Critical Tools: Using Technology to Augment the Process of Literary Analysis

Abstract

When it comes to the arts and sciences, Northrop Frye argues that “it is clear that the arts do not stabilize the subject in the same way that science does. . . The stabilizing subject of science is usually identified with the reason; the unstabilizing subject is normally called the imagination”. Since the nineteen eighties, with the institutionalization of Humanities Computing research, there have been attempts at combining humanistic questions with technological innovations, and by extension, scientific concerns. Within the digital humanities there is a tension between these two positions that often results in the neglect of the human analyst and an elevated use of technology when applied to tool design. This can be seen in the current trend of distant reading, which is the batch processing and analysis of text corpora using machines. This approach stands in stark contrast to close reading which traditionally in English studies has entailed looking at individual words and their relation to a text as a whole in terms of not what the text means, but how it means. In this thesis I argue that the bridge between technology and literary criticism can be built using digital tools as long as those tools allow access to both the reason of science and the imagination of art. I present four digital projects that each investigate this problem in a novel way: (1) I use an algorithmic approach to investigate T.S. Eliot’s own theoretical position in terms of his diction, (2) I designed and developed a visualization of the English language, LDNA, that can be recovered back into the original text, (3) I conducted a study with 14 expert literary critics to analyze their current methods and used these results to design a tool, MetaTation, that can be integrated into the literary critical process, and (4) I also demonstrated how evidence-based testing of literary theory can be done in the context of Engineering writing by conducting a study that tests the feminist theory of false universals in human-computer interaction literature. I use these projects to present a hybrid approach that answers the question: How do we reconcile the specificity and human dependent nature of an unstable and imaginative close reading with the historic breadth and reason of a distant reading approach?