Send message to Author

Paper short abstract

This paper examines selected 19th-century practices such as phrenology, craniometry, and electrical study of emotion to interrogate assumptions about surface and depth, fixity and mutability, in 21st-century biometric technologies using facial recognition.

Paper long abstract

The human head—as living head, desiccated skull, or expressive face—was a privileged bodily site for past medical and anthropological sciences as they attempted to categorise and predict individuals’ heredity, criminal inclinations, or medico-nervous susceptibility. Phrenology, craniometry, physiognomy, and electrical study of emotions are some notable examples from the nineteenth century. Many of their ambitions persist in today’s emerging biometric technologies, as does the primacy of (a part of) the human head, as visually captured and analysed through automated processes facial recognition. But when a ‘face’ is photographed and evaluated using AI, is it simply the visible surface of the face that is taken to encode relevant bodily markers? What might that mean for attaining precision or stability in measurement and classification? This paper teases apart some ways issues of surface and depth, fixity and mutability were configured in historical practices of medicine or surveillance, notably 19th-century anthropology, physiognomy, and physiology. For instance, where the ‘criminal man’ or electrically and hypnotically induced emotions were intended to be perceptible at a glance, from the surface of the face, physical anthropologists like Paul Broca insisted that only skulls provided sufficient stability, only painstaking measurement protocols sufficient precision, to identify hereditary anomalies. Ultimately, the paper suggests differences between these traditions remain important for understanding assumptions, values, and norms inscribed into 21st-century facial biometrics.