Case Study: The MIT Media Lab

The MIT Media Lab (MIT ML) also has a long history of capitalizing on its appearance of newness as a way to assert its ontological newness, despite the fact that it too draws on the long genealogy that informed modern labs and informs many contemporary hybrid labs.43 The MIT ML is particularly important for our book as a whole because it opened in 1984 as one of the first university-based (but largely privately funded) interdisciplinary media labs dedicated to hands-on practice. Partly housed in a school of architecture, and designed by internationally renowned architecture firm I.M. Pei and Partners, this lab more than any others was intended to be spatially responsive to and reflective of the work going on inside it. The MIT ML lab exemplifies the ways in which hybrid labs in general and media labs in particular are distinct and porous, private and public in their spatial configurations, simultaneously embracing and rejecting their historical precedents.

The MIT ML is a complex entity, consisting of a many labs grouped under the singular name of “the media lab;” it is, in fact, more a lab of labs than a single entity. It is constantly evolving and has exerted considerable influence on the creation of media labs around the world. Our analysis of the naming and space of the MIT ML, however, must by necessity be focused and brief in this chapter, although we discuss it once more in our fifth chapter in terms of the history of its founding, its leadership and its politics. Our discussion of the MIT ML in this chapter focuses on the exterior design and interior space of the first media lab building that was completed in 1985 (building E15), with some attention to the second building that was completed in 2010 (building E14) as a way to set the stage for our later, more extended analysis.

The delineation of the MIT Media Lab as a lab begins with a heady mixture of institutional policy, emergent cultural techniques, lab discourse and the ability of those in power to bring something into existence by naming it. In the early 1980s, Nicholas Negroponte, co-founder and then-Director of the Architecture Machine Group (AMG) at MIT, began working with then-MIT President Jerome Weisner to launch a lab that built on the ethos and the success of the AMG but that also embraced interdisciplinary collaboration and rapid prototyping of communication technologies rather than relying on a more staid, methodical, theoretically and historically informed academic approach. As Stewart Brand put it in Inventing the Future, his popular mid-1980s account of the lab:

Students and professors at the Media Laboratory write papers and books and publish them, but the byword in this grove of academe is not “Publish or Perish.” In Lab parlance it’s “Demo or Die”—make the case for your idea with an unfaked performance of it working at least once, or let somebody else at the equipment. “We write about what we do,” comments Director Negroponte, “but we don’t write unless we’ve done it.”44

Naming this entity “the Media Lab” signaled a desire to position it as something new in higher education. Unlike the disciplinary implications in the name “Architecture Machine Group,” Negroponte claimed “media” belong to no discipline (though Canadian Communication Studies, which was well established by that point, would take exception). Certainly the lab’s “demo or die” mantra signaled its departure from traditional modes of academic research.45 Lab discourse is hard at work here, trying to establish itself as exceptional.

While the Media Lab’s claim to novelty might hold up if we compared it to the vast majority of its contemporary labs in higher education, its infrastructural funding model was nothing new at MIT. The university has long embraced funding from ARPA, DARPA, and large corporations—think of the Radiation Laboratory from the 1940s and the Lincoln Laboratory from the 1950s, which still exists today.46 The MIT Media Lab was a continuation of the longer legacy of art and technology labs at MIT (like the Center for Advanced Visual Studies) that stood as central parts of the Cold War incorporation of artists into the university sector.47 The naming of the lab also signaled that it wanted to position itself as part of the lineage of private-sector technology labs (like Menlo Park, Bell Labs and Xerox PARC).

The Janus-faced aspect of lab discourse is at work here too. At the same time as the MIT Media Lab was trying to align it with the private sector, it was also regularly and consistently making claims to avant-garde radicality. In Negroponte’s bestseller Being Digital from 1995, he recounts the following:

The original concept for the Media Lab was to take both human interface and artificial intelligence research in new directions. The new wrinkle was to shape them by the content of information systems, the demands of consumer applications, the nature of artistic thought. The idea was marketed to the broadcasting, publishing, and computer industries as the convergence of the sensory richness of video, the information depth of publishing, and the intrinsic interactivity of computers . . . As in 1863, when the Paris art establishment declined to let the Impressionists into its official show, the founding faculty members of the Media Lab became a Salon des Refusés and had one of their own, in some cases too radical for their academic department, in some cases too extraneous to their department, and in one case with no department at all . . . We came together in the early 1980s as a counterculture to the establishment of computer science . . . 48

Despite the historical continuities from the Menlo Park laboratory to the MIT ML, much of the lab’s discursive identity hangs on its self-proclaimed assertions about its countercultural status and its belief in a philosophy not just of interdisciplinarity but of antidisciplinarity.49 For example, following on books by all previous Media Lab Directors from Negroponte to Walter Bender and Frank Moss which uniformly tout the lab’s unconventional methods and wild successes, Joi Ito’s 2016 Whiplash: How to Survive Our Faster Future begins with a curious disavowal of any influence by Edison. Ito criticizes Edison for his lack of foresight in merely inventing the phonograph without thinking to invent the recording industry, as did Eldridge Reeves Johnson. He eventually moves on to generic assertions about how “the lab . . . [has] always been something of an island for misfit toys” and even more generic, ahistorical assertions about how, as a result of its supposed countercultural status, “the culture of creative disobedience that draws innovators to Silicon Valley and the Media Lab is deeply threatening to hierarchical managers and many traditional organizations.”50

Given the lab’s consistent discourse about its exceptionality over the last thirty-five years, it should not be as astonishing as it is that when we spoke with the current Associate Director of the lab, Hiroshi Ishi, in 2016, he asserted the following:

I have no idea what you’re talking about with the general term “media” . . . Also some people use the same name . . . it is copyright infringement probably. I have no argument but . . . you’re the first person to use “media” as a general term. So it’s confusing . . . If your name is Lori, it’s very important. It’s attached to you. And the Media Lab is attached to here. So, if you can clarify, you’re talking about a general laboratory, about media research, versus MIT Media Lab . . . I encourage you to carefully use the term “media lab” because otherwise it confuses not only me but also the audience. It’s very serious.51

Under US and Canadian law, of course, a name cannot be copyrighted (though it can be trademarked), so this is pure argument by assertion (“I have no argument but . . .”): an exercise of power in an attempt to control larger and more ambiguous cultural meanings via a well-established version of lab discourse with hefty institutional backing. The MIT ML was successful in its use of this strategy for many years, though as we will see, there are definite cracks in the monolith.

We noted in the first section of this chapter the number of “labs” that have no physical space and that may simply be a group of people with common interests (whether intellectual or profit-oriented). We also noted the number of historical entities such as kitchens that were only referred to as labs if their denizens were male scientists. The space of the MIT ML bears a strong resemblance to those historical entities that explicitly referred to themselves as labs in order to give it a certain coherence. Like Menlo Park, the MIT ML contains many open, closed, and porous spaces for hands-on experimentation; a wood shop; a metal shop; areas for spare parts; a kitchen; an area for recreation and socializing; and administration offices. Unlike Menlo Park, it also has a lecture hall as well as designated areas for exhibitions, receptions, and conferences. All the same, given its fairly obvious ties to historical precedents in both name and spatial design, it is odd that its discourse desires complete ownership of the term “media lab” despite the existence all over the world of a myriad of other, unaffiliated media labs.

Regardless of Ishi’s reaction (above), the MIT Media Lab has in fact been remarkably successful in becoming practically and unquestionably synonymous with a lab dedicated to, as Brand put it, “inventing the future,” partly because of a dogged thirty-year marketing campaign whose success we can measure by the fact that almost any discussion of “the future” of technology is a discussion about some project at the Media Lab.

The specific way in which MIT ML discourse is double—claiming to be both part of a long lineage of labs while at the same time positioning itself as utterly new and singular—is evident in its documentation, design, and use of its space. As we have noted, the MIT ML was initially housed in a four-storey, cube-like modernist building (E15) designed by I. M. Pei and Partners. In an attempt to point to its longstanding dedication to interdisciplinary or “antidisciplinary” collaboration between artists, architects, and engineers, the exterior of the building is reminiscent of both graph paper and Piet Mondrian’s grid paintings, as it is covered with square, white metal panels with primary colored lines wrapping around the length of the tiles. But as the authors of Artists and Architects Collaborate: Designing the Weisner Building attest, because of the cube-like structure’s “strong and definite character,” pronounced even more by the flat, metal panels, the building has a way of immediately communicating self-containment and inaccessibility. The authors continue: “Very little that happens inside the building can be seen from outside, nor for the most part are the activities inside the building visible to one another. The building gives the sense of a set of isolated boxes packed one within another. . . .”52 It is intentionally both isolated (or at least giving the appearance of being isolated) and open to the public, albeit only open to that very specific demographic of the public involved in venture capital.

