From the History of Network Visualization, Indicators of Interdisciplinarity, to the Career of Metaphor: Formal and Computational Methods and the Illumination of Patterns Otherwise Unknowable

Yung, Vincent

doi:https://doi.org/10.21985/n2-9zc9-6m88

Work

From the History of Network Visualization, Indicators of Interdisciplinarity, to the Career of Metaphor: Formal and Computational Methods and the Illumination of Patterns Otherwise Unknowable

Public

Download PDF

Download All Files (.zip)

From visualizing thousands of social relationships, operationalizing measures that offer us insight into our theories, to characterizing cultural change over the span of centuries, computational methods appear to provide sociology a means to reveal patterns otherwise unknowable. While these methods are persistently critiqued for their inability to generate social theory, some contemporary scholars have argued that computational methods may even change how we theorize social life. My dissertation takes inspiration from these debates. I offer three self-contained studies that demonstrate how various forms of formal and computational analysis, including early forms, permit us to advance social theory in a variety of ways. My first chapter shows the importance network visualization. Existing historical accounts of the origins of Social Network Analysis tend to emphasize the foundational role of mathematics in unifying otherwise isolated approaches to the study of social structure. Visualization, a cornerstone of contemporary network research, has a relatively understated role in these accounts. Through an archival analysis of three early precursors of network research, I demonstrate how network visualizations impacted how each thought about, theorized, and communicated ideas about social structure. My second chapter compares two divergent approaches to interdisciplinarity. The Mertonian approach assesses interdisciplinarity through the intermixing of disciplinary actors, while the Geertzian approach assesses it through the intermixing of ideas across disciplinary actors. Using the case of higher education scholarship, I use co-citation network analysis and topic modelling to operationalize each approach into complementary indicators of interdisciplinarity. My findings provide systematic evidence in support of the disciplinarity of higher education scholarship, but with a few complications. More broadly, my findings demonstrate that the Mertonian and Geertzian approaches offer complementary insights and indicators that may be used to more consciously define, assess, and facilitate interdisciplinary scholarship. Finally, my third chapter examines the evolving use of the network metaphor. I offer a formal model for visualizing and interpreting what metaphors describe and do and how they change and conventionalize. Specifically, I characterize the career of metaphor in the context of the network metaphor, which I leverage as both a revelatory case for evaluating my model and a critical case for deepening our historical understanding of the network. Using computational techniques based on co-occurrence, I find that the network metaphor underwent significant changes in use over the first half of the 19th century, stood in for anatomical structures of the body in the mid-19th century, and conventionalized prior to the start of the 20th century. I use my dissertation to reflect on how computational methods may be used to theorize systematically and may impact how we theorize. I argue that, in some respects, computational methods supplement rather than succeed conventional approaches to theorizing. In other respects, however, computational methods impact how we theorize in ways that are genuinely unprecedented.

Creator