Materiality, Intentionality, and the Computer-Generated Poem: Reading Walter Benn Michaels with Erin Mouré’s Pillage Laud

Intertwined with the emergence of what Ihab Hassen first called “the posthuman,” the genesis of digital poetry can partly be traced to attempts, beginning in the 1950s in Europe and the 1960s in North America, to generate poems via computer in order to demonstrate that poetry need not be the result of a romanticist notion of inspiration. There are, of course, many competing genealogies of computer-generated and digital poetry—Friedrich Block argues, for example, that the first random or probabilistic texts were written by Theo Lutz in 1959 (his “Stochastichte Text” which appeared in the German avant-garde literary journal Augenblick) using a Zuse Z 22 mainframe computer; such experiments in computer-generated poetry were then taken up by Nanni Balestrini of Italy, Jean Baudot of Canada, and Brion Gysin and Emmett Williams of the U.S. (Block 19). Kenneth B. Newell, on the other hand, cites R.M. Worthy (U.S.) as the first to compose poetry with a computer nicknamed “AutoBeatnik” in 1962—efforts that were duplicated in 1984 by William Chamberlain and Thomas Etter’s program “Racter” and Hugh Kenner and Joseph O’Rourke’s “Travesty” and then duplicated once more in 1992 by Neil Rubenking’s appropriately named “Brekdown.”¹ The methods of Oulipo (a group founded in 1960 by French writers and mathematicians) for generating poems are also frequently cited as progenitors of digital poetry; originally paper based and often using the mathematics of Boole and Fibonacci to create poems, such works were among the first to be literalized with a computer. As Italo Calvino infamously puts it, “ [T]he aid of a computer, far from replacing the creative act of the artist, permits the latter rather to liberate himself from the slavery of a combinatory search, allowing him also the best chance of concentrating on this ‘clinamen’ which, alone, can make of the text a true work of art” (13). For example, the rigid set of rules at the heart of fellow-Oulipian Raymond Queneau’s Cent Mille Milliards de Poèmes (a matrix of ten sonnets which generate one hundred trillion poems) along with its unreadability—as Queneau himself puts it, if one read a sonnet per minute, eight hours a day, two hundred days per year, it would take more than a million centuries to finish the text—make it, in Calvino’s terms, a “true work of art” and an odd variation on postnineteenth-century anti-romantic poetics.

Despite Calvino’s emphasis in the quotation above on the expression of an artist’s intentions in relation to the computer, most accounts of computer-generated poetry assume the neutrality of the machine²—an assumption which produces debate not about the production process but about what has been produced. That is, the question repeatedly asked is if the computer–generated poem is in fact a poem and if it shows that, as one critic puts it in the title of his article, “an author’s intention is irrelevant to the meaning of a literary work” (Juhl 481). Further, and not surprisingly, the answer to these questions is invariably a no—a computer-generated poem is not a poem and neither, in contrast to Calvino’s assertion above, does it reflect authorial intent nor, therefore, does such a poem have any meaning. For instance, rather than interrogate the relationship between intention and meaning, Juhl circumvents the issue altogether (and reinstates the conventional view of poetry as an insightful message delivered to us by poets who are human exemplars) by claiming that “to ‘interpret’ a computer ‘poem’ is not to interpret a poem” (481).³ In fact, as I will discuss, critics have so resoundingly rejected the very idea of computer-generated poetry that it has not been until recently that issues about the intent of these poems (as well as their treatment as material objects) have re-emerged. This essay, then, addresses the overriding need for an alternative set of literary terms for the interpretation of computer-generated texts—more specifically, it is concerned with the...