The MIT ML expanded to a second building in 2010. The majority of the lab’s current discourse seems to disavow E15 (referring only to the six floors/163,000 square feet of space in E14 rather than the four floors/114,000 square feet in E15) in favor of celebrating the supposed transparency offered by the new building’s glass and steel exterior. As the campus news outlet put it shortly after the building opened, “The aluminum and glass curtain walls that surround the steel-framed building extend the feeling of openness and transparency to the exterior and make the building appear like a luminous jewel at night.”53 While the building might evoke the feeling of transparency, opacity masquerades as transparency in that all of the windows are covered with a tight metal mesh on the outside and, during the daytime, blinds on the inside. (We elaborate on this paradox of opacity masquerading as transparency in chapter 5.)

From an outsider’s perspective, the seams between the old and the new building materially embody the lab’s discursive shift from appearing more closed and secretive to wanting to appear more open.

Despite the physical (dis)continuity between the two buildings, the message embedded in the distinctly different style of the newer building is a disavowal of the earlier one. The tension between openness and inaccessibility continues as one enters the atrium of the older E15 and is instantly brought into a lifeless, eerily quiet, empty and forbidding space that’s punctuated only by skylights above, one small balcony for each floor, and a few interior windows. Nearly ten years after writing his breathless account of the lab in Inventing the Future, even Stewart Brand had to admit that the “Media Lab’s atrium cuts people off from each other. There are three widely separated entrances . . . three elevators, few stairs, and from nowhere can you see other humans in the five story high space. Where people might be visible, they are carefully obscured by internal windows of smoked glass.”54

Originally where you would have found the majority of the lab’s denizens was in a structure simply called “the cube.”

While the exterior and the atrium of E15 is notably unwelcoming and devoid of any evidence of human activity, the Cube is an open, two-floor structure embedded deep in the heart of the lab. Thanks to the office cubicles located peripherally around a large, central open space that could be and was used as a place to “work, live and eat,” openness and accessibility is turned inward (Tereshko). Certainly this spatial design is reminiscent of, though not identical to, its predecessors insofar as it’s effectively closed to the outside world and open as well as flexible largely to and for its inside world. The design of “the cube” also seems to be a direct answer in the affirmative to a Christopher Alexander poses in the pattern called “Flexible Office Space”: “Is it possible to create a kind of space which is specifically tuned to the needs of people working, and yet capable of an infinite number of various arrangements and combinations within it?” (690) In other ways, the cube indicates the answer to Alexander’s question is no. There are limits on spatial possibility; infinite flexibility is a postmodern fantasy that is driven by the desires of management (for a rhizomatic “flat” neoliberal workplace) rather than the needs of individuals in a building.

Greg Tucker, Director of Facilities at the lab from 1985 to 2014, described the Cube as follows in our interview with him:

Architecturally, it’s a raised floor that doesn’t fill the space, so you get this two-story thing. That idea is just about wrapping as many people as possible around a social space. . . . in the early days of computing . . . the way mainframes worked was you had a room down the hall that was air conditioned to the moon and back. And there was a giant machine in there. And there was a room next to it where all the terminals were, those old VT100s and stuff. And in order to do any work, everyone had to be in that room, not in their office. The offices became somewhat of an afterthought, a place to throw your coat. And because everyone was in the same room, lots of interactions went on. And Negroponte was just going, “This is cool. This is the way this is supposed to work.” So that’s the model. All these lab spaces that you see, and it’s more obvious in the other building, it’s a bunch of offices wrapped around an open space that everyone’s supposed to share.55

The Cube was designed to insist on (more than ensure) interactions and collaborations around specific technologies.

