Gary Edmond* and Natalie Wortley

London’s Metropolitan Police has recently established a team of “super-recognisers” to identify suspects. Limited attention has been given to the use to which their evidence may properly be put during investigations, formal interviews and prosecutions. This article explores the ways investigators have approached the identification of persons of interest in crime-related images and the use of this evidence at trial. It explains that the courts have largely been inattentive to scientific research; particularly notorious difficulties and the (un)reliability of much image interpretation and comparison. Following a review of admissibility jurisprudence in England and Australia and relevant scientific research, it concludes that the strategic use of those with enhanced abilities — to recognise familiar faces and to match unfamiliar faces — would improve the reliability of identifications and offer the potential to circumvent the dangers of unreliability, bias and contamination that threatens current police and expert practice.

identification; images; expert evidence; CCTV; police familiars; super-recogniser; Code D
Click here to read extracts of the article

This article reviews the ways investigators in England and Australia use crime- related images to identify persons of interest (POIs).1 It considers the ways in which identifications by investigators, police familiars, facial mappers and other putative experts are used in investigations, police interviews and prosecutions. It also explores the emergence of a relatively new type of actor, the police super-recogniser.2 Following a review of the very different approaches to evidence of identification by police officers in England and Australia, the article provides a summary of scientific research relevant to facial recognition and face matching. This review is revealing because it suggests that the legal regulation of image interpretation in both jurisdictions has been, and continues to be, largely inattentive to what is known by scientists about the difficulty and (un)reliability of facial perception and comparison.
Our review of the admissibility jurisprudence and relevant scientific literature questions practices in both jurisdictions. Liberal English approaches appear too accommodating and Australian approaches are, as we explain, unnecessarily exclusionary. In consequence, we offer a partial solution that might help to repair a vexed area of law in ways that largely avoid the need to rely on the identification evidence of investigating police officers, police familiars and facial mapping experts. Strategic use of independent super-recognisers offers the potential to discipline the interpretation of images by the state while circumventing reliance on conventional identifications that tend to be compromised by uncertainty, inadequate regulation and indifference to scientific knowledge.