PhD defence on ” Discursive Cartography: Computing and Interpreting Maps of Online Communication”

SODAS PhD student Thyge Ryom Enggaard will defend his dissertation "Discursive Cartography: Computing and Interpreting Maps of Online Communication" on Friday 16th June 15.00-18.00 CET.

Title: Discursive Cartography: Computing and Interpreting Maps of Online Communication

Date: 16th June 2023, 15:00-18:00 CET

Zoom-link: https://ucph-ku.zoom.us/j/68168086760?pwd=RjhLbnJNRTlyRnhwNUtjS2M3T3cvUT09

Abstract:  The use of computational models has become a promising and fairly established part of many social scientific approaches to analyzing texts. Fundamentally, however, the quantification involved in computation seems incompatible with how we naturally read and understand texts, including being insensitive to the many contexts surrounding the text. This dissertation contributes to the development and application of computational methods for exploratory discourse analysis, where researchers do not seek to test pre-established theories, but rather seek to learn about various properties of a given text, potentially with the aim of subsequently developing new theories. In particular, the dissertation advocates a cartographic approach to exploratory discourse analysis, in which the role of computational methods is two-fold. First, computational methods are used to track and model the associations between elements in the text, such as the associations between words or the ways authors are associated with words. These include counting and tracing selected associations, as well as embedding entire matrices of associations. Such computational methods are useful because they help characterize one particular context, namely the ways in which textual elements are used within the text itself. Second, computational methods are used to produce visual maps of the computed associations between textual elements. These maps serve as a compression of the modelled associations, that help facilitate the identification and interpretation of patterns in the text at hand. The strength of these maps is the ease with which they can be shared, discussed and criticized, within and outside a research project. Since a computational analysis of a text cannot restore the wider contexts that surround the text, the dissertation also experiments with how a cartographic approach to exploratory text analysis can be fused with (n)ethnographic observations of the wider settings in which the texts emerge.