monolithic entity. Although designers sometimes referred to projects as if the projects belonged to them, in the end the firm’s work bore the simple attribution of “Skidmore, Owings & Merrill: Architects & Engineers,” with no identification of the various studios or offices involved (Chicago, New York, and San Francisco were the big three). Internal communications, however, reveal both less polish and less consistency. On the green cover of a volume of printed material related to the engineering of the Hajj Terminal roof, for example, the firm’s name appears inside a small typographical box in an ecumenical grid, following the names of both the construction manager and the general contractor (Figure 5). If SOM’s role seems clear, as the “architects and engineers,” the confusing series of names of contractors and consultants that follows undermines this notion. Why does Owens-Corning—a company based in Toledo, Ohio—appear in the guise of a Jeddah-based subsidiary? Why is URS listed as a “consulting engineer” while Geiger Berger Associates—also a consulting engineering firm—is given almost twice as much space and outfitted with the catchy subtitle “New York, New York”?

Figure 5

Skidmore, Owings & Merrill, cover of a book of engineering analysis printouts for the Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, 1979 (SOM Box No. [Chicago/Struct] SO374-1, SOM, Chicago).

Figure 5

Skidmore, Owings & Merrill, cover of a book of engineering analysis printouts for the Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, 1979 (SOM Box No. [Chicago/Struct] SO374-1, SOM, Chicago).

Close modal

This thick volume is but one of many. Indeed, volumes of engineering documents arrived by the boxful throughout the design process; these are now held in SOM’s archives, along with more boxes of loose documentation (letters, memos, reports, notes, and so on), each document stamped, dated, and marked as “approved.” No fewer than four stamps adorn the green volume on the engineering of the Hajj Terminal roof. One identifies the book as belonging to SOM’s Chicago office. Another specifies that it belongs to the Structures group, that it was accepted only pending changes, and that the project as a whole belonged to the Chicago and New York offices.

While the interior pages provide further clues, they also deepen the mystery surrounding the team who actually designed the project. Dozens of volumes with similar green covers, for instance, feature endless tables of numbers documenting innumerable calculations for the shape of the Hajj Terminal roof and possible failure scenarios. Often generated by software and including only rudimentary, abbreviated annotations, these tables are now essentially incomprehensible. Occasionally more sophisticated graphic printouts accompany these pages, sometimes stylized like architectural drawings (Figure 6). Typically, there is little or no indication of who exactly produced all of these documents—someone at SOM, or Geiger Berger Associates, or URS? It seems reasonable to consider them as internal working documents produced by a team of engineers and architects not entirely contained within SOM. If we want to understand the Hajj Terminal as a project emerging from within the flow of documentary paperwork generated by a bureaucratic organization, rather than designed from the top down by an architect, we have a fair amount of reconstructive work to do.

Figure 6

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, roof node map printout, 1979 (SOM Box No. [Chicago/Struct] SO374-2, SOM, Chicago).

Figure 6

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, roof node map printout, 1979 (SOM Box No. [Chicago/Struct] SO374-2, SOM, Chicago).

Close modal

The Hajj Terminal formed only a small part of a much larger endeavor, the King Abdulaziz International Airport, an immense project about the size of the island of Manhattan, certainly surpassing any prior SOM project in scale (Figure 7).28 The final design for the Hajj Terminal itself ended up roughly a quarter of the size of New York’s Central Park, but in the grander scheme it featured as merely a small functional unit: a rectangle initially referred to only as Element 20. Other components of the airport included an air force base, a main terminal for year-round passenger traffic, a mosque, airplane hangars, a royal reception pavilion, a housing complex, and a health care facility, along with a desalination plant, a sewage treatment plant, and all of the services and utilities associated with a major international airport.29

Figure 7

Skidmore, Owings & Merrill, King Abdulaziz International Airport, Jeddah, Saudi Arabia, site plan produced in DRAW2D, 1977 (SOM, Chicago).

Figure 7

Skidmore, Owings & Merrill, King Abdulaziz International Airport, Jeddah, Saudi Arabia, site plan produced in DRAW2D, 1977 (SOM, Chicago).

Close modal

For SOM, participation in this gigantic project meant entering into a far-flung network of contractors and subcontractors. The project began on 16 May 1965, when the Ministry of Defense and Aviation of Saudi Arabia hired the Washington, D.C.-based Airways Engineering Corporation to build an international airport.30 Airways Engineering hired Edward Durrell Stone as an architectural consultant, but the Six-Day War in 1967 put the project on hold.31 During this pause, one of the most dramatic geopolitical episodes of the era occurred, with Saudi Arabia at its nexus. By November 1973, a related series of sociopolitical and economic events—recession in the United States, the end of the Bretton Woods Agreement regulating international exchange rates, a subsequent drop in oil prices, and, finally, OPEC’s successful effort to drive oil prices back up—bestowed extraordinary and unanticipated wealth upon Saudi Arabia.32

Key decision makers in Saudi Arabia, educated in the West, were determined to invest these financial resources in development projects led by Western experts. As elsewhere in the Middle East and across the so-called developing world, architecture became a key field for investment.33 The “race for new markets” marked the expansion of Western corporations in the post–World War II era.34 Although there were many exceptions, large corporate firms based in the West often secured larger projects while small firms focused on smaller projects, whether local or elsewhere. Some firms capitalized on name recognition (e.g., The Architect’s Collaborative, or TAC, led by Walter Gropius at Harvard University, which obtained the commission for the University of Baghdad), while others marketed their organizational and technical acumen. The biggest project of the era for the large St. Louis–based firm HOK, for instance, was King Saud University in Riyadh, Saudi Arabia, and HOK soon issued promotional materials that emphasized the computer network that linked the firm’s global offices.35 Computers thus joined mundane tools like telephones and photocopiers, as well as postal services and air travel, as enablers of expanding corporate practice. Colossal projects such as the King Abdulaziz International Airport blurred boundaries—not only those between architecture and infrastructure but also the boundaries between the different firms involved in the work. As corporate firms began to function as consultancies offering experts for hire and menus of services, the responsibility for large projects appeared to devolve on multicorporate teams of experts rather than on any single firm.

When the airport project began again in 1974, Airways Engineering selected SOM.36 During the course of design, other consultants, including a German general contractor and a U.S. construction firm, also received contracts that gave them status roughly equal to that of SOM. Various other subordinate firms included Owens-Corning, which was chosen over Swiss and Japanese fabric manufacturers. Owens-Corning contracted two engineering firms in New York as well as an architectural fabric consultant located in Buffalo. As this buildup of experts multiplied across the many elements of the airport, the project began to require more coordination than design.

SOM’s work on the airport began when the firm moved the project from its Washington, D.C., office to its much larger New York office.37 Despite Bunshaft’s dominating presence in New York, the assignment landed on the desk of Gordon Wildermuth, the New York partner responsible for Washington, D.C., projects. Understandably, SOM spent the bulk of its time not on the Hajj Terminal but on the rest of the sprawling airport complex. In fact, if the SOM team had wished to focus on “design,” the main passenger terminal would have met that need. Because the Hajj Terminal would be used only to accommodate the temporary influx of visitors each year during the hajj, the problem it presented initially appeared to be a logistical one. How would a million visitors or more be flown in and out of Jeddah over the course of a few weeks?

