The Great Gatsby Plays To Mate |
By Jennifer Shahade | |
September 29, 2010 | |
Have you ever wondered how Crime and Punishment would fare against
Madame Bovary in a chess match? OK, maybe not. But after meeting Princeton
professor D. Graham Burnett, you may
find your chess & literature merging in more contexts than overcrowded
bookshelves. Together with philosopher and programmer W.J. Walter, Dr. Burnett created Novel Chess, a program that turns novels and other texts
into chess competitors. They first described novel chess in "Reading to the Endgame" in the Fall 2009 edition of the art & culture magazine, Cabinet. A chessboard consists of sixty-four squares commonly designated by alphanumeric coordinates (a-h across the x-axis and 1-8 up the y-axis). If one were to replace the numerical assignations with a continuation of the alphabet (running, for instance, i-p up the y-axis), each square would be designated uniquely by a two-letter coordinate that we will call a "tuple." Now imagine setting up a simple computer program that knows the rules of chess-nothing more. It knows, for instance, all the moves that are makeable by a given piece, and it can keep track of a chessboard (updating what pieces are on which squares as moves are made). Suppose further that this program takes directions for making moves in the form of a pair of "tuples"-namely, one letter-pair designating the coordinates of a square occupied by a movable piece, and then a second letter-pair designating the coordinates of a square to which that piece can be legitimately moved. We now have everything in place to convert two texts into a game of chess: we simply feed the program the two novels, asking it to play one text as "white" and the other as "black"; the program searches through the white text until it finds the first tuple corresponding to a movable piece (in the case of an opening move, either a pawn or a knight), and then, having settled on the piece that will open, continues searching through the text until it encounters a tuple designating a square to which that piece can be moved. When it has done so, the computer executes that move for white, and then goes to the other text to find, in the same way, an opening move for black. And so it goes: white, black, white, black, until-quite by accident, of course, since we must suppose that the novels know nothing of chess strategy (and our program cannot help them, since it knows only the rules of the game)-one king is mated. There are more details on how this all works in the full Cabinet Magazine piece (for which Burnett is also an editor) and you can try the game out for yourself at novelchess.org. Burnett also introduced the game to art lovers and chessplayers at the New York City non-profit Chess-in-the-Schools' (CIS) Associate Board benefit, "The Algorithm", held this summer at the White Box Gallery. Burnett talked
to CLO about his own background in chess, the CIS event and the critical
Huckleberry Finn--Metamorphosis face-off. Jennifer Shahade (JS): How did you come up with the idea of Novel Chess? Dr. Graham Burnett (DGB): I suppose the whole thing probably emerged out of a long-standing interest in codes and cryptography. Not a scholarly interest or anything. Just an ordinary preoccupation. Having read a bit about various ways of encoding information within what appears to be plain old writing, I began to play around with various more and less paranoiac ways of "reading." Reading for "hidden" messages. The origin of Novel Chess lay in these experiments. What would it be like to try to "find" the chess matches encoded in War and Peace? It was at this point that I teamed up with my dear friend and collaborator W. J. Walter, a philosopher who lives in Paris, is pretty good at chess, and a skilled programmer. We went to work. JS: When Garry Kasparov lost to Deep Blue in 1997, the press went wild. Since then, computers have furthered their dominance. What does novel chess add to the idea of artificial intelligence? DGB: I think we are more interested in intelligent artifice than artificial intelligence. The "game" of Novel Chess involves a reimagining of works of literature through the chess board. Our little algorithm "reads" a pair of novels against each other on the board, converting strings of text into sequences of moves. Novel Chess turns the basic engine at the heart of a computer chess program into a tool for the production of a new kind of writing. Not wholly new, of course, since we have been inspired on this project by the work of George Perec, François Le Lionnais, and others-the French experimental writers who worked under the collective program known as OuLiPo, the "Ouvroir de Littérature Potentielle," or the "Workshop of Potential Literature." These guys, who were most active in the 1960s and 1970s, had a strong interest in the intersection of mathematics and language, and they were especially obsessed with games-chess above all. They liked the idea of "story-making machines," and that is sort of what we are after with this project. A machine that makes new stories out of old stories, and that sets constraint conditions for rewriting old stories so that they can do new tricks. JS: What types of texts seem to do well in competition? DGB:Well, as we write in our little essay on the experiment, it seems as if French novels do particularly well. Go figure. We're still running larger batch matches to see if that finding is robust. We'll see. JS: If you were to design a book to defeat all comers in Novel Chess, how
