In the first week of this course we engaged in a twitter conversation about what exactly constitutes a poem (and lyric poems in particular). The part of our eventual definition that found the most consensus was that a lyric poem spoke from the heart: an emotional glimpse of a perspective frozen in time. Because of that guiding premise, we always have the opportunity to construct a coherent mind behind the speaking voice of the poem, a persona we infer and reconstruct based on the evidence the poem presents to us. This axiomatic premise of lyric is a profoundly humanist gesture: that when human beings write, it is natural to assume both writer and audience wish to represent human beings and the things that human beings find interesting (in other words, that art is all about us, human beings).
But humanism is a premise that may not have any necessary foundation in reality. If you’ll remember back to my page on the intentional fallacy, one of the reasons why it’s possible to second-guess the author’s intention (even an author who bothers to tell us explicitly what he or she meant by writing a poem) is that language belongs not to any one individual, but to all of us. It was there before we arrived, and will continue on (hopefully) once we depart. So perhaps it’s language that writes the poems we read, and we ought to deemphasize the credulity with which we approach the author’s agency in constructing a text (or in this case, the “author”), and for that matter perhaps we should deemphasize our own agency in coming up with authoritative readings of texts — perhaps both author and reader are mere conduits, ventriloquized by language. Take that alternate premise to its logical extreme, and perhaps one could remove human beings from the entire poetic equation (whereby you get posthumanism). This week’s pages are about the multiple ways poets have experimented with gesturing outside the self, outside of humanist conceptions of poetry, audience, authorship.
For example, Google Poetics:
This is just the latest in a string of found poems that function strictly via anaphora, the name for a series of lines that all start with the same word or phrases. In this case, you type in a search phrase on google and wait for it to offer to finish the phrase with whatever hits are most common. This isn’t random, but you might be surprised by what the search engine suggests. Then a user who thinks the list is poignant emails a screenshot to the google poetics people and away we go. This is a species of found poetry, but instead of finding a different text to make into poetry, the premise here is that (because of the anaphora), it already is a poem. Also, the question of who wrote the original series of suggested links is up for grabs. It wasn’t google, or the designer of the search algorithm (they are doubtless as surprised at any of us about what comes up!). And it’s not the person who found it, though they make a claim for us that it is interesting, certainly. In a sense, because the google search is populated by all users using the search up to and including a given time, we all wrote it. Another good example of this is the Pentametron, a twitter bot engine that finds two rhyming lines in iambic pentameter from completely different users and retweets them simultaneously. Part of this is an attempt to show how prevalent the iambic pentameter line is in common English parlance. Part of it, too, is to show us all how even randomly collated tweets can be a form of poetry — even when the users themselves have no idea they’re rhyming with someone else out there. [Note: Seriously, if you check out only one twitter account reference on this page, check out Pentametron and scroll through the tweets it collects. It is amazing.]
Here’s a more obscure example: the markov chain, an algorithm by which someone might, say, take all of the works of Emily Dickinson, pour the collective vocabulary into a document, and have a computer randomly select words and put them into her recognizable syntax. (In case you’re not familiar, Dickinson’s rhythm and meter of choice was patterned off of hymns; this is why you can sing most of her poems to “Yankee Doodle”). This is what VCU’s own Tom Woodward did when he developed the Emily Dickinson Markov generator. It’s a completely randomly generated set of poems and — of course!! — not all of them work. But every so often it does make sense, and when it does it’s fascinating — who is the author of such a poem? It’s not completely Dickinson, although she did supply the “words”: it’s the diction available to a woman like Dickinson which she internalized to the extent that those words appeared in her body of work. And it’s her familiar rhythm and meter, so…. But no, she did not write those poems. Neither, strictly speaking, did Tom: he merely designed a site and sat back to watch the poems unfurl. Next good question: Is the computer-generated poem interesting on its own, or only because we say it is interesting? And is it only poetry when it most closely resembles Emily Dickinson’s originals? There’s an element of the Uncanny Valley to all of this as well: the freaky feeling that machines and humanity are converging ever so slightly. Other examples of twitter-based Markov chains: MoPoBo, This is Just to Say bot, and EzraPoundSign.
It has gotten to the point where human beings are actually trying to simulate computers’ ability to badly simulate human beings, as in the case of the Horse eBooks phenomenon: