Authority of Ideas: Aesthetics of Biofilms and the Anthropocene

People who want to make knowledge about the physical world have relied on various technologies and representational styles to argue, communicate, and catalog their observations, experiments, and theories. There are feedback loops between how subvisual and abstract scientific concepts are represented in visual representations and how new questions are approached, developed, and incorporated into bodies of scientific knowledge. Those feedbacks underpin our ability to converge around models of the natural world that are not directly observable with human senses. Historians, philosophers, sociologists, and metascientists study those feedbacks with a variety of qualitative and quantitative tools. However, there are few computational tools that can be applied to studying visual representations directly. 
 
We apply computational methods to study the evolution of scientific representations. We develop and benchmark new methods from image processing, nonparametric multivariate statistics, and computer vision to trace the development of images from scientific publications in ways that are interpretable and that shed light on the evolution of the scientific concepts of “microbial biofilms” and “anthropocene.” We aim to understand how the emergence of novel technologies, questions, and concepts across different research communities in microbiology and climate science influenced the visual language of their research publications, and how innovation emerges in representations of subvisual and abstract concepts in scientific research over time.