I recently finished readin Eden Medina’s “Cybernetic Revolutionaries: Technology and Politics in Allende’s Chile”, a history of Projecto Synco (Project Cybersyn), often billed as “the socialist internet”, though reading Cybernetic Revolutionaries I think “the socialist AWS” might be more accurate. Put roughly, Synco was a technological and managerial infrastructure built under Chile’s sociliast president Salvador Allende, starting in 1971 and ending in September 1973 during Pincohet’s military coup. The project was deeply informed by the framework of cybernetics, and many of its key architects believed that these cybernetic principles would lead to a form of national economic coordination which could transcend (sublate?) the “centralized vs decentralized control” paradigm that had previously dominated the global discourse around socialism.
I described “Revolutionaries” as a history, but it might be better to call it a case study. The book closely follows both the sociopolitical forces which influenced the design of Synco, but also how the developing product “functioned” socially; which of its design goals it achieved and what the historical impact of the project ended up being. Reading this analysis it is made clear that the concept of “proper function” is not well-defined when you take a step back from the material technology. For me, this was a serious readjustment of my thinking around technology. As a toy example of this ambiguity, Medina references a rock, which might alternately function as a weapon or construction tool depending on context. The operation of this “technology” depends entirely on the social configuration that surrounds it, and so it’s maybe more appropriate to define “technology” as the human procedures surrounding some object (or collection of objects), rather than the object itself. At the very least the thing we call “the technology” is the combination of the two, rather than either one alone.
It’s really cool to me to think about statistics, and mathematics more broadly from this viewpoint. Operating with this dual definition of technology, math seems like a special case, as the mathematician is simultaneously the object around which the technology operates (what Dr. Legasov would have maybe called a biorobot) and also the human person performing operations on that object. Statistics maybe has a more concrete object at its center in that it’s defined around “data” and its associated calculations, or at least a more concrete set of protocols, but I think it’s still very much a human technology. Even if you’re using R to record data and perform calculations, ultimately the human operator is the “thing” which is performing the statistical analysis; in the language of cybernetics, the computer is just a subsystem. Thinking about statistics as a technology leads to the obvious question “what does it mean for statistics to function properly?”
There’s a lot of ink spilled trying to answer that question, but I think it’s typically spilled trying to answer a very specific interpretation of it. For theoretical statistics “function properly” means to satisfy some pre-specified roster of epistemoligical desiderata. But we could also consider “function properly” as some kind of memetic success; how do different statistical paradigms emerge and die? Under Sewell causal statistics fell flat, but after Judea Pearl the methodology is experiencing a revival; why?
The last chapter of “Cybernetic Revolutionaries” takes up a question kind of like this, describing Synco’s public reception and highlighting some interesting interactions between its architects and both the Chilean and global public. Indeed gaining public approval was a major challenge for Synco. As I mentioned above, its designers had a very particular view of how Synco would be operated to enable workplace democracy, but the public was frequently skeptical or even ignorant of this claim. Spoiler alert: what ends up happening is kind of a failed paradigm shift, with the cybernetic underpinnings of Synco failing to take root in Chile and abroad. Synco faced other challenges, and its designers certainly were not perfect, even when it came to meeting their own ideals, but you have to wonder what would have happened if their ideas had gained more traction, or if it was even possible for them to do so.
Despite a downer ending I found the book overall to be really inspiring, and its lessons seem highly relevant in the era of “data for social good”. The prose was also very engaging, and Medina does a wonderful job bringing the historical characters to life (although several of them are still in fact alive). If you have any interest in the history of statistics, the theory of design, or the technological aspects of socialism I would highly recommend checking adding “Revolutionaries” to your reading list if it isn’t there already.