In the earliest plans, the Hajj Terminal consisted of nothing more than a few giant but apparently featureless rectangles caught up in a complicated infrastructural apparatus. Upon arrival, passengers were to be ushered off planes and then immediately boarded onto buses that would transport them to a large waiting area. The key issues had to do with flows of people and vehicles, which were worked out in diagrams with arrows and programmatic blocks.38 A large clinic would satisfy health concerns, open marketplace areas (labeled imaginatively as “souks”) would serve aggregate needs for rest and recreation, and faceted corners and smooth paths would facilitate the movement of people, who would flow like a gentle stream of water pouring through a series of channels (Figure 8). A vast computerized security system would monitor the whole complex. Given the many intricate problems demanding attention in the initial phase of design, the building’s appearance was a relatively minor concern.

Figure 8

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, plan, 1978 (SOM, Chicago).

Figure 8

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, plan, 1978 (SOM, Chicago).

Close modal

It is likely that the Hajj Terminal would have been built in a default SOM corporate style if a series of contingencies had not ruled out this possibility. Following the approval of Wildermuth’s master plan for the airport in 1976, the first complete design for the Hajj Terminal featured a vast rectangle punctured by courtyards (Figure 9).39 Overall, the scheme resembled a scaled-up version of the Connecticut General Life Insurance Company Headquarters of 1957, itself representative of a set of exurban corporate office buildings designed by SOM. Instead of undulating Noguchi landscapes, the courtyards centered on prismatic forms (perhaps inspired by the Kaaba in Mecca, itself a black cube). It was conceived as a conventional steel-frame structure, sealed and air-conditioned in typical fashion.

Figure 9

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, plan of an early scheme with steel structure, 1975 (SOM, Chicago).

Figure 9

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, plan of an early scheme with steel structure, 1975 (SOM, Chicago).

Close modal

However, the designers soon abandoned this familiar corporate style, as it proved to have multiple flaws. A fully air-conditioned building did not make much sense: in addition to presenting significant energy costs, it would not fit the building’s purpose, which was to serve as a space where pilgrims would transition from climate-controlled airplanes to more austere accommodation in tents. For many pilgrims, this would be the first time they had traveled by plane and the first time they had occupied a building designed around American customs. The designers also recognized that shaded spaces would clearly be preferable to open courtyards.

These realizations triggered a complete redesign. Concrete columns with giant waffle capitals aggregated into a roof replaced the steel-frame structure (Figure 10). Following the proposal of both porous and air-conditioned versions, the SOM team selected the open version as more appropriate. With its vast field of columns, the space would have been something like the prayer hall of the Great Mosque of Córdoba, with echoes of Pier Luigi Nervi’s Gatti Wool Factory (1953) and Frank Lloyd Wright’s Johnson Wax Administration Building (1936–39) (Figure 11).

Figure 10

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, plan of an early scheme with concrete structure, 1976 (SOM, Chicago).

Figure 10

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, plan of an early scheme with concrete structure, 1976 (SOM, Chicago).

Close modal
Figure 11

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, model of an early scheme with concrete structure, 1976 (L’Architecture d’Aujourd’hui 223 [Oct. 1982]).

Figure 11

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, model of an early scheme with concrete structure, 1976 (L’Architecture d’Aujourd’hui 223 [Oct. 1982]).

Close modal

In terms of materials, concrete, like steel, appeared at first to be a safe bet. In previous decades concrete had become a reliable symbol of modernity for architects working in the Middle East.40 The preferred structural systems for most of SOM’s buildings included concrete as well as steel-frame construction. Yet it soon became clear that the Saudi Arabian concrete industry did not have sufficient capacity to supply this massive project, and the infrastructure alone would require years to build.41 SOM’s architects were also anxious about the heat absorption and radiation generated by such a large thermal mass. Not only would the cavernous concrete interior require substantial lighting, but it would also be crowded with hundreds of thick structural columns.

If SOM had little left to work with in terms of design, in the meantime the team had largely sorted out the program of the building. The Hajj Terminal evolved essentially into two separate projects—plan and envelope—and the only remaining question was how to cover all the activity happening on the ground. Having reached an impasse, the SOM team asked Bunshaft to take over the design of the terminal enclosure.42 An architect on the project, Roy Allen, had been experimenting with structural possibilities, including fabric roofs.43 As Bunshaft later reminisced, the designers had pursued these experiments only in a “kind of casual and lazy” way, so he came on board to make decisions and speed things up.44 Bunshaft recalled being seized by a dream of simply covering the whole site with “fabric spanning 450 feet.”45 The team singled out one of Allen’s roofs and quickly worked it up into a complete proposal (Figure 12).

Figure 12

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, model of an early scheme with large fabric roof, 1977 (SOM, Chicago).

Figure 12

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, model of an early scheme with large fabric roof, 1977 (SOM, Chicago).

Close modal

Typically, Bunshaft refrained from elaborating on any precedents he may have had in mind, and thus details are lacking about why this particular design was judged to be the most appealing solution. Nevertheless, two sources of inspiration appear obvious. One is the many indigenous tensile structures of the Arabian Peninsula. Besides the thousands of tents set up annually for the hajj, many of the long-established nomadic pastoral tribal societies on the Arabian Peninsula have their own characteristic tent forms. Philip Drew studied these in a quasi-ethnographic manner and presented his findings to architects in 1979 (two years after Bunshaft and Allen’s design), but Drew’s research marked the culmination of many years of fascination with indigenous lightweight structures among architects.46 It may be mere coincidence that the early design for the fabric roof bears a striking resemblance to the so-called black tents of northern and central Arabia, but it is not inconceivable that Allen or Bunshaft had encountered photographs or drawings of these structures.47

The second obvious source of inspiration for SOM’s first fabric roof scheme is the work of the noted German architect and structural engineer Frei Otto. Drew published a book on Otto in 1976 that presented an alluring variety of projects, mostly unbuilt.48 The most iconic tensile structure of the period, the Munich Olympic Stadium, which relied in part on Otto’s engineering, opened only a few years earlier in 1972. Otto’s Institute for Lightweight Structures at the University of Stuttgart directed a flurry of structural experiments that captured the imaginations of many contemporary architects.49

Yet SOM’s first fabric roof design for the Hajj Terminal was also discarded. One reason was that Saudi officials associated fabric roofs with temporary dwellings, not permanent infrastructural buildings.50 The clients also likely regarded the fabric roof as both anachronistic and culturally imperialist in its connotations. If the roof resembled local tent forms associated with the Saud family’s history within a nomadic-pastoral society, it also served as an unwelcome reminder that Ibn Saud had conquered Jeddah, Mecca, and the surrounding territory in 1926. The Kingdom of Saudi Arabia was formed shortly afterward, in 1932—recent history in the 1970s. Furthermore, Jeddah had a long history as a port city with commercial and governmental structures consistent with the emerging national orientation toward international commerce.51 If formal connotations mattered, a local urban form would have offered a more symbolically appropriate choice. Finally, the government of Saudi Arabia wished to construe the hajj in apolitical terms, seeking to counter radical Arab nationalists who had recently attempted to embarrass the Saudi regime by disrupting the pilgrimage.52 Thus it made little sense to associate the project with traditional tents.

