Data-Realism: Reading and Writing Datafied Text

In this forward-looking and somewhat tentative call for a realism of data - or a data-realism - we chart out three literary experiments all relating to one of the most pressing aspects of the contemporary conditions of writing, namely the role of data-driven support software including machine learning. Questioning how to read the text’s data - or the data’s text - we develop our approach in dialogue with three works of e-lit, arguing that these works present three distinct, yet related, ways of engaging with the underlying data-infrastructures of computerized language. Thus, we read these three works of e-lit in order to examine (1) how predictive keyboards change the notion of who is writing (or texting) whom in Erica Scourti’s THINK YOU KNOW ME; (2) how Google’s AdWords algorithm creates and manages a linguistic capitalism as investigated by Pip Thornton’s {poem}.py; and (3) how the style of generative text is becoming both recognizable and re-mixable, forming a machine learning rhetoric, evident twitter user @KeatonPatti’s viral “Olive Garden Tweet.” Refraining from presenting a closed conclusion, we instead speculate the relevance of our data-realist approach by presenting an e-lit meditation of the concept, taking the form of the latest iteration of The Poetry Machine: “The Oracle from Selphie.”