Interview by Ilana Gershon
Questions for Ben Peters
If you found yourself at a picnic with linguistic anthropologists, and one of them was sensible enough to bring tasty corn and seemed very interested when you mentioned briefly that you had written about the history of thwarted Soviet attempts to develop the internet, how would you explain your book?
Hey ling anth friends, please pass the corn!
(Between messy bites) I think my interest in the Soviet internet story owes a sideways debt to the fact I’m from deep corn country, USA. Coming from a small college town in the Midwest licenses one not only to know their corn (delicious!) but to appreciate life removed from the urban centers of global action. This appreciation once struck me as a 20-year-old service volunteer living in a small city called Balakovo along the Volga river in the post-industrial rust belt of Russia. There I realized not all “middle of nowheres” are similar (Ian Frazier’s delightful and meditative masterpieces The Great Plains and Travels in Siberia make this point). In addition to the earnest people, the agricultural base, and scenic landscape I was used to back in Iowa, Balakovo, a former secret Soviet city, had decaying military-industrial factories for the cosmonaut industries, an enormous hydroelectric dam, a line of nuclear reactors looming on the horizon, and much else. Where did this outsized industrial infrastructure come from? I remember wondering on the edge of an enormous dammed reservoir: who first thought it a good idea to plan so much electrical power in out-of-the-way Balakovo, and why?
(Holding up half-eaten corn cob) I have since recognized that the big ag corn industry in the US—and perhaps even all advanced modernity in its debt to large institutions—participates in the basic question behind the Soviet internet book: why do some institutions organize our lives and others not? Why, in particular, wasn’t there a Soviet internet? Given its failure was neither inevitable nor natural, what was the story of the Soviet scientists and leaders that planned to network the nation with computers anyway? Who were these cyberneticists? What did they want when they started to imagine networking the planned economy with computers in an ambitious project called the All-State Automated System ( or OGAS for short)? And why, despite thirty years of attempts at the height of the cold war tech race, did this outsized informational network infrastructure for the Soviet people not take root?
(Hint: my answers are not censorship cultures, technological backwardness, or inefficient hierarchical states.)
Throughout the book, you explain that while many cyberneticists modeled their national network projects after the human mind, the projects were quite different, in part because how they understood this metaphor varied so much. Could you explain some of the differences?
(Sets corn cob down in order to focus.) Sure thing. There are many brain-computer metaphors—and none of them are right. I think the twentieth-century pedestaled the wrong image of the ideal computing processor: the ideal computing processor is not the human brain. (Moreover this towering intellectual hubris—or what brains think about themselves—builds naturally on a troubling early modern vaulting of western individualism.) Cybernetics, after World War II, enabled strong neural-computer network analogies to be at work in Warren McCulloch’s influence on Paul Baran’s distributed network design at RAND, in Stafford Beer’s influence on Allende’s Cybersyn network in Chile, and in (as I detail in the book) Viktor Glushkov’s influence on the OGAS Project in the Soviet Union. Each had different consequences in different places: to put it in a nutshell, the ARPANET designers imagined their nation as a single distributed brain of users, while the OGAS Project designers (not entirely unlike Beer in Chile) imagined their network as a nervous system for layering onto their nation as an economic industrial body of workers, with the state as the brain. To suggest that the first American computer network was modeled after an imagination of a national brain and the Soviet networks were after a national body not only rehearses the mid-century emerging information and industrial cold war economic differences—but it also obscures the on-the-ground story of both. I suspect linguist anthropologists have much to teach me here in particular: I think the biggest difference lies not in the metaphors but in the distance between all those brain-computer metaphors and the embodied practices of building and institutionalizing computer networks. No national network projects resemble the human mind in practice.
Economists and economic cybernetics play a much larger role in the book that I imagined when I first started reading. It left me wondering: what kind of dilemmas do economists have to solve when they don’t presume that the market is the best way to distribute goods and determine value?
They stump up against some of the hardest dilemmas I know. In all semiotic-material discourse (of which the economy is of course just one mode), every evaluation is also an executable fabrication that itself acts on other evaluations. And among the resulting chains of operations, in which there are many dilemmas, the ones that matter most in this book do not fall along the cold war economic liberal language of private markets versus public states. (Hannah Arendt, for example, nudges my conclusion to deconstruct and begin rebuilding network discourse beyond the tired cold war triumphalism around markets, liberty, and commerce.) As you note, I spend a couple chapters developing how economic relations did not work as planned in the Soviet context: continuous and partial reform among battling schools of thought, nonlinear command and control dynamics, informal power networks, vertical bargaining, and other sources of organizational dissonance. I dub all this, borrowing from McCulloch and David Stark’s language, “heterarchy.” In a heterarchy, every node is subject to competing regimes of evaluation and the resulting logics by which value is determined cannot be described or mapped onto simple two-dimensional models (markets, hierarchies, and so on). Perhaps our behavior can be mapped onto a higher order in n-dimensional spaces, suggests McCulloch, or perhaps not at all. How we determine value is a complex measure of how modern humans interact, and indeed how any actor responds to contradictory demands (do I write, prepare for class, go for a walk, or have another piece of corn?) reveals more than our negotiated compromises to that contradiction.