But there was another reason for the abandonment of SOM’s first fabric roof proposal—and here computation enters the scene. The story revolves around a murky episode involving the engineering consultant Horst Berger. In the spring of 1977, as Allen was first considering tensile structures, he invited Berger to SOM’s New York office to seek his advice.53 Berger had worked with Allen as a consultant before, so Allen knew that the company Berger ran with his partner, David Geiger, had become a leading authority on the design and engineering of fabric structures.54 Geiger specialized in air-supported structures: the United States Pavilion at the 1970 World Exposition in Osaka had been his breakout project, and his Pontiac Silverdome had opened two years before in Atlanta. Concurrently, Berger had developed a specialization in tensile structures. Many were similar to designs by Otto and others, although Berger also took an unusual approach. Otto’s largest projects were cable-net structures consisting of two distinct layers; the Munich stadium, for instance, featured acrylic panels attached above a structural mesh of cables. Berger instead focused on projects that interwove cables with fabric, merging enclosure with structure in a single sheet.

When Bunshaft took over the Hajj Terminal project, he confirmed that Berger would be the structural consultant for the giant fabric roof. Berger later recalled being “ecstatic—this was the reward for many years of patient preparation.”55 But then when Berger received no news for a few weeks, he contacted SOM and learned he would not be needed. What happened remained a mystery to Berger until the fabricator Owens-Corning contacted him a year later to ask if he would serve as the subconsultant for a major tensile roof project—which turned out to be a new scheme for the Hajj Terminal. Fazlur Khan, SOM’s star engineer in Chicago, assumed direction of the design of the roof structure, although Bunshaft was still directing the project as a whole from New York. Berger signed on, and eventually, while working on the project, he had a chance to have breakfast with Khan in Stuttgart. Khan described what had happened—this is how Berger remembered it years later:

Bunshaft had shown Khan his design concept. … Khan had told him that [it] would not work. Bunshaft’s response was: “You are an engineer, make it work!” Khan then tried to explain that the form of a tensile structure cannot be arbitrary; that a tensile structure won’t work unless you shape it correctly. Bunshaft didn’t like that; in fact he was furious. “Then you design it!” he yelled at Khan. “So I did,” he explained to me. That’s why the design had moved to Chicago. And that’s why I had never heard of it again until the time of the bid.56

The precise relationship between Khan’s and Berger’s designs remains murky in this account. Certainly Khan’s solution seemed strikingly similar to projects that Berger had worked on as an engineering consultant, including a modest tent for the Great Adventure amusement park in New Jersey (Figure 13).57 Berger had engineered two other widely known structures for the bicentennial celebrations in Philadelphia the year before, one of which featured a similar shape (Figure 14). A feature article on Berger’s work and “the discipline of tent structures” published in Architectural Record in 1975 exhibited several additional similar forms, including, as its very first image, a small wireframe view that was essentially identical to the units of the Hajj Terminal roof.58

Figure 13

Geiger Berger Associates, Great Adventure amusement park tent, Jackson, New Jersey, 1973 (Geiger Engineers, New York).

Figure 13

Geiger Berger Associates, Great Adventure amusement park tent, Jackson, New Jersey, 1973 (Geiger Engineers, New York).

Close modal
Figure 14

Geiger Berger Associates, Independence Mall Pavilion, Philadelphia, 1976 (Geiger Engineers, New York).

Figure 14

Geiger Berger Associates, Independence Mall Pavilion, Philadelphia, 1976 (Geiger Engineers, New York).

Close modal

According to the reasoning of all parties involved, the striking formal similarity between these roofs emerged as the result of similar materials, formal constraints, and engineering methods. It would have been absurd to accuse SOM of copying Berger’s design because “design” never really entered the picture. The text of Berger’s article in Architectural Record describes an engineer analyzing and solving a given problem in a rational way, without geometrical creativity. Engineers at SOM stretched fabrics into the same shapes as those in Berger’s work—the method dictated the form. When the introductory text to Berger’s article mentions that “there is more to [his] design than meets the eye,” the term design should be understood as a verb (a form-finding process), not as a noun (an iconic form that could potentially be someone’s intellectual property).59

With this caveat in mind, it is easy to imagine that Khan’s team considered Berger’s tensile form as a ready-made available to be appropriated. After all, Khan and other engineers at SOM sought something that would be guaranteed to work. Given that they had accepted a challenge from Bunshaft that strained their capacity to consolidate a new area of expertise on short notice, they did not have many options. Their sketch models revealed some variety, but the most common features—square grids repeating relatively small pyramidal units—had already been tested by Berger (Figure 15).

Figure 15

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, early fabric roof models, 1977 (SOM, Chicago).

Figure 15

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, early fabric roof models, 1977 (SOM, Chicago).

Close modal

However it happened, an element from deep within the labor hierarchy—a relatively standard element of tensile structural engineering—assumed an unlikely role at center stage in the Hajj Terminal project. An array of Berger’s Great Adventure tents became the project.

The key to understanding why SOM hired Berger, then let him go, then hired him again as a subconsultant on a project that closely resembled Berger’s own previous work has to do above all with software. Two sets of software, really: one at SOM and another at Geiger Berger Associates. Although the Hajj Terminal roof consisted of an array of units that in theory were fairly straightforward to construct, the project hinged on the risk of failure involved in the process of scaling up the size of the individual units and then repeating them 210 times in an unfamiliar environment. The only way to ensure that nothing would go wrong was to calculate the possible scenarios, and this required specialized software.

In the decade before the building of the Hajj Terminal, the design and engineering of tensile structures had encountered a computational bottleneck. Otto, the most prominent figure in the field, famously worked without digital computers—in the mid-1970s his example provided a cautionary tale. For the medium-size German Pavilion at the 1967 World Exposition in Montreal, Otto had used physical computation methods to great effect. His team carried out the engineering using elaborate models with tiny tension meters to measure the forces acting on dozens of scaled-down cables (Figure 16).60 When Otto served as a structural engineer on the similar but much larger roof for the Olympic Games in Munich in 1972, however, he discovered that these methods did not easily scale up. Despite his team’s incredible sophistication at building physical models, their measurements proved insufficiently precise to enable the construction of the enormous roofs.61 The project was ultimately built as envisioned, thanks to the arrival of engineers specializing in geodesy and finite element analysis—along with their digital computers.62 The resulting design process also proved to be very expensive—feasible only for a national prestige project.

Figure 16

Frei Otto, German Pavilion, World Exposition, Montreal, 1967, engineering model created at the Institute for Lightweight Structures, University of Stuttgart (FO_KB-P_1965-07_02-03-3, Südwestdeutsches Archiv für Architektur und Ingenieurbau, Karlsruhe).

Figure 16

Frei Otto, German Pavilion, World Exposition, Montreal, 1967, engineering model created at the Institute for Lightweight Structures, University of Stuttgart (FO_KB-P_1965-07_02-03-3, Südwestdeutsches Archiv für Architektur und Ingenieurbau, Karlsruhe).

