On the Cybersyn Article Controversy: We Need Best Practices October 11, 2014Posted by Will Thomas in Uncategorized.
Evgeny Morozov recently wrote an article for the New Yorker about management cybernetician Stafford Beer (1926-2002) and Project Cybersyn, an ambitious early-1970s attempt to use information-handling technologies to manage the Chilean economy under Salvador Allende.
Eden Medina published a fine book on the subject in 2011. Morozov mentioned Medina’s book in his article, but only in an off-hand manner. Historians, writing on Twitter and the SIGCIS message board, were incensed — the article, they believed, was, in effect, intellectual theft. Morozov’s reply was a post on his tumblr describing his original research process, while acknowledging the importance of Medina’s work.
I’ll offer my own take up front: as a courtesy, Morozov should have acknowledged Medina’s book much earlier in the essay, and should have signalled to readers that it is an authoritative source on the subject. I reach this conclusion regardless of whether the essay was a proper book review or not (which has been a subject of some confusion, given that the essay was a “Critic at Large” piece, which appears in the book review section).
My conclusion is drawn from my belief that Medina’s book constitutes a canonical source. I will, however, refrain from joining my historian colleagues in their visceral disgust at Morozov’s essay, because this principle of the canonical source (as opposed to an originating source) does not actually exist. I have just made it up now, and it would not be charitable to expect people to be bound to either a principle or a canonical status that nobody has ever agreed upon or even discussed.
My proposition: if there exists a large (i.e., book-length), thorough, deeply researched treatment of a topic — even if one disagrees with aspects of it or goes beyond it, and especially if it is recent (say, the last 15 years) — one is obligated, regardless of publishing genre or venue, and regardless of whatever replication and supplementation of research one has done, to acknowledge that canonical status clearly.
Now, for an examination of some particular issues in the case at hand…
My colleagues’ stronger objection to Morozov is based on the fact that they believe Morozov more or less derived his essay from Medina’s work. It does seem to be the case that it was inspired by a reading of Medina. Here’s how Morozov puts it in his tumblr post:
In a sense, I was lucky because there’s an excellent – and yes, entertaining – history of Project Cybersyn. It’s Cybernetic Revolutionaries by Eden Medina. It came out in 2011, so it’s not particularly “recent” to fit into the New Yorker’s slot. But we found a way.
I interpret this to mean that, while Medina’s book is fresh to the scholarly community, it was too aged for the New Yorker to review directly. It could, however, feature within the context of a larger “Critic at Large” essay about Project Cybersyn, and its resonances with present-day problems of information-handling.
In his post defending his piece, Morozov makes a great deal of the extensive research he put into the essay, which I find exonerating, mainly because he follows the above quote with the declaration:
[Medina] was not my first entry point into Beer though. I had already read some of [Beer’s] books, essays, and interviews, as I’ve also been studying other like-minded cyberneticians – specifically, his older colleague Heinz von Foerster and one of his close early collaborators Gordon Pask…. I’ve also done a bit of work on the history of cybernetics so I didn’t have to waste time on establishing the basics.
The crucial point for me is that (as a historian of operations research and related areas) I would agree that Medina would not be able to claim to have done the basic originating spadework on Beer and Project Cybersyn, although it might appear that way to most historians because they themselves learned about these subjects from people they ordinarily read, i.e. Medina and Andy Pickering.
In fact, though, Beer and Project Cybersyn are pretty legendary within certain circles, and, in those circles, the basics of Beer’s career and Project Cybersyn have long constituted a kind of common historical property. The essentials can, for example, also be derived from the writing of Beer’s protégé, radical operational researcher Jonathan Rosenhead, who wrote Beer’s obituary for the Journal of the Operational Research Society in 2003, as well as an essay in the 2011 Profiles in Operations Research volume from Springer.
Also, here’s a documentary on Project Cybersyn that appeared on vimeo in 2009, two years before Medina’s book was published:
Note that the creators were aware of Medina’s work, and offered a quick thanks at the end.
Now, none of this should be taken to mean that Medina’s extensive labor is unoriginal or anything other than top-notch, or that it can, in good conscience, now be acknowledged with only an off-hand mention (even if it is a crusty three years old). You could get away without citing Rosenhead, because of the brevity of his accounts, but, in my mind, you have to give Medina credit.
So, how should the issue be handled?
I share STSer Lee Vinsel’s dissatisfaction with Janet Browne’s declaration that the issue is “now resolved” to the effect that there is no foul on Morozov’s part because he was writing in a “highbrow” literary outlet. Morozov, who has long worked as a writer, is presently a PhD student in Harvard’s History of Science Department, where Browne is department chair (and where I received my own PhD in 2007). So, she would be eager to put the issue to bed. But why should we accept this rote declaration?* It’s not clear.
At the same time, I’m really uncomfortable with historians’ rush to shame Morozov publicly. Historians have rather ruthlessly used Morozov’s pre-PhD identity as a journalist to deny him any benefit of the doubt that they would ordinarily extend to one of their own. In fact, some tweeters have used it to reinforce a portrait of him as an untrustworthy outsider (historians do have a propensity to be clannish).
They are repulsed that academic norms of thorough citation have not been accommodated in other genres (despite the conclusions of every STS study ever on the translation of standards across sociocultural boundaries), and take it for granted that Medina and Pickering are the originating sources on Beer and Cybersyn, which must be acknowledged even in a watered-down version of academic norms that prevail in vulgar “highbrow” publications. (The unfortunate implication is that if historians aren’t aware of it, it mustn’t count.)
Personally, while I agree that Morozov has sinned, I believe the sin is venial. I think the prior existence of a substantial lore on Beer and Cybersyn does constitute an extenuating circumstance, and, for this reason, I have been forced to essentially invent a principle to express the professional discourtesy that I, like others, sense has taken place.
The fact is, we have no good guidelines for adjudicating issues of this sort.
To me, the whole episode points to a more general failure to have pointed, civil, and conclusive discussions about policies/best practices of attribution, as well as any means of ascertaining what constitutes something that is generally “known” and what needs to be attributed to a specific author.
One might, for instance, just as easily ask (to take a recent example from this blog) whether Simon Schaffer should have given a better sense of existing critical thought about the historical role of automata. Personally, I think his argument was different enough that he was OK, though I do think it would have been ideal to have given some sense of the depth of existing thinking about them.
The point, though, is that we really don’t have any sort of guidance on issues such as these, and we foreclose the possibility of creating them if issues are simply declared “resolved”.
Given that some of our most senior scholars do have connections to the “highbrow” literary world, there is even a possibility that we could establish with their editors certain protections for when subjects we write about enter the less-well-documented forums of public discourse. But that’s not something we can expect to happen implicitly or automatically.
And these issues do matter. I am a member of the History and Traditions Committee of the Institute for Operations Research and the Management Sciences. One of the other members emailed the rest of us last night to tell us about the very interesting article in The New Yorker on Project Cybersyn, which, he allowed, covered a lot of the same ground as the Rosenhead profile that he already knew about. It would have been nice if more than a single breadcrumb had been left to lead him to Medina’s excellent book.
*In my opinion, historians have a bad habit of declaring issues they don’t want to discuss anymore resolved or passé, and counting on others to follow their lead on the pain of being passé. Michael Bycroft and I feel this way, for instance, about the common claim that the internalism-externalism problem reached a satisfactory conclusion, and that we are now so beyond that.