March 10, 2016

· Sociology · Politics · IT · Misc

You may have heard the news that Lee Sedol, a Go Master, has been defeated by a computer program created by a group of Google engineers. A second match is underway today. The Google/DeepMind team has a technical paper in Nature describing AlphaGo, the program they wrote. Various commentators have remarked on the sometimes surprising but extremely effective moves that AlphaGo made. And of course there’s the usual half-serious musings about the inevitable robot uprising that this victory portends.

On the way to work on the bus this morning I was listening to a podcast where the hosts expressed some relief that being able to make a computer that is really, really good at Go is at least still a very long way off from accidentally creating Terminator’s Skynet, the global AI that takes control of military computers, mostly destroys the world, and fights a war against the remaining human resistance. The other point of reference here, especially when it comes to malevolent game-playing AIs, is Joshua, the strategic computer in War Games—the one that, to everyone’s relief, eventually settles on “a nice game of Chess” instead of playing Global Thermonuclear War for real. Interestingly, unlike Chess, Backgammon, Checkers, and Poker, I think Go is not on the list of games that Joshua knows how to play. Besides, as someone may have mentioned on the podcast, it’s not as if skill at Go is something that would be of much use to a Skynet or Joshua anyway.

Scott Boorman’s The Protracted Game (Oxford University Press 1969).

Scott Boorman’s The Protracted Game (Oxford University Press 1969).

Or perhaps it might. In 1969, Scott Boorman published a book called The Protracted Game: A Wei-chʻi Interpretation of Maoist Revolutionary Strategy. In it, he argued that there was a fruitful and rigorous parallel to be drawn between the game of Go (or Wei-chʻi, its Chinese name) and the military strategies pursued by Mao Tse Tung and others, particularly in their execution of long-run campaigns of guerrilla warfare that often involved extended, complex frontiers of conflict (rather than focused battlefronts) and rapid shifts of apparent territorial control (as opposed to decisive victories brought about by single strategic errors). The game of Go had long been played and analyzed in China, Japan, and Korea, and was explicitly invoked by Mao himself on more than one occasion in his discussion of his strategy. A theme of the book is the value of thinking in terms of Go rather than Chess (or Poker) as a metaphor for ongoing conflict on a large and open field:

[There are] three structural characteristics basic to the motifs of the game’s higher strategy: wei-ch’i is a protracted game; it is a “war of jigsaw pattern”; and it is a game in which victory and defeat are relative phenomena. … Although one player may be defeated tactically in one part of the board, he may recover his position by strategic out-manuevering of his opponent. In the best-known Western board games of strategy, chess and checkers, on the other hand, a single mistake in tactics is—given optimal play on the part of the opponent—fatal to the blunderer. … [T]he board is blank, or nearly so, at the beginning of the game … Consequently there is no restriction on the ability of either player … to play his stones deep in hostile spheres of influence or behind any emerging “front line”. Second, the encircling mode of capture creates complex patterns of encirclement and counterencirclement and induces the development of mutually discontinuous groups … Any conscious or unconscious assumption of zones of safety will result in severe disorientation ….

The book goes on to specify the analogy in more detail and apply it to a series of Mao’s guerrilla campaigns in the 1940s. It closes with some cautions about the limits of formalizing an approach to interpreting strategy in this way, but also an argument for its advantages, in particular “the most important single feature of the formalist approach: a logical and consistent point of view from which to analyze one facet” of strategic decision making in insurgency.

The Protracted Game is one of those books that it is hard not to admire even as one struggles to know quite what to make of it. Boorman wrote it when he was nineteen years old.1 One of the intuitions you can see driving it is the effort to keep the fluidity of real conflict squarely in view while still getting some sort of analytical purchase on what is happening. While the book’s reach may exceed its grasp, I could not help but be reminded of it this morning as I listened to a discussion emphasizing the huge gulf between Go and anything of serious interest or concern in the real world of politics and conflict. Scott Boorman tried to find a way across that gulf in 1969 as a Harvard sophomore. He did not quite succeed, but perhaps we are a little closer today.

  1. His earliest publications, from several years before, are reviews of works on Chinese military history co-authored with his father. Boorman went on to publish several seminal papers in social network analysis as a student of Harrison White’s, as well as doing pioneering work on optimal searching in networks, the evolution of cooperation and other topics. His early work really is quite astonishing. He even shows up in a footnote early on in A Theory of Justice, where Rawls thanks him for clarifying an issue about how the difference principle is evaluated by individuals in the original position. [return]
All Categories


I am Professor of Sociology at Duke University. I’m affiliated with the Kenan Institute for Ethics, the Markets and Management Studies program, and the Duke Network Analysis Center. Learn more.



To be notified of updates, you can subscribe to the  RSS feed for the site.