“Super-Recognizers”: A Crime-Fighting Super-Power


photo: Kevan, creative commons license

The ability to recognize faces is a neurological trait that some people are simply better at than others. You can test yourself here. People at the lowest end of the spectrum lack this perceptual ability altogether. In these extreme cases, mothers cannot recognize their own children; colleagues don’t recognize someone they’ve worked with for years. At this level, the condition is called prosopagnosia, “face-blindness,” and some degree of difficulty recognizing faces may affect about 14 million Americans.

For many years, interest in this trait focused on people who have problems recognizing faces. When recent scientific advances indicated the trait exists on a continuum, this opened interest in people who have a superior ability to recognize faces. Detective Chief Inspector Mick Neville of London’s Metropolitan Police Service (the Met) thought he had a job for them: identifying criminals.

London is the perfect place to test Neville’s idea, according to a fascinating article by Patrick Radden Keefe in The New Yorker. London has the densest concentration of closed circuit television (CCTV) cameras in the world—more than a million of them, mostly in the hands of homeowners and businesses. Keefe quotes former London Mayor Boris Johnson as saying, “When you walk down the streets of London, you are a movie star.”

Crime fiction writers will have a field day with this. The “super-recognizers” seem ideally suited for solving cold cases and identifying suspects in real time. On the other side of the courtroom, smart defense attorneys—I’m thinking Mickey Haller here—might chip away at the facial-recognition ability of “eye-witnesses.”

In the 1990s, installation of cameras was promoted throughout London as a crime prevention measure, but it turned out to be a weak deterrent. There were too many images, they were too hard to analyze, and though the camera recorded lots of crimes, nothing came of this evidence, because the images couldn’t be matched to specific people. Last weekend, NewYork/NewJersey bombing suspect Ahmad Khan Rahami was captured on camera at both Manhattan bomb sites, but it was the fingerprint left at the scene that led to his identification and the match with the man seen on camera.

Early on, Neville headed a unit that analyzed this CCTV footage, trying to make identifications. It was slow work. But when he learned about super-recognizers, he saw the potential benefit of recruiting people who might be extra-skilled at the process.

Now a small, dedicated unit within the Met is assembling an image database, which has more than 100,000 pictures of unidentified suspects in crimes recorded by CCTV. Unit experts compare these images with mug shots of known criminals. They collect images of the same individual at different crime scenes; if the person in one of the images is finally identified, multiple crimes are solved. And, knowing when and where multiple images of the same person were captured gives clues to a criminal’s behavior patterns.

This is, says Scientific American, a very special super-power.

Friday: The Future of Facial Recognition: Man vs. Machine?

Information vs. Confession: New Police Interrogators

Punch & Judy, police

photo: Dan Dickinson, creative commons license

Mystery and crime fiction readers (and writers!) may soon encounter a new approach to police interrogation that may be more effective at producing solid information and valid confessions. Until the mid-1930s, suspected criminals were subjected to the “third degree,” which often included bodily harm or at least the threat of it—like dangling a suspect out of a window (!).

Currently, police mostly use confrontational techniques “a rusty, stalwart invention that’s been around since the days of JFK,” says reporter Robert Kolker in the current issue of Wired.

These supposedly more scientific techniques are based on psychological manipulation, in which police attempt to persuade their suspect that confession is their only reasonable choice. Hallmarks of the technique are the claustrophobic interview room in which detectives appear absolutely convinced of a suspect’s guilt and present a damning version of facts (and even made up “facts”) that paint the suspect as the culprit. (If you want to see a memorable demonstration of this technique, check out this terrific YouTube clip from The Wire.)

The developers of confrontational interrogation justified the use of false information and other tricks because they—and many cops trained in their methods as well as judges and prosecutors—were convinced an innocent person simply would not confess to a crime he did not commit. This post demonstrates what a tragically wrongful conviction that was. Evidence against its reliability started piling up when DNA analysis became available and a large number of convictions were thrown out, even though the accused at some point “confessed.” Further, and contrary to expectation, Kolker says, “The more confident police officers are about their judgments, the more likely they are to be wrong.”

Now a growing number of police departments, starting with the LAPD, recognizes the shaky science behind these methods and are moving to an “investigative” approach more similar to that long used in England and Canada. As a joint effort of the FBI, CIA and Pentagon, the High-Value Detainee Interrogation Group (HIG) studied interrogation techniques around the world, with an eye to producing valid confessions and avoiding false ones among terrorists. Bottom line: “If you want accurate information, be as non-accusatorial as possible.” Now they are trying to spread the word throughout domestic police departments.

I can see changes in fiction—plots where one officer is trained in the HIG techniques, but the partner resists; repeat criminals unnerved by the change in police attitude; and the expansion of information police have to work with when their questioning causes suspects to simply clam up. Of course, in both fiction and real life, many skilled interviewers have used these techniques for years, without official sanction. (Fictional detective Lt. Colombo comes immediately to mind as a possible, possibly extreme example.) Any attempt to change the culture of policing is ripe for drama.

Think the Truth Protects You?

Texas, guns

(photo: C. Holmes, CC license)

Douglas Starr in the December 9, 2013,  New Yorker, describes how the most commonly used confrontational interrogation technique used in the United States leads to false confessions. The method relies on detectives’ observing non-verbal behavior, looking for (or creating) anxiety, never giving the suspect a chance to voice a denial, minimizing the crime and trying every trick to make it easier for the suspect to admit it, even claiming to have evidence they don’t have, with their right to lie to suspects in many circumstances protected by a 1969 Supreme Court decision. The goal is simply to get a confession.

Psychologists became suspicious about the issue of false confessions several decades ago and began studies on it. And experienced detectives have begun to doubt it, as they’ve seen suspects mold their statements to fit the information the detectives have fed them.

In Britain, police don’t try for confessions, they go for information. They focus on the content of what is said, not nonverbal behavior or anxiety (proved to be not correlated with lying). Instead, they look for inconsistencies: “For the suspect, lying creates a cognitive load—it takes energy to juggle the details of a fake story.” It’s hard to keep it up. Nor are the police allowed to lie about what evidence they have.

In the United States, out of 311 people exonerated by post-conviction DNA testing, more than a quarter had given false confessions—perhaps most notoriously, the Central Park Five. (Ken and Sarah Burns film on the case). Why do they confess? Worn down by the interrogation, an innocent suspect “fabricates a story to satisfy his questioners.”  This is most likely what led to the false statement made by Amanda Knox, which, although she recanted, has been used against her ever since.

This post originally appeared February 3, 2014.