Super-recognizers and the Legal System

Abstract

Contemporary research into exceptional face recognition performance has largely been driven by practical demands. Consequently, so-called “super-recognizers” have occupied various forensically relevant operational roles, yet their potential utility as expert witnesses in the courtroom is currently under debate. Here we critically evaluate this suggestion, focusing on the relevance principle (e.g., R v Turner, 1975); specifically, can super-recognizers reach face recognition conclusions beyond the reach of the standard juror, which would allow them to present facial recognition opinion evidence as fact? To do so we critically probe the empirical literature to assess whether super-recognizers use different mechanisms to recognize faces than typical perceivers, whether they can consistently outperform the latter on face recognition tasks, and whether the two groups are limited by the same perceptual biases. Last, we compare the accuracy of face recognition judgments derived via innate and trained routes to assess whether super-recognizers perform comparably to the humans and machines currently permitted to provide relevant evidence in the courtroom. Interrogation of the literature suggests that we should proceed cautiously. We argue that the full utility of super-recognizers cannot be addressed until screening methods and testing batteries are standardized across investigations.

Alejandro J. Estudillo

Principal Academic in Psychology

My research interests include face and object recognition, numerical cognition, neuropsychology, etc.