In an essay on Cordite poetry review – ‘Picture becomes text, becomes writing: software as interlocutor’ – Christopher Funkhouser and Sonny Rae Tempest transform an image into text, encode it and then translate the outcome to produce the poem ‘Exit Ducky?’
It is, in effect, a repurposing of information from image to text, from representation to poetry. It is an experimentation that’s increasingly familiar, both within the tradition of conceptual poetics and in recontextualising information. Our understanding of politics in literature is often tied to the content or ‘message’ of a work – so what does this mean for generative poetry?
In artistic practice, there is a history of repurposing and plagiarism that predates the digital. The pseudonymous Comte de Lautréamont, a French poet who died in 1870, whose work later influenced the Surrealists and Situationists, said:
Plagiarism is necessary. Progress demands it. Staying close to an author’s phrasing, plagiarism exploits his expressions, erases false ideas, replaces them with correct ideas.
A man of his word, Lautréamont plagiarised in his two major works: Les chants de Maldoror and Poésies.
Throughout the twentieth century, procedures for producing texts were developed by the Surrealists and Dadaists. As computers became available, they were quickly adopted by poets to automatise pre-existing algorithms. Nowadays, the ubiquity of computers makes these processes far more accessible.
It is almost too obvious to mention that our relationship to information has changed because of computers: white-collar jobs, for instance, all include some form of collating data, processing information or creating new information. All this pre- and post- processed data surrounds us, structuring whole sections of society from law to literature. Ultimately, we have a staggering amount of information that can be recontextualised, remixed or reused.
Kenneth Goldsmith, conceptual poet and founder of UbuWeb, argues that with the current state of digital culture in literature we see a situation where the mainstream and the avant-garde ‘find ourselves in the same boat, grappling with new questions concerning authorship, originality, and the way meaning is forged’. What is interesting about these methods of writing is the way in which they eschew the romantic notions of the writer and instead take on forms that mimic production. Instead of mixing ink with ‘great ideas’, the raw materials of writing is writing (or data) that is reprocessed into new writing. It presents a challenge to authorship – an undermining of the specialist writer – through a process that allows anyone to write.
But this means that there is no reason for the romantic notions of the specialist to persist. The reduction of all labour to unskilled labour would be a reasonably favourable outcome for capitalism: after all, what publisher wouldn’t want bestsellers at the push of a button? So while it may be a radical position to see literature as production, the concept implies little about the politics of that position.
Generative writing, however, is never purely production, and is never merely a process of ‘push button’ (though that may be an interesting experiment in itself). The form, like most forms of literature and art, is actually a lot more complicated: it requires the selection of the source texts, the choice or creation of the method/algorithm/program and, often, the editing of the final works, then followed by the process common to all creative writing – throwing away the pieces that don’t work.
Late last year the Gnoetry daily people published a chapbook of their poetry composed from various computational methods of generating poetry. The poems are composed from different source texts and use a range of different computer programs to produce the works. The Gnoetry daily poems also highlight one of the other forms of recontextualised text that have been increasingly enabled by technology – the leak. Poet Eric Elshtain, in the introduction to the chapter of poems ‘Executing Poetry Politically: Using a Machine to Comment on the States of the World’, writes of his interest in ‘the ways in which language generated/manipulating machines can make poetic statements politically’. The chapter, which features his own poems but also those of other poets, uses WikiLeaks diplomatic cables, Hurricane Katrina sources and other fundamentally political and literary texts.
Using content from WikiLeaks to produce political poetic texts immediately questions the politics of those cables. Goldsmith, in Uncreative writing, writes ‘in its self-reflexive use of appropriated language, uncreative writing embraces the inherent and inherited politics of the borrowed words: far be it from conceptual writers to dictate the moral or political meanings of words that aren’t their own.’ So what are the politics inherent in a leaked document? Goldsmith cites as one illustration Vanessa Place’s Statements of fact, wherein a set of legal documents concerning a sex offender are recontextualised as literature (Place is a criminal lawyer who defends sex offenders). The confrontational nature of the text is directly tied to the lack of exposure these forms of documents typically get. Language is dictated by context and so then, suggests this work, are the politics. The leak, by being revealed and published, transforms itself from a document that hides an injustice to a revelatory document that may help to challenge that injustice.
The politics in literature, therefore, can be read perhaps as contextual to the political climate. The writing of early Surrealists, for instance, can be stripped of their political content or context as we become temporally removed from them. Using WikiLeaks documents as a source of poetry, then, has a political dimension precisely because of the context those leaks have at this moment in time. They aren’t letters from the seventeenth century; they are the activities of our governments now. Charles O Hartman writes in the ‘unconclusion’ to his book Virtual muse: experiments in computer poetry: ‘What I find interesting is that these experiments, which are so simple from a computer science point of view, can help remind us in a new way of things we already knew about poetry and about language.’ When applied to leaked diplomatic cables, military documents and political speeches, these new texts remind us of things we already know about. Decade-long wars, for example.
This post was originally published on the Overland literary journal website as part of the Meanland project: Reused, rewritten.