Cybernetic Revolutionaries: Technology and Politics in Allende’s Chile
by Eden Medina
MIT Press, Cambridge, Mass.
Chile’s Cybersyn project has long needed a thorough, critical history, and now we have one. Eden Medina’s carefully researched book lays out the project’s origins, structure and struggles as a “case study for better understanding the multifaceted relationship of technology and politics.” If there’s a lesson in Medina’s account, it’s that the task of designing a politics into a technology is not only necessarily social and unwieldy, but best understood at the micro-level. By doing good history here, Medina has also provided something of a design brief, a guide to just how broadly and deeply designers need look when seeking to embed their work within a particular political vision.
Designed by a collaborative, interdisciplinary and international team, Cybersyn (or Synco as it was known in Chile) was many things to many people. The project originated in a conversation between Fernando Flores, technical and economic advisor to then newly-elected Allende, and Stafford Beer, the renowned management consultant and cybernetician. Faced with a newly nationalized system of production and distribution, Chile’s Popular Unity government needed a way to facilitate shared, open and efficient conversations about labor and goods at an unprecedented scale. Materials were in short supply; workers were still new to the Socialist system and sometimes wary; and the United States was encouraging disinvestment as a way of suppressing the perceived threat of a socialist state in South America.
Beer threw himself into the project with increasing fervor in what became, by most accounts, a life-transforming experience. Together with Flores and a small, young team, he applied his earlier work to designing a new sociotechnical system for Chile that would not simplistically solve complex problems, but rather model possible scenarios for a collective (or at least representative) decision-making body. The Cybersyn team sought to build a system composed of static technical and social components that allowed for flexible relationships between those components. One might reorganize the order in which different parties are notified of changes in production, for example, but the essential composition of the social groups, and the technologies in which they participate, remain the same. Through such an approach, Chile’s citizens might enjoy what Beer called “effective freedom,” in which all were guaranteed as much autonomy as possible for the greater good. Medina points out how closely this approach, which preserves the potential for hierarchy in the service of the people, mirrored that of Allende’s understanding of his own executive power.
Cybersyn’s components included a new communication network (“Cybernet”), an operations room, and two new software packages: one for statistical analysis (“Cyberstride”) and another for modeling economic scenarios (“CHECO”). The software ran using spare time on the government’s four existing mainframe computers; the network utilized 400 surplus Telex machines; and the control room eventually found a home in an unused atrium once used as an automotive showroom. The new network facilitated real-time exchange of data up and down supply chains, updating managers at various levels about production quantities, supply shortages, labor speeds. The software package, designed with the sub-contracted help of Ron Anderton, offered a custom-built modeling and simulation environment for computing possible outcomes of management decisions made in response to new data. The iconic, ultra-modern Ops Room, designed under the leadership of Gui Bonsiepe, was where such deliberations would take place.
By confining automation to the processes of information exchange and decision modeling, the designers of Cybersyn hoped to harness the powers of new technologies while preserving an essentially socialist, participatory space of decision-making and control. Thus we see a manifestation of Beer’s hope to balance centralization and decentralization in management systems, applying networked technologies to the solution of large-scale problems while avoiding technocracy, or the dictation of human behavior by machine.
Medina’s book is largely dedicated to the task of answering whether the realized design truly manifested this political aim. The fact that Cybersyn was never fully implemented before Pinochet’s violent rise to power (and the system’s subsequent dismantling) presents no barrier to the historian’s efforts, and probably affords her ever more opportunity for scrutiny and reflection. Medina looks to at least three different areas to discover the project’s politics: first to how the design itself reflected the socio-political particularities of the team’s efforts, situated as they were in a certain geopolitical, historical context; then to how the project, even half-completed, found ready use by government in a time of crisis; and lastly, how the project, at later stages, found vastly different descriptions even among its proponents, let alone by its critics. In each of these areas Medina makes substantial contributions through deep archival and interview work, as well as careful synthesis and reflection.
