ABSTRACT
In this chapter, we implement the methods described in Chapter 3, which combine NLP, social network analysis, and an ethnographic rhetorical analysis inspired by actor-network theory and the practice of “controversy mapping.” We undertake a detailed analysis of the most influential superusers in the 141 million tweets related to the COVID-19 pandemic and the initial chaotic debate about the public health crisis and its effect on society from January 21, 2020, to May 2020. Through our analysis, we identify that the dueling forces of emotion and reason in our models cannot be neatly encapsulated in a single preset vocabulary. Rather, our method defines emotion and reason as emergent properties of the language of the tweets composing our network graphs, as structural features of the network distribution of how language is used by different individuals and communities. The dynamic and unstable nature of the patterns we identified in the use of reason and emotion in the digital public domain reflects the insights of historical scholars of rhetoric who have long recognized that the power of rhetorical appeals to persuade lies in their malleability and their chameleon-like ability to be adapted and molded to any given line of argument.
