In a word, decentralization. I had the pleasure of presenting a working paper and my first stab at such an argument at the Harvard-MIT-Yale cyber-scholars working group at the Berkman Center Tuesday, October 21, 2008. The other presentations were stimulating, and all took place leavened by warm food and the smart and generous folk at Berkman. It was a real treat. A paper draft is available for comment at bj[insert my last name here]@gmail.com and a video of the presentation is available online (here).
Building on the fantastic “InterNyet” article (here) of MIT historian of science Slava Gerovitch, I argued, in brief, that the Soviet attempts to build a non-military nationwide computer network (namely Victor Glushkov’s 1964 proposal for a hierarchically-structured information network that could harvest and manage all economic data for the entire Soviet socialist economy) in the 1950s and 1960s need to be understood in the context of decentralized politics, administrative structure, and network design. Decentralized networks are obviously different than centralized networks but what many forget is that they are also importantly different from distributed networks as well (for this distinction, see Paul Baran “On Distributed Communication” 1964). The fact that Soviet state structure was decentralized hierarchically, and that ministries did not share information or funding sources between themselves, offers both an explanation of why no comprehensive computer network design could survive fractured implementation as well as a cautionary tale today for our own largely decentralized world saturated, as it may be at times, with comparable levels of talent, enthusiasm, and vision that too could prove shortsighted. One particular case study, the origins of the Central Economic Mathematical Institute in Moscow and the irony of purposeful funding (both the lack of money and a channeled flood of money can kill a brilliant project), was examined in particular.