Images, Evidence, and the Arts of Attention

• In Event: Reproducibility and Other Problems: Practical and Institutional Responses to Contemporary Crises in Science - I

Abstract

This paper investigates the development of new technologies for scientific representation in order to understand the circumstances under which we allow ourselves to accept claims made using constructed, synthesized, and ML-generated images. When do we believe the truths that fictions can offer? Decades of research have illuminated the social processes behind the construction of scientific representations, and recent studies have shown how decidedly analog practices impact the development of scientific visualizations built from digital data (Coopmans et al. 2014; Lynch and Woolgar 1990; Myers 2008; Vertesi 2015). Using an ethnographic study of a team of archaeologists working to develop new 3D/4D tools and a focus on their collaborations with law enforcement to develop new methods to capture crime scenes, this project leverages a unique object of study – a group of scientists aiming to develop new methodological tools to serve both a clearly bounded scientific community and an unruly, high-stakes lay community – to investigate the ways scientists use imagery to show what they know and, crucially, how they go about the work of convincing others to know it themselves. I show how an orientation to scientific discovery, a disciplinary tradition of manual craft, and a commitment to justice each contribute to the production and experience of visual veracity across epistemic cultures, and how diverse arts of attention are used in the construction of proof.

Author

• Alison Gerber, Lund University