Blackout Poetry + Drunk Poet = Blackout Drunk Poems

The other day, I was on Reddit and stumbled across this example of applying the “blackout poetry” method to a page CS Lewis’ Narnia novel, The Silver Chair. Fair warning, it’s pretty lewd:

For those unfamiliar with blackout poetry – well, you’re probably able to piece together how it works from the picture above. It’s the practice of taking one text and covering up all but a few words in a sort of reverse-Mad Libs fashion, and seeing what you end up with. I’ve never tried my own hand at it before, but I would imagine that blackout poets exercise varying degrees of precision while scribbling out words. You could black them out at random, or in a fixed pattern, or with a particular outcome in mind – e.g., turning a page from children’s book into a slice of erotic fiction.

It occurred to me that blackout poetry has quite a bit in common with the computer-generated poetry we’ve been looking at. For one thing, both genres are often limited to a small, predetermined corpus of words. Yet both also accomodate a massive number of potential outputs, even with a dictionary a few dozens of words large. It’s not inconceivable that if I took the same un-marked page from The Silver Chair, I could scribble out the words to make it read like a horror novel, or an action movie.

One key difference that I see between the two methods is that whereas many computer generated poems effectively work to make sense out of nonsense – that is, take a pool of words that are meaningless on their own, then apply some template and create (at least somewhat) meaningful sentences from them – blackout poetry takes something that already makes sense, and tries to twist it so that it makes a different kind of sense. If I’m playing by the rules, I don’t get to cut the page up and rearrange the words how I like them. My sharpie is my only weapon. I’m not sure that we’re at a place yet where computers are able to perform this kind of abstract thought, at least not without a good deal of human assistance.

But hey, maybe I’m wrong. Maybe we can make some great literature with nothing more than random number generator and, say, a Donald Trump speech.
Well, I gave it a whirl.
I copy/pasted the last few paragraphs of the speech (original text here), then used Word to get a word count (427). Next, I generated 15 random, unique numbers between 1 and 427. After blacking out all of the words except those corresponding to the 15 random numbers, here’s what was left:
Screen Shot 2015-09-16 at 8.57.07 PM
For comparison, here’s one I made on my own – without random numbers or any other sort of computer assistance – using the same excerpt. I cheated a little, marking out just the ’n’ in “we need” to make “weed”, but let’s be honest, if the computer knew how to do that too, it would have.

Leave a Reply

Your email address will not be published. Required fields are marked *