Close modal

Compared to the Hajj Terminal, the design of the Munich Olympic Stadium was also simpler in one important way. Steel cables formed the primary structural element of the Olympic Stadium roof, creating a relatively even grid with acrylic panels mounted above for enclosure. SOM essentially decided to use fabric to replace the cables going in one direction in the terminal roof’s structural net. This made for a simpler-looking roof—collapsing fabric and cables into one layer—but it also made the engineering and construction much more difficult. SOM now had to contend with the interaction between two structural materials, one of which was new. Although Berger and a few others had used Teflon-coated fiberglass fabric for several years—and it was thus generally understood to be a viable material—the use of this fabric nonetheless represented a risky proposition. The piles of paperwork that SOM and its consultants produced in the design of the Hajj Terminal mostly involved the testing of various scenarios of fabric failure.63

In the course of design, SOM and Hochtief, the general contractor, made the crucial decision to insist that the contract with the fabric manufacturer include a ten-year guarantee, and Owens-Corning was the only manufacturer willing to take on that risk.64 This was considered highly unusual; engineering liability was usually located with the chief engineer, in this case SOM.65 With this decision, SOM off-loaded much of the engineering risk to Owens-Corning. However, Owens-Corning in turn contracted it out to others. Birdair, the firm headed by Walter Bird, the elder statesman of architectural fabric in the United States, performed strength testing. URS and Geiger Berger Associates performed engineering calculations. Contractors in Chicago built a detailed physical model and shipped it to the University of Western Ontario for wind tunnel testing. Stress calculations were conducted across sets of four units and sets of twenty-one units. Eventually the engineers deemed it prudent to build and test two of the roof units at full scale, a seven-month process. Owens-Corning produced a sleek manual that detailed every step of the process, from fabrication and testing to the packaging, shipping, and erection of the fabric, followed by testing again on-site (Figure 17).66

Figure 17

Owens-Corning Saudi, illustration from a series showing how to erect the fabric roof, 1978 (Prototype Testing Manual for NJIA Haj Terminal Fabric Roof System, 1978, SOM Box No. [Chicago/Struct] SO373-2, SOM, Chicago).

Figure 17

Owens-Corning Saudi, illustration from a series showing how to erect the fabric roof, 1978 (Prototype Testing Manual for NJIA Haj Terminal Fabric Roof System, 1978, SOM Box No. [Chicago/Struct] SO373-2, SOM, Chicago).

Close modal

Even if it was clear from the beginning that the roof would eventually be made to work, the realization of such a structure posed a daunting computational task. This is where Berger and his software came in. While Otto relied on a platoon of students and an ad hoc assemblage of experts, Berger and Geiger built a small private practice in New York by automating the computational process of engineering tensile structures. In the mid-1970s, Geiger Berger Associates had developed unique software for this purpose. Beginning with an approximate shape (provided by SOM in the case of the Hajj Terminal) and prestress levels for the fabric and cables, the software ran a series of iterative calculations to determine the shape’s distortion under stress as well as its final resting form. This final model of the shape and its stresses provided the basis for engineering calculations (performance under wind loads, failure scenarios, and so on). This iterative process required significant computational effort given the large number of points. SOM provided Geiger Berger Associates with a digital mesh model consisting of about two hundred points and requiring calculations intensive enough that the software needed to be run on a new Cray supercomputer owned by a local time-sharing firm. Beyond these calculations, Geiger Berger Associates’ software also automated the crucial step of outputting patterning dimensions for the cutting of the fabric (Figure 18).

Figure 18

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, computer-generated node map of fabric roof panel produced by Geiger Berger Associates, 1979 (SOM Box No. [Chicago/Struct] SO374-2, SOM, Chicago).

Figure 18

Skidmore, Owings & Merrill, Hajj Terminal, King Abdulaziz International Airport, Jeddah, Saudi Arabia, computer-generated node map of fabric roof panel produced by Geiger Berger Associates, 1979 (SOM Box No. [Chicago/Struct] SO374-2, SOM, Chicago).

Close modal

Berger’s 1975 article in Architectural Record informed architects that Geiger Berger Associates possessed this unique capability, and many probably found it reassuring that the firm confidently offered a rather adventurous design as its primary example. A series of screenshots of a digital model of an amphitheater in Florida (designed by the architect William Morgan) resembling, in its amorphous shape, something like the Munich Olympic Stadium, filled fully half the space of the article (Figure 19).67 The text detailed the process of designing such a roof with Geiger Berger Associates’ software, noting that the engineers dismissed older methods of using “soap bubbles, stretched fabrics, and other modeling materials” as “time-consuming,” “approximate,” and “interesting.” This positioned the firm’s software as a more efficient and rigorous alternative to Otto’s academic experimentation.68

Figure 19

Geiger Berger Associates, screenshot of an amphitheater designed by William Morgan, 1975 (Horst Berger, “The Engineering Discipline of Tent Structures,” Architectural Record, Feb. 1975).

Figure 19

Geiger Berger Associates, screenshot of an amphitheater designed by William Morgan, 1975 (Horst Berger, “The Engineering Discipline of Tent Structures,” Architectural Record, Feb. 1975).

Close modal

A sidebar accompanying the article mentioned that a student at Columbia University, where Geiger was teaching, produced the evocative screenshots. This serves as a reminder that the U.S. government’s enormous investment in computer technology in the decades after World War II had, by the 1970s, trickled down into an ecosystem of small computer firms.69 Some programmers founded time-sharing firms that bought expensive computer hardware and leased out computation time. Geiger Berger Associates used United Computer Services, located on Sixth Avenue in New York.70 The proliferation of specialized technical consultancies defined the “go-go years” of the “software boom.”71 Architects and engineers joined in on the action, and in the prevailing mood of optimism they provided services ranging from automating specifications to creating spatial planning simulations to producing renderings.72 Geiger Berger Associates consulted on the engineering of fabric structures for numerous firms, working on projects that included pavilions, department stores, stadia, and other airports.73

It is worth emphasizing that there was nothing mysterious in the process developed by Geiger Berger Associates—the firm’s calculations matched those used by Otto, and engineers at SOM used the same methods as well. The paperwork for the Hajj Terminal project records page upon page of calculations circulating between engineers and designers at SOM, their subconsultants, and manufacturers. But some steps within this process had been codified into unique software by one firm or another. At an early stage, SOM used DRAW3D, a drafting package the firm had been developing for several years, which featured a lofting algorithm to generate the shape of the roof. DRAW3D was used to produce many of the images of the project that engineers found so alluring.74 At a later point, SOM used small pieces of engineering software that it had developed alongside a standard structural engineering program, STRUDL.75 Geiger Berger Associates used its own software, and the testing devices employed by Birdair had their own software as well. Nothing prevented SOM from developing these capabilities internally; the equations and procedures were well known, and the Computer Group at SOM, initially under Khan’s supervision, led innovation in software for architects.76 But it was probably prudent for SOM to allow some areas of expertise to remain external. In these years, the firm worked with IBM to develop an architectural software package, SKYLINE, that incorporated too much and proved to be too sophisticated. SKYLINE ended up being priced at roughly twenty times what its rival AutoCAD cost, and it never took off.

Because it distributed expertise among other firms, SOM encountered difficulties projecting an appearance of control, and it had to forgo some of the glory associated with the Hajj Terminal roof. Before the terminal was constructed, an article in Architectural Record on “tent structures designed to endure” included an ambiguous discussion of who had done what work on the project, prompting Khan to write a letter to the editor to set the record straight, clarifying that “the structural concept, design and engineering was developed by SOM under [his] personal supervision.”77 The article identified SOM ambiguously as the “architects and engineers” of the project while naming Geiger Berger Associates clearly (and wrongly) as the “structural engineers.”78 Even with this correction, however, there was no changing the overall arc of the article’s dramatic narrative: when SOM hired Owens-Corning to supply the roof, Owens-Corning turned to Geiger Berger Associates with a difficult problem that only “computer analysis” could solve. The writer at Architectural Record correctly picked up on a key moment that SOM would have preferred to have kept hidden beneath its corporate umbrella.

