: Horse_Ebooks and Algorithmic Translation
By Manya Scheps
I will tell you a true story. If you believe me you will be well rewarded. If you don t believe me, – 20 Nov 2011
There is no form or method. There is only emotion. – 21 Oct 2011
Horse_ebooks is a semi-popular spam feed that utilizes the fractured syntax of robots to advertise e-books about equestrian care and maintenance. While there are thousands of spambot twitter streams, horse_ebooks is decidedly more special, given its 12,000+ followers and hundreds of retweets. Perhaps it is the brevity of the tweets that is appealing (“Being” 16 November 2011). Perhaps it is their length (“Today s women have an inward turn of mind. Identify themselves with the world around them. Create imaginative worlds. Look at the world” 19 November 2011). Perhaps we just love indeterminacy presented by animals.
To the extent that we can talk about reality and facts, we should: a Russian man named Alexei Kouznetsov owns horse-ebooks.com in addition to 197 other e-book domains like net-ebooks.com, horror-ebooks.com, action-ebooks.com, diet-ebooks.com, affiliates-ebooks.com, cholesterol-ebooks.com, all purchased through GoDaddy. These generate revenue through affiliate programs and advertising clicks (and perhaps the occasional sale of Fancy Fillies Horse Racing System). Many of these have had Twitter streams at one point, though Twitter’s spam protection shut most of them down. The ones that still exist (@AffiliatesEboox, @SolersSportsEbk, etc) consist mainly of retweets from other ebook streams, with roughly 30% of the content being self-generated. However, to say “self-generated” is misleading -— all of horse_ebooks’s tweets come from one of the ebooks available for purchase on the site. The algorithm that picks the content remains a mystery, though it is assuredly a sophisticated rendition of a term frequency–inverse document frequency measure.
The algorithm that produces the horse_ebooks stream, like most spammic algorithms, relies on user interaction to grow more effectual. It interprets text as data, and determines which keywords might best promote an outcome like the sale of Cialis or Horse Medical Records. Just as with many of our more popular and less insidious internet applications, the more interaction the algorithm gets, the smarter it becomes. The growing popularity of horse_ebooks has reciprocally allowed it to become better at generating tweets like “The Fear Of lowlife criminals With Environmental Protection” (October 30, 2011; 49 retweets).
Of course, the entire notion of pinpointing the author (it’s Alexi, it’s the algorithm) is meaningless — horse_ebooks exemplifies a specific type of authored non-authorship with roots in chance operation procedures and instutitional subversion (what the Situationists called détournement). Or, to put it less tautologically, horse_ebooks acts as translator, mimetically parlaying the source text (ebooks about horses) into tweets. This mirrored translation embodies further propagation of the origin, the form of translating that Walter Benjamin posited in The Translator’s Task. But horse_ebooks goes beyond mere self-reflexivity or generational coherence. Hundreds of people retweet each horse_ebook aphorism, consuming them with vigor and spitting them into their own streams flowing with foreign content. There is even a fan fiction Tumblr dedicated to horse_ebooks tweets, as well as an artist who illustrates the most vivid into comic-strip format. These further translations – the hip winds that derive from one ambiguous source – realize the form, perfectly consummating the Fancy Fillie.
To speak of tweeting and retweeting spam as a form may seem overly analytical, or meaningless, or both. However, I suggest that it is through such modes of communication that content becomes significant. It closes in on pure language (die reine Sprache), breaking free of the context that would otherwise be rendered by surrounding sentences. Or, as Rodolphe Gasche writes in an essay on Benjamin:
Translatability, as an objective category of the work of art, points beyond the original itself. Rather than aspiring to a fulfillment of the original, translatability indicates the work of art’s search for a fulfillment in something other than the original itself. Translatability, as a call in the work of art, calls for a liberation of the work of art from itself.[1]
“Get a leg” (21 Oct 2011) is assuredly part of a less ambiguous sentence in the ebook from which it comes. Through the algorithm’s translation, however, the term takes new meaning (which is, indeed, no meaning). Despite the apparatus of a stream, the surrounding tweets (“WHAT S THE ABSOLUTE WORST THING THAT COULD” and “Wind”) do not offer any context, nor do they interact with one another in any mode that re-commits the language to a specific signification. It is this pure language that we conceptualize as poetic. The algorithm has no artistic intention, but we cannot help but read its production as such – as something wholly original (false) and innocent (even more false). The tweets communicate gesture and impulse, facilitating an oeuvre familiar, though unrelated, to those that exist in outsider art.
We can, of course, draw some parallels between Twitter translation and contemporary visual practices – Sherri Levine’s appropriative photography, for example, or Louise Lawler’s documentary frames. However, these citational modes of working still maintain strict adherence to the idea of an author, whereas horse_ebooks (if we can even pinpoint this singularity) neither confirms nor denies such a concept. While it is true that the algorithm publishes tweets that were, at some point, somehow, written, it is non-author in the sense that it defies the binary border between ebook and reader. It imagines non-authorship in a way that even social media, with its dissolution of anonymity, can enable.
We anthropomorphize horse_ebooks, reading wonder and purity into every sentimental “Everyone is happy” (18 Nov 2011). Conflating bad, calculated tweets with vernacular art, we consume the stream with pitied bemusement – “it knows not what it does.” Like the horse it represents, the feed follows with due diligence, working ever harder to translate itself, awaiting our reiterated instruction.
[1] Rodolphe Gasché, “Saturnine Vision and the Question of Difference : Reflections on Walter Benjamin’s Theory of Language.” In Rainer Nägele, ed., Benjamin’s Ground : New Readings of Walter Benjamin (Detroit : Wayne State U. P., 1988), p. 90.
Manya Scheps is a graduate of the University of Pennsylvania, where she studied fine arts. She is the editor of New Asshole and co-founded Talking Pictures, an art theory reading group, in West Philadelphia. Her critical focus is on DIY art, its production, and its social implications.