Born-digital documents are a challenge for the field of textual scholarship. Technological obsolescence, privacy concerns and overwriting practices hinder access to and reconstruction of born-digital writing processes (Kirschenbaum 2009; Richle 2007; Ries 2018; Bénédicte Vauthier 2019). When several textual stages have taken place within the same digital document, that has not been saved as different versions, it is no longer possible to distinguish phases based on the characteristics of separate documents (such as De Biasi’s typology of genetic materials). The medium also affects writing itself. Digital text is produced, evaluated and revised in smaller cycles than on paper, making it more recursive (Goldberg, Russell, and Cook 2003; van Waes and Schellens 2003). Also, there is a greater attention to the ‘cosmetic’ side of the document - spelling and other formal matters are revised continuously. This begs the question if and how we can make meaningful phase demarcations within a continuous, fluid process taking place in a single Word document. Within our larger research project, Track Changes: Textual Scholarship and the challenge of digital literary writing, we are using the keystroke logging software Inputlog to document the work processes of literary writers (Leijten and van Waes 2013). This provides us with large quantities of fine-grained data that require computational filtering and analyses. A reconceptualisation of ‘phases’ can help us on our quest for new analytical tools to better handle the challenge contemporary writing poses. In this talk, I will propose a number of complementary methods to analyse the born-digital writing process. My case study is the writing process of the novel Roosevelt, by the Flemish author Gie Bogaert, registered by Inputlog. I will attempt to use the amount and types of revision, the chronology, the linearity of text development, and the usage of metamarks as indications of writing phases.
|Status||Gepubliceerd - 2019|