This episode provoked a minor existential crisis at SOM. In response to the article and a conversation with Owens-Corning, John Zils, the project engineer at SOM for the Hajj Terminal, wrote a letter in which he offered a detailed timeline of the structural engineering of the roof and a list of credits he clearly hoped would be definitive, including the following:

Architect:

Skidmore, Owings & Merrill

Structural Engineer:

Skidmore, Owings & Merrill

Mechanical Engineer:

Skidmore, Owings & Merrill

[…]

Engineering Consultant to Owens-Corning Saudi:

URS/Corporation,             Geiger-Berger Associates

Zils relegated Geiger Berger Associates to a decidedly inferior position in several respects (including rendering its name incorrectly, with a hyphen) and concluded somewhat plaintively that, “in summary, SOM did develop and perform the initial design to include rigorous and complete structural engineering analysis and design for the tent roof.”79 In this affirmation, we sense that some of the real work happened elsewhere.

Khan and Zils also seem to have missed the point. SOM had vast in-house engineering capabilities, to be sure, but asserting “complete structural engineering” authority suggests a blindness to what SOM lacked, as well as the significance of that lack. Geiger Berger Associates possessed unique software, and the ability to draw upon that software within an ad hoc team for the Hajj Terminal project represented a crucial aspect of SOM’s operation and its success as a corporate firm. In this sense, SOM did not need to be able to do everything—the idea was not consolidation but coordination.

Taking a wider view, however, the wrangling around who deserved credit for the structural engineering of the Hajj Terminal roof may as well have been left behind the scenes. For the most part, the general media attributed the project to Khan, as the star engineer.80 Bunshaft also received credit as the architect in charge of the airport project as a whole (even if its most interesting aspect turned out to be more a matter of engineering than of architecture).81 Within the field of structural engineering, Berger managed to claim a share of glory for himself. In a monograph of his work, he presented the project in a sufficiently ambiguous way to give at least some readers the impression that the project was essentially his.82

In the mid-1970s, the genius of Geiger Berger Associates was to package all this tensile engineering stagecraft, produce a compelling image for it, and offer it on the market to architects as a specialized consulting practice. SOM’s genius, in turn, was to draw a loose and heterogeneous team together while keeping SOM and its principals at the head of the credits (at least most of the time).

Most important for the longer sweep of twentieth-century architecture, this process also involved the packaging of a distinct concept of “the computer.” The computer-as-calculator prevailed as the most widespread notion of computers in the era.83 In his later recollection of the design of the Hajj Terminal, Bunshaft expressed a typical outlook in this regard. In an in-depth oral history interview conducted in 2000, he mentioned computers only twice, and one of these was when he acknowledged that “without advanced computers, [the Hajj Terminal] could never have happened.”84 He was aware that computation played a crucial role in the design, but he seemed unsure exactly how this took place—he knew only that Khan and Owens-Corning had scores of people “working on it” with their engineering equations and calculating machines.85

Other concepts of the computer vied with that of the computer-as-calculator. Many revolved around the notion of the computer as a tool of one sort or another. In architecture trade periodicals in the 1970s, computers appeared in the sections on office supplies and equipment. When computers showed up in the context of design, spreadsheets often provided a means to visualize them, as at SOM around 1970.86 By the early 1980s, drafting software modeled on common drafting tools dominated the architectural imagination.87 In one sense, these tropes showed the limits of imagination, as architects assumed the essential equivalence of computers to the tools they already used.88 In another sense, however, they showed the persistence of imagination. Conceptual figures—or “images” in psychoanalytical terminology89—do not always disappear even when they have become obsolete. Sometimes they are repurposed: drafting tables become drafting software, paper ledgers become digital spreadsheets, handheld calculators become calculating programs.90

Geiger Berger Associates’ screenshots of glowing spiderweb forms depicted in evocative perspective views offered a distinct concept of “the computer” to accompany these other ideas. Geiger Berger Associates presented the computer as a simulation environment where geometrical forms suffused with engineering intelligence could be directly manipulated. This trope goes by other names, such as “the interactive computer” and “digital modeling.” Around 1980, this signified a remediation of normal physical modeling practices in architecture as well as implying a sort of physical computation akin to Antoni Gaudí’s upside-down catenary models of the Sagrada Família and Otto’s elaborate models of tensile structures. It was these types of models and their associated techniques that were transferred into the computer. This was also the computational imaginary at work in the first computer-aided design software, Sketchpad, meant to assist in the design of engineered forms. Geiger Berger Associates’ images publicized not only the firm’s expertise in structural engineering but also a specific way for architects to conceptualize computers.

By the same token, these images also revealed a distinct way of conceptualizing architecture. The images produced by Geiger Berger Associates point to a temporary loophole in the labor hierarchy that produced the Hajj Terminal, one that allowed the ideal of the anonymous engineer to be valued above the ideal of the creative designer; the firm’s software acted as an intermediary that standardized the process of arriving at compelling forms through the application of impersonal physical laws. Software not only cut the genius architect out of the process but also eliminated the architect’s competitor, the star engineer. Producing compelling forms not attributed to anyone in particular (who can claim authorship of a form derived from nature?), software could arrive at these forms without making any momentous decisions. Recall that the Hajj Terminal was not so much actively “designed” as passively “arrived at” through a process of elimination and the churn of bureaucratic engineering paperwork.

According to the ideology of form-finding, in order to appear “undesigned” and “natural,” a project should be amorphous. Geiger Berger Associates produced asymmetrical and idiosyncratic screenshots, making visible the way in which external constraints affected the shape of a roof. In a similar vein, Khan did not claim authorship of the Hajj Terminal but described experimenting with “the infinite possibility of creating forms simply by pushing and pulling at different locations [on a piece of fabric] to achieve two-directional curvature.”91 Form-finding allowed for the creation of shapes that appeared to be anonymous but were still interesting enough to be iconic.92

Modeling software thus became a design partner, just as the pioneers of computer-aided design had dreamed.93 But this was a partner of a very specific type: ingeniously inventive, rigorously mathematical, and fastidiously bureaucratic. Its operator—in this case Berger—took on similar characteristics, as both a charismatic spokesperson and a paper-pushing engineer. In the resulting reconfiguration of architectural values, paperwork moved up a notch while the genius designer was demoted. In fact, the two could even sometimes reverse positions, as in cases when a compelling project originated in bottom-up form-finding rather than top-down decisiveness.

This reconfiguration of values provides us an opportunity to rethink Max Weber’s classical sociological distinction between charismatic authority and bureaucratic authority. Genius and bureaucracy can be conceived as moments that occur in succession rather than as polar opposites. Charismatic authority, Weber says, is “authority resting on devotion to the exceptional sanctity, heroism, or exemplary character of an individual person, and of the normative patterns or order revealed or ordained by him.”94 In the succession of government types that Weber describes, government devolves into rigid officialdom once the charismatic leader exits the scene. Bureaucracy maintains order, but this order preceded bureaucracy in the leader who “revealed” it. The relationship between charismatic authority and bureaucracy can be described as an oscillation between revelatory magic and routine maintenance.

Returning to the Hajj Terminal, at least three charismatic moments reveal three orders that prevailed in architecture ca. 1980. The first had to do with form-finding, as a way of configuring architecture around a specific type of modeling software (“smart” modeling software, combining interactive manipulation with engineering calculation). Another order involved a certain vision of the relationship between human subjects and their technological environment. The Hajj Terminal belongs alongside, for example, Superstudio’s Supersurface and Archigram’s many plug-in projects (Figure 20). What these share is a relationship between autonomous nomadic individuals and the vast technological infrastructure that supports them. This pervasive Western vision of the 1960s fell out of favor as a result of the widespread technophobic backlash of the 1970s.95 Importantly, the figure of the tent survived this mood shift because tents escaped overtly “technological” connotations. More than other projects of the era, lightweight structures were appealing because they revealed the desire of the “free” individual situated in relationship to a comprehensive catalogue of practices that were supposedly timeless but at the same time were uniquely cutting-edge. This strange conflation of timelessness and modernity appears clearly in Otto’s drawing of the global evolution of structural systems, in which tents from cultures around the world appear as a “universal” type but nevertheless undergo final development in the hands of contemporary engineers (Figure 21). The related complicity between “counterculture” and “cyberculture” was represented and enabled by such publications as the Whole Earth Catalog.96 Gilles Deleuze memorably fashioned the resulting subject as the “nomad,” though it could be more simply associated with the cosmopolitan, jet-setting consultant of exactly the type enabled by computational practices.97 It is crucial to note that computers, although rarely seen, played a key role in this imaginary. They were the devices—construed almost as a force of nature—that enabled the vast infrastructure to operate.98 Beneath the Supersurface, presumably, stands the mainframe. The Hajj Terminal aptly thematizes this order: the relationship between a global technological infrastructure and a specific contemporary subjectivity.

Figure 20

Superstudio, Life–Supersurface–“Fruits & Wine,” 1971 (AM2000-2-156, Musée National d’Art Moderne, Paris).

Figure 20

Superstudio, Life–Supersurface–“Fruits & Wine,” 1971 (AM2000-2-156, Musée National d’Art Moderne, Paris).

Close modal
Figure 21

Frei Otto, drawing of the history of building construction, ca. 1980 (Südwestdeutsches Archiv für Architektur und Ingenieurbau, Karlsruhe).

Figure 21

Frei Otto, drawing of the history of building construction, ca. 1980 (Südwestdeutsches Archiv für Architektur und Ingenieurbau, Karlsruhe).

Close modal

The third sociotechnical order revealed by the Hajj Terminal had to do with the profession of architecture as exemplified by the corporate structure of SOM. It is significant that the software complications outlined above were relegated to obscurity. In the end, the Hajj Terminal project reverted to star designers: Khan (or sometimes Berger), the star engineer who came up with an ingenious form based on the universal truths of engineering; and Bunshaft, who held to a vision of simple, compelling forms. Untold piles of paperwork assumed a new identity as heroic design.99 And to a certain extent, the corporate body of SOM obscured Khan and Bunshaft themselves. All of this conceptual packaging suggests a way to think about expertise in the context of corporate architecture. SOM’s reputation for reliability depended on the firm’s allowing ideas to emerge from within internal project teams made up of architects and technicians as well as various dispersed external experts, and then putting these ideas center stage in important projects. Ideas could also be ushered offstage again when necessary to avoid a messy battle of wills. Anyone could come up with an idea. When convenient, it might be attributed to Bunshaft or Khan, but ultimately the idea simply belonged to SOM. When the iconic screenshot of the Hajj Terminal was reproduced time and again, it was not to celebrate form-finding or to give credit to Berger, Khan, or anyone else; rather, the image served as a useful tool for advertising the firm’s unique computational (and, by extension, organizational) expertise. SOM presented an image of expert stagecraft, deftly packing the actual behind-the-scenes work of numerous experts into its corporate black box.

By allowing compelling form to arrive from anywhere within the labor hierarchy, SOM was able to offer excellent design without the usual risks that genius entails. SOM refined the ideal of bureaucratic architecture expounded by Hitchcock in the mid-1940s.100 Bureaucratic architecture attested not to routine banality but to processes that allowed charismatic forms to emerge out of a system of rules and paperwork as opportunities presented themselves.

1.

Oddly enough, Hitchcock did not mention SOM by name. Henry-Russell Hitchcock, “The Architecture of Bureaucracy and the Architecture of Genius,” Architectural Review 101, no. 601 (Jan. 1947), 3–6.

2.

Stanford Anderson, foreword to Gordon Bunshaft of Skidmore, Owings & Merrill, by Carol Herselle Krinsky (Cambridge, Mass.: MIT Press, 1988), ix.

3.

See, for example, Reinhold Martin, The Organizational Complex: Architecture, Media, and Corporate Space (Cambridge, Mass.: MIT Press, 2003).

4.

Compare this approach with that of The Architects Collaborative (TAC), which was more radical than SOM in its style of collaboration. Michael Kubo, “The Anxiety of Anonymity: Bureaucracy and Genius in Late Modern Architecture Industry,” in New Constellations/New Ecologies: Proceedings of the 101st Annual Meeting of the ACSA, ed. Ila Berman and Edward Mitchell (Washington, D.C.: ACSA, 2013), 810–17.

5.

“SOM’s Computer Approach,” Architectural Record, Aug. 1980, 84–91.

6.

Nicholas Adams, Gordon Bunshaft and SOM: Building Corporate Modernism (New Haven, Conn.: Yale University Press, 2019), 1–5.

7.

Bunshaft’s role models were Ludwig Mies van der Rohe, Le Corbusier, and Frank Lloyd Wright, three of the four architects who were becoming canonical “modern masters” in this period. Carol Herselle Krinsky, Gordon Bunshaft of Skidmore, Owings & Merrill (Cambridge, Mass.: MIT Press, 1988), 3–4; Edwin Hoag and Joy Hoag, Masters of Modern Architecture: Frank Lloyd Wright, Le Corbusier, Mies van Der Rohe, and Walter Gropius (Indianapolis: Bobbs-Merrill, 1977).

8.

Jay Wickersham, “Learning from Burnham: The Origins of American Architectural Practice,” Harvard Design Magazine 32 (2010), 18–27.

9.

See, for example, Michael Kubo, “The Concept of the Architectural Corporation,” in OfficeUS Agenda, ed. Eva Franch i Gilabert, Amanda Reeser Lawrence, Ana Miljacki, and Ashley Schafer (Zurich: Lars Müller, 2014), 37–48; Ann Lok Lui, “Experimental Logistics: Extra-architectural Projects at SOM, 1933–1986” (master’s thesis, Massachusetts Institute of Technology, 2011).

10.

For an early study, see Peter F. Drucker, Concept of the Corporation (New York: John Day, 1946).

11.

Thomas Walton, Architecture and the Corporation: The Creative Intersection (New York: Macmillan, 1988).

12.

“SOM’s Computer Approach,” 84.

13.

Skidmore, Owings & Merrill, SOM’s Computer Capability (Chicago: Skidmore, Owings & Merrill, 1980), 8–9.

14.

“SOM’s Computer Approach,” 84.

15.

This is standard obfuscation about the supposed mysterious essence of architecture. See, for example, Reyner Banham, “A Black Box: The Secret Profession of Architecture,” New Statesman and Society, Oct. 1990, 22–25.

16.

This is akin to the logical process of “abduction”—the type of speculative leap characteristic of the archetypal detective (e.g., Sherlock Holmes). Igor Douven, “Abduction,” in Stanford Encyclopedia of Philosophy, last revised 18 May 2021, https://plato.stanford.edu/entries/abduction (accessed 20 July 2021). See also Lionel March, “Introduction: The Logic of Design and the Question of Value,” in The Architecture of Form, ed. Lionel March (Cambridge: Cambridge University Press, 1976), 1–40. It should be noted that historians routinely operate in this way. See, for example, Ilkka Niiniluoto, “Defending Abduction,” Philosophy of Science 66 (1999), S436–51.