We learn, for example, how the designers of the Operations Room went to great lengths to design a space of inclusive debate, offering each of seven managers his own control pad for calling up new information on the room’s shared screens. According to Medina’s interviews, fiberglass suited the room’s designers not as a stylistic reference to futuristic or science fiction scenarios, but rather because its material seemed infinitely re-formable, plastic, and therefore appropriate to a space of experimentation. Yet we also learn of the very gendered way in the room was designed to operate. The designers excluded paper or QWERTY interfaces, for example, so as to eliminate the need for clerical labor (positions typically staffed by women). The only roles available for women in the original vision were as designers of the slides which operators called up by “thumping” their armrest controls. Women here – in this case, four graphic designers – functioned literally behind the scenes. Beer even explicitly referenced “Gentlemen’s Clubs” as an example of how the room should feel. Original designs included a bar.
We also learn how the Cybernet Telex network, perhaps the most realized portion of the project, sprung into action during a general strike to facilitate real-time coordination of non-striking workers, management and government. It seems likely that without this mode of communication, the Allende government might not have outlasted the strike. In Medina’s view, this particular system doesn’t stray far from Taylorist approaches to labor, time and space. She points out Cybernet’s similarity to the magnetic-tape based numerical control systems in place elsewhere at the time. From a temporary control center at State Development Corporation, throughout the month-long strike managers read paper tape from the Telex machines in order to send new tape back to the factories with instructions to speed up, slow down, or alter course.
Medina’s account of Cybersyn’s public debut is of particular note, as we see therein an encapsulation of debates often found around cybernetics; few approaches can so easily be described as liberating by one and subjugating by another. Working as a publicist, Beer had his work cut out for him in early 1973, when the Orwellian nightmare of a computer-controlled society was nearly a cinematic and literary genre. He was profoundly frustrated with how critics misunderstood Cybersyn’s mission, and increasingly blamed not only Chile’s incomplete deployment of the system, but the United States’ efforts to stall Chile’s efforts through economic pressure. Building on Gerovitch’s history of Soviet cybernetics, Medina points out how Cybersyn’s critics might also have worried about a Socialist state pushing the balance of cybernetic research in an autocratic direction. (Meanwhile, the Cybersyn team saw the Soviet example as too centralized, bureaucratic.)
Medina concedes that the reconfigurable nature of Cybersyn’s components did leave the system available for deployment by a variety of political interests. Instead of simply defending Beer or others against attacks, she takes the time to study who makes what accusations, and why. The panoply of attacks from journalists and academics are reminiscent of contemporary debates over the right of developing nations to nuclear armament, in that we see Chile portrayed as both not advanced enough to deserve such a project and not authorized to do so. If Cybersyn’s critics misunderstood the project – mistaking the Operations Room for a control room, for example, or interpreting the CHECO simulator as a governor – they did so often from States where other forms of centralization and technocracy were in place, or well on their way to realization. Even Cybersyn’s most vehement critics – those who took missiles to Allende’s palace and knives to the slides in the Ops Room – later appealed to Chicago school economists, disinvesting in federal infrastructure and deregulating commerce while enforcing a new regime of centralized accounting, surveillance, and murder through disappearance.
Sovereignty and autonomy then are available as values to defend by all parties in the rush to the assumption of power. In this case, Medina’s account of how a sociotechnical system, operating at large scale, still affords capitalization by individuals or rival states might lead some to cynicism about the potential for designing a politics into a technology. But this would be a mistake, a missed opportunity to learn more about the lively life of software, hardware, interfaces and systems as they leave the drawing board.
Cybersyn’s story may leave one large nagging question for some – if the system had functioned as designed, would it have achieved the desired political end? The largest gaps to be closed included some factors as technical as computer processing power, and other matters more political – such as inclusion of worker input in the design of the CHECO modeling software. Here’s how the system should have worked: in the Ops Room, a representative group should have been able to receive fast, accurate data from the spaces of production via Cybernet, from which it could construct fast, representative simulations of possible future scenarios using CHECO, resulting in recommendations for action delivered back to the spaces of production. At the contact points of each stage of this process lay complex information transfer, and one could imagine the manual process of displaying such information growing increasingly automated as a sort of “dashboard” for each group in the chain of decision.
Three points in this picture bear most of the weight in achieving success – that of the visualizations as trusted representations of data; that of the simulation as a trustworthy model of the problem at hand; and that of the Ops Room managers as a representative and skilled team. Through Medina’s lens, we might easily see where technology comes in and out of attention at each of these stages. Does one trust the program or the programmer? Were the managers selected according to a sensible system? If a worker has never been to the Ops Room, can she trust the managers’ word via Telex (or telegram, or smartphone)? As one closes the technical gaps of Cybersyn, the effectiveness of the system relies just as much on human expectations of technology and the relationship of users to their tools. Medina’s detailed accounts of how differently each player approached the same component here are particularly useful.
