In this article I analyse the hidden genealogical link between portrait photography, used for criminological and psychiatric purposes, and contemporary systems of biometric identification of the human face. The aim is to highlight the shift between the emphasis on the importance of the human ‘expert eye’ in recognizing the face when talking about nineteenth-century photography and the use of computer technology that produces and reads digital facial images. In both cases, however, these are modes and variants of reducing and flattening the human face; the face itself disappears under the onslaught of technologies of vision and mediation, becoming a mere data set. Special attention is devoted to the pose, the frontal view, which is the technological a priori of the empirical possibility of recognition, articulated through the history of visualization of the face from early portrait photography to facial recognition systems. Consequently, what we call the ‘face’ is a simulacra of individual the face presented to the apparatus: the ‘real’ face is transposed upon multiple technological layers (camera – plates – photographic surface), losing its characteristic features to be re-written according to specific techniques of measurement.