The Liminal Space Between Art and Data: Du Bois’s Data Graphics as Disciplinary Misfits

Abstract

At the 1900 Paris Exposition, renowned sociologist and civil rights activist W. E. B. Du Bois presented over sixty hand-crafted charts, graphs, and maps documenting the conditions of African Americans since the end of slavery. Part data and part art, these data graphics do not fit neatly under the scholarship of a single discipline. This thesis analyzes the significance of these graphics using an interdisciplinary approach, applying semiotic, psychoanalytic, and socio-historical frameworks that are inspired by the methodological traditions of art history and Visual Culture Studies along with the graphic theory of Jacques Bertin. In a novel interpretation of Du Bois's data graphics, this thesis elucidates how through the medium of data, Du Bois visualizes African American “Progress” as non-linear and complex. This study broadly discusses how visual encodings capture cultural meanings that inherently complicate the narrative of the data itself, presenting a counterargument to several of the conventional data visualization principles established by Edward Tufte. This study also theorizes a novel semiotic framework, in which visual elements are evaluated based on a continuum of semantic complexity, to analyze how data visualizations communicate meaning beyond the data.