Cent mille milliards de poèmes

In English, the title translates to One hundred thousand billion poems, or else One hundred million million poems, or else One hundred trillion poems. The French have a crazy way of expressing numbers. (For instance, the series of quantities 4, 20, 9, 89, 10, 9, 19, 99, 96, 80, 16 would be said aloud as quatre, vingt, neuf, quatre-vingt-neuf, dix, neuf, dix-neuf, quatre-vingt-dix-neuf, quatre-vingt-seize, quatre-vingt, seize. Apparently with enough wine you can pronounce a hyphen.) In the case of this particular book, however, such combinatory flexibility is quite apt, for the author Raymond Queneau here proves his mastery at leveraging the power of recombination.

I don’t know much about poetry, but when I imagine a book of one hundred trillion poems, I imagine a very thick book. As we read in the postface by François Le Lionnais, who assisted Queneau, “The work you are holding in your hands represents, itself alone, a quantity of text far greater than everything man has written since the invention of writing, including popular novels, business letters, diplomatic correspondence, private mail, rough drafts thrown in the wastebasket, and graffiti.” In fact, at one poem per standard 20-pound sheet of paper, my calculations suggest the book would be approximately 6 million miles thick. (Thanks to the electronic god-brain Wolfram Alpha for confirming my math.) But no! Though rather wide and tall, Queneau’s book isn’t very thick at all:

How, you ask, does the genius Queneau do it? The magic happens with ten sonnets and a pair of scissors. Each sonnet not only follows the same rhyme scheme, but bears the same rhyming sounds in the same places—so that one can replace, say, the final line of one poem with the final line of another, without breaking any formal rules. Cutting slits on either side of each line of each poem yields a fanfare of recombinable lines, easy to rearrange into one hundred trillion possible sonnets:

In his introductory “Mode d’emploi”—a term that would resonate through the OuLiPoan tradition that this book, in 1961, set rolling—Queneau says that his little book allows “everyone to compose” these trillions of sonnets, and that he hopes his little sonnet-machine will always produce poems with a formal and grammatical regularity, as well as “a theme and a continuity” to hold them together. There is a gratifying viscosity, a density in moving through this thicket of sonnet material—quite like getting chased by tigers through a bamboo forest:

As with so many strange books, this one proves difficult to handle. My fingers are so accustomed to the wholeness of the sheet, they become trembly and useless with these strips splaying themselves out everywhere. Happily, Gordon Dow has produced a more user-friendly online version of the original text, and Beverly Charles Rowe provides multiple digital versions in English. The internet seems as apt a destination as any for this strange book; scholars of new media writing have long seen Queneau’s many sonnets as a precursor to the textual combinatorics that hypertext and related technologies enable. (Indeed, Noah Wardrip-Fruin and Nick Montfort provide an extremely limited excerpt of Queneau’s book—percentage-wise, probably the smallest excerpt ever—in their wonderful The New Media Reader.) But when the day is done, there’s nothing quite like physically holding the uncanny slimness of this very big, very strange book. You can get such a “bookbound” copy here.

A Million Random Digits with 100,000 Normal Deviates

First published for the RAND Corporation in 1955 by the Free Press of Glencoe, Illinois, A Million Random Digits with 100,000 Normal Deviates contains exactly what you’d suspect: an extremely large table of random digits. One can hardly resist the impulse to open this hefty volume, but good luck actually reading it. Viz.:

Production of the so-called RAND Book began at the Los Alamos National Laboratory in 1947. Scientists there were developing a new branch of mathematics called the Monte Carlo Method, which uses random sampling to model complex systems. As it happened, the main complex system in need of modeling was the chaotic reaction inside the thermonuclear weapons that RAND scientists were then helping to design at Los Alamos. Of course, this is one of several important random number tables, but its connection with Cold-War weaponization is particularly interesting. Stanislaw Ulam is credited with the initial idea for Monte Carlo mathematics, and he developed it in collaboration with John von Neumann and the astoundingly named Nicholas Metropolis. Enrico Fermi had already used random sampling techniques in some of his experiments at the University of Chicago in the 1930s, but only in the late 1940s and 1950s at Los Alamos would they become formalized and see routine use. As the Introduction to the first edition puts it, “the applications required a large supply of random digits or normal deviates of high quality,” and the RAND Book was produced to answer this need. It was created using an electronic roulette wheel in conjunction with a random-frequency pulse. The random pulse likely was a gas noise tube, though as Tom Jennings’s wonderful reconstruction proves, a geiger counter trained on uranium ore would have worked equally well—while also lending the whole affair a delightful circularity. (Jennings has previously reviewed the first edition of the RAND Book.) Monte Carlo mathematics has since become important for complex systems theory, chaos math, environmental science, and a variety of other disciplines, and the RAND Book remains among the most trusted and widely used tables of random digits. A copy is now available online, but for those who prefer to feel this volume’s heft in their hot little hands, the RAND Corporation graciously issued a second edition in 2001, available on Amazon. The tables themselves remain unchanged, of course—in fact, like the first edition, they were not conventionally typeset but reproduced directly from computer printouts, by photo-offset, in order to avoid typographical errors. The new edition does, however, provide an updated Foreword in celebration of this very strange book. Also available is a thrilling sequel by David Dubowski, who uses computer algorithms to provide an equally large table, now with more perfectly random distribution—whatever that means.