17.

Neil Garston, “The Study of Bureaucracy,” in Bureaucracy: Three Paradigms, ed. Neil Garston (Boston: Kluwer, 1993), 5.

18.

Max Weber, “Bureaucracy,” in Weber’s Rationalism and Modern Society: New Translations on Politics, Bureaucracy, and Social Stratification, trans. and ed. Tony Waters and Dagmar Waters (Basingstoke: Palgrave Macmillan, 2015), 74.

19.

David Graeber, The Utopia of Rules: On Technology, Stupidity, and the Secret Joys of Bureaucracy (New York: Melville House, 2015), 176–78.

20.

David Wengrow, “Archival and Sacrificial Economies in Bronze Age Eurasia: An Interactionist Approach to the Hoarding of Metals,” in Interweaving Worlds: Systemic Interactions in Eurasia, 7th to the 1st Millennia BC, ed. Toby C. Wilkinson, Susan Sherratt, and John Bennet (Oxford: Oxbow Books, 2011), 135–44.

21.

Ralph Kingston, Bureaucrats and Bourgeois Society: Office Politics and Individual Credit in France 1789–1848 (New York: Palgrave Macmillan, 2012), 1.

22.

A key precedent for such accounts being Franz Kafka’s 1925 novel The Trial.

23.

See, for example, James R. Beniger, The Control Revolution: Technological and Economic Origins of the Information Society (Cambridge, Mass.: Harvard University Press, 1986).

24.

“105-Acre Fabric Roof Will Shelter Moslem Pilgrims at Jeddah,” Architectural Record, Dec. 1978, 37.

25.

“Invitation to the Haj,” Progressive Architecture 63, no. 2 (Feb. 1982), 116–22.

26.

Krinsky, Gordon Bunshaft, 331–34.

27.

On the notion of “the project,” see Antoine Picon, “The Ghost of Architecture: The Project and Its Codification,” Perspecta 35 (2004), 8–19.

28.

The SOM projects closest in size to the airport were the Oak Ridge New Town in the 1940s and the U.S. Air Force Academy in the 1950s.

29.

A more complete list of the program can be found in Yasmin Sabina Khan, Engineering Architecture: The Vision of Fazlur R. Khan (New York: W. W. Norton, 2004), 288.

30.

“Chronology: March 1, 1965–May 31, 1965,” Middle East Journal 19, no. 3 (1965), 348.

31.

Aleksandr Bierig, “AIA Honor Award 2010: 25 Year Award—Hajj Terminal,” Architectural Record, June 2010, 83–128.

32.

For an excellent summary of the politics and economics of this period, see Khan, Engineering Architecture, 267–70.

33.

On the ideology of “developmentalism,” see Tony Smith, “Requiem or New Agenda for Third World Studies?,” World Politics 37, no. 4 (July 1985), 533–34.

34.

Donald McNeill, The Global Architect: Firms, Fame and Urban Form (London: Routledge, 2008).

35.

Alfred M. Kemper, ed., Pioneers of CAD in Architecture (Pacifica, Calif.: Hurland/Swenson, 1985).

36.

Bierig, “AIA Honor Award 2010.”

37.

The Washington, D.C., office was largely subsidiary to the New York office. John Zils, SOM project engineer for the Hajj Terminal, interview by author, 8 Jan. 2019.

38.

This was “programming” in the classic architectural sense.

39.

An earlier, undeveloped scheme separated the terminal from the holding area. See “Invitation to the Haj.”

40.

Sandy Isenstadt and Kishwar Rizvi, “Introduction: Modern Architecture and the Middle East; The Burden of Representation,” in Modernism and the Middle East: Architecture and Politics in the Twentieth Century, ed. Sandy Isenstadt and Kishwar Rizvi (Seattle: University of Washington Press, 2008), 3–36.

41.

Zils, interview by author.

42.

Krinsky, Gordon Bunshaft, 264.

43.

Krinsky, 264; Zils, interview by author.

44.

Gordon Bunshaft, “Oral History of Gordon Bunshaft,” interviewed by Betty J. Blum, 2000, rev. ed., transcript, 198, Chicago Architects Oral History Project, Ernest R. Graham Study Center for Architectural Drawings, Department of Architecture, Art Institute of Chicago.

45.

Krinsky, Gordon Bunshaft, 264.

46.

Philip Drew, Tensile Architecture (Boulder, Colo.: Westview Press, 1979). One of the most iconic books in this genre is Lloyd Kahn’s Domebook One (Bolinas, Calif.: Pacific Domes, 1971).

47.

The Bedouin of the Arabian Desert, with their black tents, were charismatic figures of the popular imagination. See, for example, Donald Powell Cole, Nomads of the Nomads: The Al Murrah Bedouin of the Empty Quarter (Chicago: Harlan Davidson, 1975).

48.

Philip Drew, Frei Otto: Form and Structure (Boulder, Colo.: Westview Press, 1976).

49.

See, for example, Daniela Fabricius, “Material Models, Photography, and the Threshold of Calculation,” Arq 21, no. 1 (2017), 21–32.

50.

Wildermuth recalled in a recent interview that Saudi officials expected a fully enclosed “building.” Gordon Wildermuth, interview by Aleksandr Bierig, 2010. (I would like to thank Aleksandr Bierig for sharing his transcript of the interview.) Nicholas Adams also notes that “acceptance of the shift to a tent … was not immediate.” Nicholas Adams, Skidmore, Owings & Merrill: SOM since 1936 (Milan: Electa, 2007), 263.

51.

David E. Long and Sebastian Maisel, The Kingdom of Saudi Arabia, 2nd ed. (Gainesville: University Press of Florida, 2010), 1–24.

52.

Long and Maisel, 60.

53.

Horst Berger, Light Structures, Structures of Light: The Art and Engineering of Tensile Architecture (Bloomington, Ind.: Authorhouse, 2005), 77.

54.

Another authority was Walter Bird, whose firm, Birdair, was the oldest; Birdair, however, did not have the same “design” focus as Berger and Geiger’s firm. See, for example, “Tomorrow’s Life Today,” Life, Nov. 1957.

55.

Berger, Light Structures, Structures of Light, 77.

56.

Berger, 77.

57.

On this and other projects by Berger, see Berger, Light Structures, Structures of Light.

58.

Horst Berger, “The Engineering Discipline of Tent Structures,” Architectural Record, Feb. 1975, 81–88.

59.

Recall also Karl Marx: “A spider conducts operations that resemble those of a weaver, and a bee puts to shame many an architect in the construction of her cells. But what distinguishes the worst architect from the best of bees is this, that the architect raises his structure in imagination before he erects it in reality.” Karl Marx, “The Labour-Process and the Process of Producing Surplus Value,” in Capital, vol. 1 (1867), https://www.marxists.org/archive/marx/works/1867-c1/ch07.htm (accessed 20 July 2021).

60.

Fabricius, “Material Models, Photography.”

61.

Wolfgang Brand, “Designing the Membrane Roof of the Munich Olympic Stadium Using Supercomputers,” in Computer Simulations and the Changing Face of Scientific Experimentation, ed. Juan M. Durán and Eckhart Arnold (Cambridge: Cambridge Scholars, 2013), 201–32.

62.

On this fascinating story, see Sean Keller and Christine Mehring, Munich ’72: Olympic Art and Architecture (New Haven, Conn.: Yale University Press, forthcoming).