Looking ahead to our third chapter, it’s worth noting the roles that other parts of the extended lab model—especially lab apparatus—play in the production of lab space. The people in the lab adapt to the dictates of the lab machinery, despite the discomfort it can produce, and this practice becomes formalized as policy by a Director. Further spaces develop according to this thoroughly posthuman logic, where machines have as much or more agency than the people working with them, and the lab produces different kinds of subject sand different kinds of knowledge accordingly. Even when the need for a terminal room went away in the era of the personal computer, Tucker maintains that Negroponte insisted on remaining dedicated to the model of an open workplace:

. . . it was all about this very simple thing: use space to force people to consider this as a collaborative exercise . . . when he built the new building, and he put glass on the walls in every office, people on the design side were like “Well, what about their privacy?” Negroponte’s response was classic. He was like, “I’m not offering these people privacy. I’m offering them an opportunity in a collaborative enterprise that was built according to our rules.”56

Once more, there is a tight fit between the design of E15, Negroponte’s views on spatial design, and Alexander’s Pattern Language. The latter asserts that “the totally private office has a devasting effect on the flow of human relationships within a work group, and entrenches the ugly quality of office hierarchies”; thus, one should always “avoid closed off, separate, or private offices. Make every workroom, whether it is for a group of two or three people or for one person, half-open to the other workgroups and the world immediately beyond it.”57

Now that the lab has expanded to E14, the Cube is almost completely empty of any equipment or people, yet the practical difficulties in what Alexander calls “half-open” spaces come to the fore in different parts of E15. On the opposite side of E15’s third floor, the tension between the lab’s much-touted openness—partly expressed via decades of assertions about the wildly creative and disobedient collaborations taking place there—and its closedness is striking in the ways that the lab’s denizens have attempted to undermine the original building’s spatial design and infrastructure. For example, Ethan Zuckerman, the Director of the Center for Civic Media housed in the MIT ML, has an office with a door that cannot close, as his predecessor removed the door closer because of their belief in “Latour’s argument about the agency of the door closer as the embodiment of certain systems about what is supposed to be open and closed.”58 After pointing out the idiosyncrasies of how the physical space of the lab ends up embodying research projects and after touring the seemingly open and glassed workspace outside of his office, Zuckerman asked us back our own question: “Is this an open or closed work space? It’s mandated to be an open work spaced . . . [by] the powers that be because we are supposed to look busy and active and energetic. But when students are trying to get work done, they cocoon.’59 The theme of undermined openness also continues throughout the space via wall-less workspaces dominated by large meeting tables and surrounded by panels that have been hung up to create a very small measure of privacy or (ultimately ineffective) boundedness.

The design process for an additional building that would house the MIT ML in E14 began in 1998, with architect Fumihiko Maki; the building eventually opened in 2010. While the exterior of the new building might look as if it is disavowing the old building, Maki designed the interior such that it would replicate the structure of the cube in E15. More specifically, directly above a large public atrium that opens onto the campus street, seven two-story glass cubes are arranged around an upper atrium, with each cube intended to be home to a particular lab or research group. Each is vertically staggered so that moving from one lab to another involves returning to the central atrium and using the stairs or elevators. Since all of the walls facing the atrium are made of glass, the staggered verticality appears to be a creative way to make the individual workgroup spaces only half-open. Regardless of the stacking, the space of E14 is unnervingly exposed, with the balance between open and closed weighing heavily on the side of open as one can look into and onto nearly any space one wishes.

The space of the lab’s central, upper atrium also adheres to Alexander’s recommendation in the pattern “Common Areas at the Heart” to “[c]reate a single common area for every social group. Locate it at the center of gravity of all the spaces the group occupies.”60 However, despite the design of the central meeting space, with its apparatus of couches, armchairs, and ping-pong table, it turns out that most users of the space engage solely in individualistic work, perhaps because E14 is already over-ridden with a kind of tyrannical openness. Other attempts at communality throughout the building, such as the “food cam” or the “resuse cam”, also produce individualistic rather than community-oriented interactions, as those with extra food or electronics leave their items on a designated counter, push a button, and a camera above sends a notice to people who are in the lab that there is available food and/or electronics.

Hearkening back to Edison’s Menlo Park laboratory, the most open and communal areas in E14 are the machine shop and wood shop on the lower levels. These spaces are open to any MIT ML student at any time and have nearly any kind of equipment (from shop bots to water jet cutting machines) to allow students, especially those enrolled in the “Make Anything” class, to learn how to do just that.