would you go about it? DGB: In the newest version of the software-the one currently online -you can actually upload any text you want. This makes it possible to write novels or short stories specifically structured to defeat others. This is a moderately entertaining, if slightly insane, exercise. At one level it's a little like writing poetry, since you have to encode the moves in particular letter combinations, and you have to do so while keeping track of the overall frequencies of letters in the text, since these ultimately determine the coordinate plane on which the games play out-a recursive problem. It is a bit of a mind-screw. It is easier to make editorial "adjustments" to a given novel, to help it defeat some other one. In our original version Kafka's The Metamorphosis defeated Huckleberry Finn. That drove us crazy. So we tarted up the first chapter of Huck Finn so it could mop the German. Very satisfying. Could you write a novel or short story that was especially good against various opponents? Maybe, but it would be very tough. What gets selected as a possible move changes depending on the configuration of the board, and that is a function of the opponent. JS:What is your background and interest in chess itself? DGB: Actually I learned the game from Walter himself, who taught me as a boy. I am a solidly poor player. But a wannabe. The kind of guy who shows up at the chess clubs in the West Village, gets humiliated, goes home, goes back. I used to think someday I might be good, and in college I used to play these endless games with another guy who thought he was a genius. We took it very seriously. Games went on for days. We both sucked, of course. I have a chess clock on a shelf in my living room, but it is basically for show. JS: I recently read an article about a chess program in which people did not know which piece they were playing with until they clicked on it! This reminded me of your project and also two of mine, roulette chess & poker chess, since they all play with randomness and chess. Critics of such projects say that chess is a pure game because it's 100% skill and that introducing randomness into it makes it like so many other games, backgammon, poker, etc. How would you respond to a statement like that from a chess purist? DGB: I love this
(Quantum Chess)! Back in college, when
I was still fantasizing about my future as a grandmaster, I used to devour
myself in towering and preposterous rages when I would lose one of our big
showdowns. At the heart of that anger was a sense that chess measured
intelligence. This is, of course, quite
silly, or at least that's what I like to say now that I have accepted that I am
a crap chess player and always will be.
The Italian writer Primo Levi once argued that chess players and writers
were both very irritable sorts, because, in his view, they shared a
predicament: there is no one but yourself
to blame for failure. In that sense, I
am with the purist: there is something
simultaneously terrifying and magnificent about the games of "pure" skill. JS: Tell us about the event you partnered with Chess-in-the-Schools with. How did you present novel chess as a public spectacle? DGB:It was pretty great, I think. We did it at this gallery downtown, and there was a really nice crowd. A French ukulele band played a couple of sets, and they were excellent! The organizers had these huge projection screens on the walls where Novel Chess matches were unfolding at all different speeds-it was a party. I spoke for a while, made people laugh, and eventually got heckled by an intoxicated guy in a cowboy hat who, it turned out, had been Rothko's studio assistant back in the day. Seemed like a bum, but owned the building. Welcome to the Big Apple! JS: What do you think serious chessplayers can learn from this game? DGB: I am in awe of competitive chessplayers, and hesitate to project from my pea-sized chess brain any suppositions concerning the inner life of truly skilled players. I guess I would mostly hope that they would find the thing amusing, a little leavening for their grey matter. Those players who have an appetite for experimental literature, literary (post) modernism, and the relationship between contemporary art and information theory may be more patient with the whole madcap enterprise. Play around with the game yourself at novelchess.org and read the original Cabinet Magazine article. Find out more about D.Graham Burnett on his Princeton homepage. See more photos from the Algorithm on the CIS Associate Board website. The next event will be held at Mark Murray Fine Paintings on October 14, 6:30-9:30. Contact [email protected] for more info. |