Back to the Soviet case: it is no surprising revelation that the on-the-ground practical relations for determining and planning values in the Soviet planned economy did not function as they promised to on paper. Still, this mundane fact had consequences in at least two directions: it frustrated the rationalizing impulses of technocratic economic reform attempts such as the OGAS Project. It also ensured that economic bureaucracies could actively resist reforms because they were free to pursue their institutional self-interest in the status quo. The Soviet network story is thus an uneasy mix of technological genius and futuristic foresight leavened with mutinous ministries and institutional infighting. Or, to restate the book hook, while the US ARPANET took shape thanks to state funding and collaborative research environments, the Soviet contemporary projects broke against the rock of unregulated competition among self-interested institutions and bureaucrats. The first global computer networks took shape thanks to cooperative capitalists, not competitive socialists.
As your story unfolds about why the Soviets never developed a national computer network, another deep irony emerges – that a system built around centralization consistently over time ended up undercutting any possibility of a centralized computer network. You suggest that this is a story about the tensions between belief and practice, between a system that was touted as a centralized and well-regulated bureaucracy and in practice a complicated mixture of differently structured hierarchies. Could you discuss the tensions between Soviet centralization and Soviet bureaucratic fiefdoms that lie behind your history of non-events?
That’s a beautiful question central to the project—one that I’d like to tweak in two ways as a way of responding.
First, commonplace understanding of the Soviet state as centralized and well-regulated is empirically wrong. Let’s think instead of the Soviet state as trying to turn a complex field of decentralized fiefdoms into a single field of decentralized fiefdoms. Even the Politburo rarely endorsed totalizing centralization, and certainly none of the Soviet network projects in the book call for fully centralized networks. With the exception of one short-lived radial network proposal, all of the proposed Soviet network projects between 1959 and 1989 interestingly resemble decentralized pyramids (just like the official economic plan at the time).
These network still recognized a central command in Moscow while also permitting real-time remote access between any two authorized nodes on the national network. This is a key corrective (especially in light of the romance of flat organizational networks in the west): in both principle and practice, the Soviet Union was not too top-down or rigidly hierarchical. And in administrative practice, it was too messy and pernicious.
Second, the book is a negative history of real events, not a hypothetical history of non-events (although it does color the vision of a futuristic electronic socialism that never was). The reason I think this re-characterization matters is because I am openly interested in helping normalize the study of failed projects among scholars of human relations, complex institutions, media studies, and their adjunct interests.
Or as the book puts it,
contingent histories also help focus public debate better than do popular histories of technology that parade about hackers, geniuses, and geeks marching to the Whiggish beats of technological progress. In negative histories failures, even epic breakdowns, are normal. Astonishing genius, imaginative foresight, and peerless technical wizardry are not enough to change the world. This is one of the lessons of the OGAS experience. Its story places the conventional concepts of technological successes and failures on the wobbly foundations of the accidents of history. The historical record is a cemetery overgrown in short-lived technological futures: stepping of its beaten paths leads us to slow down and take stock before we rush to crown the next generation of technologists as agents of change. (197)
Perhaps the hard moral of the story is this: no one sensitive to the suffering all around us cannot want to reform the world for the better. Yet, in the multivariable calculus of social reform, the only thing more certain than our desire to change the world (and media and language are among those ways) is to admit that there is no guarantee that any given effort ever will.
This is a beautifully written book about bureaucracy, one in which you even manage to make a bureaucratic meeting, and the fact that two people happened to be absent, suspenseful. I was wondering if you could talk a bit about your writing strategies for making compelling some topics that at first glance might seem like good insomniac aids. How do you make bureaucracy and its genres compelling?
(Setting down now finished corn cob.) Thanks! That means a lot. Writing is one of my favorite demons. Bureaucracy, because it (like computing programming) is made out of writing that is meant to be executed but never really read, often deadens its observers to the real fertility and force of the written word. Perhaps students and scholars of mind-numbingly dull technical systems should indulge in great stylists in English and other languages as temporary antidotes against the occupational hazard that is prose pollution. Outside of my own awe for language (which I take up more directly in my brand new edited volume Digital Keywords: A Vocabulary of Information Society and Culture, which Princeton published last week on the 40th anniversary of Raymond Williams’ Keywords), I don’t have any sure-fire strategies, although the normal ones will probably do: I recommend reading voraciously and strategically, slavishly imitating and scrupulously doubting the masters, writing for the smart and interested eighth grader, and then rewriting with an ear tuned to the cadence of language. Of course I rarely manage to pull all that off, but it’s a good thing to try and a better thing to have sympathetic readers to share it with.
Thanks for the picnic!
Benjamin Peters is the author of How Not to Network a Nation: The Uneasy History of the Soviet Internet (MIT 2016) and editor of Digital Keywords: A Vocabulary of Information Society and Culture (Princeton 2016). He is assistant professor of Communication at the University of Tulsa and an affiliated faculty at the Information Society Project at Yale Law School. Work site: petersbenjamin.wordpress.com. Tweet at him @bjpeters.