Trained eyes,coded faces: Visual instruments and expertise in the sciences of facial and bodily expression

Abstract

This article traces the divergent approaches to the scientific analysis of facial and bodily movement from the 1950s to the 1970s. Drawing on the writings and exchanges of anthropologists Margaret Mead and Ray Birdwhistell, as well as psychologists Paul Ekman and Silvan Tomkins, it shows how their intellectual divide was sustained through different uses of visual technologies and coding practices. Cultural anthropologists working in dialogue with cybernetic thought prioritized the interpretive role of culturally patterned observers and remained cautious about relying on visual instruments to extend human perception. Psychologists, in contrast, oriented their attention toward the measurable physical changes captured in film and photographs, treating images less as interpretive mediators than as naturalistic recordings of human behavior. This divide crystallized around the concepts of code: either as culturally shaped rules that could be deciphered by trained observers or as discrete, machine-readable units inscribed through computerized techniques. By examining how professional identities, technological instruments, and the definitions of facial expressions as a scientific object took shape in the mid-20^th-century, this article demonstrates why these historical dynamics remain crucial for understanding the stakes of living with our faces coded in the era of automated facial expression analysis.