"The real story about Soviet computing networks has far less to do with the technology itself than with the institutional, political, economic, and social networks that made up the knowledge base and innovation infrastructure in a country and a culture."
Peters begins with what is titled "a global history of cybernetics" but which, though it touches on Norbert Wiener, the Macy Conferences and cybernetics' failure as a discipline in the West, largely concentrates on the Soviet Union. Here public criticism by Stalin was followed by rehabilitation, beginning in 1952 with Kitov's discovery of the one copy of Wiener's Cybernetics relegated to secure access and with key publications by Kitov, Lyapunov and Sobolev in 1955. There was considerable reworking for the Soviet audience and context, notably in "efforts to silence the social implications of the theory":
"the technical and abstract mathematical language of Wiener's cybernetics thus served as a political defense against Soviet philosopher-critics and as ballast for generalizing the coauthors' ambitions for scientists in other fields".
Institutional success brought a "slouch toward the intellectual mainstream": "the failure of cybernetics to cohere intellectually actually rested on the runaway growth of the discipline institutionally", even as it became "a nearly empty signifier".
A specifically Soviet discipline of economic cybernetics emerged, which became enmeshed in debates between traditionalists and liberals. It was all very well to argue for an experimental approach — model and calculate and see what happens — but:
"It was self-evident to the economic bureaucracy that computers were not value-neutral: cyberneticists ran them, and no state resolution could convince the bureaucrats to behave like rational bureaucrats in ceding power to cyberneticists."
Peters appeals to the concept of heterarchy, between market and hierarchy and "ordered complexly in ways that cannot be described linearly", to describe the workings of the Soviet economic administration, but for me this remained too vague to be at all useful.
The three earliest civilian networking proposals appeared in the context of the Cold War. Anatoly Kitov proposed the reuse of military infrastructure to serve civilian (industrial) purposes, through Economic Automatic Management Systems. A first letter to the Central Committee in 1959 was successful; a second was not, and saw his fall from grace (for bypassing the military hierarchy) though he survived a show trial and became a "civilian network entrepreneur" (an odd-sounding status Peters fails to elaborate on). In 1962 Aleksandr Kharkevich proposed an ambitious unified communications system (ESS), but the proposal died with him in 1965; the final proposal was a poorly documented and short-lived suggestion by N.I. Kovalev.
"Perhaps the signal lesson to take from these early Soviet network proposals is that there is no inherent connection between the designs of technological and political systems."
The most ambitious Soviet networking project was the All-State Automated System or OGAS: its origins go back to a letter to Khruschev in 1962 and a Politburo resolution in 1963, and its chief visionary was Viktor Glushkov.
"More than a network, the OGAS Project as formulated by Glushkov outlines a daring technocratic economic imagining that was meant to operate in a future Soviet information society by digitizing, supervising, and optimizing the coordination challenges besetting the national command economy."
Among other ambitions Glushkov proposed an electronic financial system with no physical currency, and he had the idea of brain-inspired computers as an alternative to the von Neumann model. But he also had "a tenacious and pragmatic administrative acumen", building coalitions and using "the governing logics of blat and personal politics": key players were Nikolai Fedorenko's Central Economic-Mathematical Institute (CEMI) as well as Glushkov's own Institute of Cybernetics, and the economist-mathematician Vasily Nemchinov.
A turning point came at a key Politburo meeting on 1 October 1970, from which both Brezhnev and Kosygin were absent, at which the OGAS Project was "neither fully rejected nor approved". Soviet administration had, Peters suggests, "too much, not too little flexibility in its capacity to generate organizational dissonance", with a variety of "entrepreneur"-managers pushing in different directions. In the end the CEMI gave up on networking in favour of microlevel linear modelling of enterprises (and other less controversial projects). The OGAS project itself went into "repose" and by 1989 had "slipped into history". (As a parallel, Peters touches here on Botvinnik and the Soviet computer chess programme.)
Peters tries a little too hard to make his story exciting. He highlights coincidences such as Kitov and Glushkov's children marrying and Sergei Lebedev's computing laboratory, the "newest icon of Soviet atheism", being housed in a monastery. He speculates that with successful Soviet networking "the current global network culture could have looked very different" in areas such as online privacy. And he plays up too much "the conceit" of his book, that:
"It was not the absence but the presence of vibrant unregulated markets of conflicting forces driven by self-interested administrators that kept the Soviets from networking their nation and command economy."
His own account suggests to me that the reason for the failure of the OGAS project was less any aspect of Soviet institutional culture than its status as a mega-project coupled to the reform or even reconstruction of the entire Soviet economy. The most striking thing about the ARPAnet, in contrast, is not that it was government funded, but that it was in its early years the plaything of distracted graduate students, given no specific goals at all. A comparison of these pioneers with the young, enthusiastic Soviet mathematicians and cyberneticists of the same period might have been interesting, but the key difference here was surely that the United States had the resources to leave the infrastructure needed for a national network — computers and telecommunications links — just lying around. (There is no discussion at all in How Not to Network a Nation of general Soviet computing and telecommunications, or of whether network planning ever got as far as details such as addressing and routing protocols.)
Peters brings together some fascinating material in How Not to Network a Nation and makes it into a coherent story. I found some of his theoretical framework unconvincing, but his efforts can be appreciated without that and anyone with an interest in Soviet political and institutional history should find plenty in it to chew on.
November 2017
- External links:
-
- buy from Bookshop.org
- buy from Amazon.com or Amazon.co.uk
- share this review on Facebook or Twitter
- Related reviews:
-
- books about Russia + Russian history
- books about networking
- books published by The MIT Press