Cybersyn might seem more unique in its relegation of computational modeling to the role of advisory image, and Beer went to great efforts to demonstrate this point in the system’s defense. This feature does indeed distinguish Cybersyn’s approach from that of Norbert Weiner, for example, where the cybernetic algorithm often found a role as governor. For Beer, the CHECO simulator was only a true “homeostat” when in use by the workers – it could not function as well on its own. Yet even here, might not a smoothly-functioning system eventually begin to fall into specific rhythms of use in which an algorithm’s reliability renders it less visible, and therefore less debated? Indeed, might not this describe the sort of modeling common to contemporary financial markets or even intelligence analysis?
A key issue here seems to be how one ever understands the model – or the visualization, or the representative body – in the first place. Is CHECO one of Sherry Turkles’s “objects-to-think-with,” as Medina describes the role of electronic circuitry for Beer? Is it a self-contained entity that exists apart from use, one of Pickering’s “ontological machines” that functions as an agent and not only as a mirror? Is it a rhetorical device? Beer seemed to have a clear sense that a problem can never be fully solved through simulation, but rather only abstracted via metonym or metaphor. He placed the weight of interpretation on the users, but perhaps underestimated the force of other powers on that interpretation, and failed to account for how one approaches a model from deep within a particular frame of gender, geopolitics, and sensation.
In the Ops Room there was no paper with which to record one’s thoughts on the varying inputs at hand, the various models of outcome offered. Debate over suggested action seemed to rest in the realm of speech, preferably augmented by food and drink. In this context, perhaps the most useful object in the Ops Room was the metal board on which lay a collection of magnetized geometric shapes. Here, one might have not only diagrammed out possible outcomes, but debated the temporarily assigned significance of each shape and color in a particular argument. Devoid of text, each symbol system might inevitably have fallen away to accrue new meaning in a rambling conversation scaled by human memory. Through good design, a model’s definitions and confines might have stayed present even as it remained useful in discourse.
It’s striking that the Cybersyn system seemed to have no place for stored memory, even of the active, ephemeral kind explored by the sympathetic Austrian-American cybernetician, Heinz Von Foerster. How might the system store information about past successes and failures in order to adjust for different future outcomes? In other cybernetic systems, the algorithm itself learns, retains traces of the past. If Cybersyn’s designers had hopes for such memory, there is no mention of it in Medina’s book. Instead, the task of remembering seems to have fallen to the workers, in their roles as speakers, debaters, and physical laborers. As in the aesthetics of early Soviet Constructivism, where cinema and radio served as network, the fluidity of form at the expense of memory seems to be the order of the day for Cybersyn. In such a vision, the people might reinvent themselves each day, with new ways of using the same blank slate.
Yet this utopian vision of memory is haunted by subsequent events. In her final chapter, Medina relates the account of how, during Allende’s violent ousting by the military, two members of the Cybersyn team scrambled to collect paper documentation of the system – artifacts forbidden in the Ops Room – in order to preserve some memory of the project. This final act of storage anticipated a period of a great offensive against memory in Chile under Pinochet, and made possible a contemporary account as detailed as Medina’s. In light of these events, the neglect of storage capacity in Cybersyn’s design seems to be the project’s weakest link, and presents a prime opportunity for re-invention.
Kevin Hamilton (complexfields.org) is an artist, researcher and scholar with the New Media Program at the University of Illinois, Urbana-Champaign, where he also co-directs the Center for People and Infrastructures. He is currently at work on a critical archive of American nuclear test and training films, and recently completed a commissioned public artwork about Illinois’ historic center for cybernetics research, the Biological Computer Lab. Awards include grants from the National Science Foundation and National Endowment for the Humanities, as well as fellowships at University of Southern California, Cornell University, and the Banff New Media Institute. At Illinois he also holds affiliations in the Department of Media and Cinema Studies, the Center for Advanced Study, and the Center for Arms Control, Disarmament and International Security. Born and raised in South Carolina, Kevin later completed his studies in Rhode Island (RISD) and Massachusetts (MIT); he has lived in Urbana, IL since 2002.