Between the Lines. A Digital Critique of Literary Representation

Traditionally, the critique of literary representation hinges on ‘symptomatic’ forms of reading in which textual elements are regarded as expressive of deeper lying ideological issues (Jameson 1981, Althusser & Balibar 1997). In the Digital Humanities, a textual approach is operationalised which mainly targets measurable elements on the textual surface (Best & Marcus 2009). For a digitally oriented critique of representation, this creates a tension between the text’s surface and it’s ‘hidden’ depths that reside between the lines. Digital methods for literary studies are evidently useful for describing, categorizing and measuring surface elements of the text, but the study of literary representation mostly focuses on that what is not directly visible or apparent. How could digital methods contribute to the analysis of ideological constructs that lie dormant in the margins of the text?

In this talk, the merits and pitfalls of a Digital Humanities critique of representation are considered by focusing on character representation in contemporary Dutch literature, as representational hierarchies between characters have been a fruitful point of departure for symptomatic critiques of representation (e.g. Pattynama 1994, Meijer 1996, Pattynama 1998, Minnaard 2010). The focus lies on a sample corpus of 170 recent Dutch novels containing ± 2100 characters, of which extensive demographic information (gender, age, place of birth, place of residency, level of education, profession) has been gathered  (Van der Deijl, Pieterse, Prinse, Smeets 2016), as well as thematic roles (family, lover, colleague, friend, enemy) between all characters. Through a network analytical approach, every character in the corpus is ranked according to statistical centrality measures in order to provide a quantitative view on the centrality of certain character demographics.

The resulting character rankings will be used to explore the character representation of gender, ethnicity and class in the corpus. To what extent do such empirically informed, data-driven patterns on the textual surface reveal something of the deeper lying ideological issues in the text? It is argued that symptomatic and surface reading are not mutually exclusive, but can potentially enhance one another, provided that one stays away from problem solving rhetorics (Cecire 2011) and neopositivist assumptions (Ramsay 2011, Tom Eyers 2017).