63.

One of the fabric panels did rip while the roof was under construction, and then the paperwork really piled up.

64.

Zils, interview by author.

65.

Berger, Light Structures, Structures of Light, 79.

66.

Owens-Corning Saudi, Prototype Testing Manual for NJIA Haj Terminal Fabric Roof System, 1978, SOM Box No. (Chicago/Struct) SO373-2, SOM Archive, Chicago.

67.

Berger, “Engineering Discipline of Tent Structures.”

68.

Berger, 82.

69.

Martin Campbell-Kelly, From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry (Cambridge, Mass.: MIT Press, 2003), 29–55.

70.

Berger, Light Structures, Structures of Light, 81.

71.

Campbell-Kelly, From Airline Reservations to Sonic the Hedgehog.

72.

Eric Teicholz, “Architecture and the Computer,” Architectural Forum, Sept. 1968, 58–61; Allen Bernholtz and Steve Fosburg, “Spatial Allocation in Design and Planning,” in Proceedings of the 9th Design Automation Workshop (New York: ACM, 1972), 181–89; Matthew Allen, “Arata Isozaki and the Invisible Technicians,” CCA: Origins of the Digital, 2016, https://www.cca.qc.ca/en/issues/4/origins-of-the-digital/40596/arata-isozaki-and-the-invisible-technicians (accessed 20 July 2021).

73.

See, for example, “Tent Structures Designed to Endure,” Architectural Record, mid-Aug. 1979, 86–93.

74.

Kristine K. Fallon, “Early Computer Graphics Developments in the Architecture, Engineering, and Construction Industry,” IEEE Annals of the History of Computing 20, no. 2 (1998), 20–29.

75.

For context, see David E. Weisberg, The Engineering Design Revolution: The People, Companies and Computer Systems That Changed Forever the Practice of Engineering (n.p.: 2005), 5–7, http://www.cadhistory.net (accessed 20 July 2021).

76.

Nicholas Adams, “Creating the Future (1964–1086): How a Passionate Group of SOM Architects and Engineers Came Together to Envision Their Profession through the Lens of Technology,” SOM Journal 8 (2013), 120–36; Lui, “Experimental Logistics.”

77.

Fazlur Khan to the editor of Architectural Record, 31 Aug. 1979, SOM Box No. (Chicago/Struct) SO693-2, SOM Archive, Chicago.

78.

“Tent Structures Designed to Endure,” 86–93.

79.

John Zils to Robert Kline of Owens-Corning, 12 June 1980, SOM Box No. (Chicago/Struct) SO693-2, SOM Archive, Chicago.

80.

See, for example, “Invitation to the Haj.”

81.

The airport is clearly considered Bunshaft’s project in the first monograph devoted to his works. Krinsky, Gordon Bunshaft.

82.

Berger, Light Structures, Structures of Light, 77–91.

83.

The computer-as-calculator was a more mundane version of the computer-as-electronic-brain, which was a common trope of the Cold War era. Paul N. Edwards, The Closed World: Computers and the Politics of Discourse in Cold War America (Cambridge, Mass.: MIT Press, 1996).

84.

Bunshaft, “Oral History,” 201.

85.

Bunshaft’s understanding of computers as calculating brains was typical in its broad contours, but he added an unusual twist. Strangely enough, the one other mention of computers in his oral history is in relation to his own brain. Explaining why architects should absorb as many visual precedents as they can—whether “natural, paintings, sculpture, graphics, [or] architecture”—he muses that “our brain, a computer, absorbs all this whether we’re aware of it or not. When you get down to doing some work, the more you see, the more this computer may throw out. You don’t say, ‘I want to do the cathedral at Chartres,’ you’ll just subconsciously have more choices than you would if your computer was half-empty.” Bunshaft, 61–62. Bunshaft’s computer-brain, it seems, was less a calculator than a database of formal precedents with some sort of mysterious algorithm for offering them up randomly at opportune moments.

86.

G. Neil Harper, “Highrise Office Design: It Can Be Systematic,” AIA Journal, Mar. 1971, 41–44. It has been suggested that Bunshaft’s denigration of office buildings as “a mathematical calculation, … three dimensional investments” originated in his experience with SOM’s Building Optimization Program, which was essentially spreadsheet software for designing skyscrapers. Adams, “Creating the Future,” 122.

87.

SOM developed its own drafting software (DRAW2D and DRAW3D), which saved a significant amount of time on the Hajj Terminal project. Kristine Fallon describes this in “The Genesis of Digital Design at SOM,” 2012, SOM website, https://www.som.com/news/the_genesis_of_digital_design_at_som (accessed 20 July 2021). See also Fallon, “Early Computer Graphics Developments.”

88.

This is Christopher Alexander’s argument. Christopher Alexander, “A Much Asked Question about Computers and Design,” in Architecture and the Computer: Proceedings of the First Boston Architectural Center Conference, December 5, 1964 (Boston: Boston Architectural Center, 1964), 52–54.

89.

Jacques Lacan, “The Mirror Stage as Formative of the I Function as Revealed in Psychoanalytic Experience,” in Écrits (New York: W. W. Norton, 1977), 73–81.

90.

This is evidence in support of Marshall McLuhan’s dictum that “the ‘content’ of a medium is always another medium” and Lev Manovich’s observation that the advent of personal computers and packaged media software gave birth to a new conceptual figure: the computer as “metamedium.” Marshall McLuhan, Understanding Media: The Extensions of Man (New York: McGraw-Hill, 1964); Lev Manovich, Software Takes Command (New York: Bloomsbury Academic, 2013).

91.

Fazlur Khan, quoted in “Invitation to the Haj.”

92.

For a good description of this peculiar type of “monumentality,” see Sean Keller, Automatic Architecture: Motivating Form after Modernism (Chicago: University of Chicago Press, 2018), 99–146.

93.

See, for example, Steven Coons, “Computer Aided Design,” in Architecture and the Computer, 26–28.

94.

Max Weber, Economy and Society: An Outline of Interpretive Sociology (New York: Bedminster Press, 1968), 215, originally published as Wirtschaft und Gesellschaft: Grundriß der verstehenden Soziologie (Tübingen: Mohr, 1922).

95.

For an instructive case study in art, see Luke Skrebowski, “All Systems Go: Recovering Hans Haacke’s Systems Art,” Grey Room, no. 30 (Winter 2008), 54–83.

96.

Fred Turner, From Counterculture to Cyberculture: Stewart Brand, the Whole Earth Network, and the Rise of Digital Utopianism (Chicago: University of Chicago Press, 2006).

97.

The following two publications reveal the shift in Deleuze’s thinking about freedom and control: Gilles Deleuze and Félix Guattari, A Thousand Plateaus: Capitalism and Schizophrenia, trans. Brian Massumi (Minneapolis: University of Minnesota Press, 1987); Gilles Deleuze, “Postscript on the Societies of Control,” October 59 (Winter 1992), 3–7.

98.

This insight is from Antoine Picon, Digital Culture in Architecture: An Introduction for the Design Professions (Basel: Birkhäuser, 2010).

99.

One letter in the archives describes how SOM’s New York office dealt with a pileup of files on the office stairways: it burned the files. See SOM Box No. (Chicago/Struct) SO359-1, SOM Archive, Chicago.

100.

John Summerson, “Bread & Butter and Architecture,” Horizon 6 (Oct. 1942), 233–43; Hitchcock, “Architecture of Bureaucracy.”