Recognizing Faces

Recognizing an offender is one of the most important tasks for a witness. If the witness gives a clear and accurate description, the police's chances of finding the guilty party increase. If the description is incorrect, however, there is a risk that an innocent person becomes a suspect or is even convicted.

Research shows that eyewitness testimony is sometimes the only evidence in a criminal case. A review of verdicts in England and Wales found hundreds of cases where eyewitnesses were the sole evidence. In roughly half of these cases there was also only a single witness. It is therefore important to understand how reliable human memory actually is.

Recognizing is not the same as describing

Most people have no trouble recognizing a person they meet every day, such as a coworker or a teacher. But if they are asked to describe that person's face in detail, the task becomes considerably harder.

Recognizing a face and describing it appear to use different cognitive processes in the brain. Our descriptions therefore tend to be fairly general, such as "dark hair", "normal nose", or "brown eyes". Such descriptions rarely help the police identify a suspect.

When we recognize a face, the brain needs to do several things:

• determine whether the face feels familiar,
• remember where we recognize the person from,
• and finally connect the face to the correct name or identity.

It is common to succeed at the first step but fail at the second. We recognize the person, but cannot remember where we have seen them before.

When memory places the face in the wrong context

An experiment shows how easily this can happen. Participants read an article about a sexual offense. The article contained two photographs: one of the offender and one of a good Samaritan who had intervened to help the victim.

A week later, participants were shown six photographs and asked to identify the offender. Half of the participants were shown the actual offender among the pictures, while the other half were instead shown the good Samaritan.

When the offender was among the photographs, he was identified by about 40% of participants. When the good Samaritan was shown instead, about 50% pointed him out as the offender.

The participants had recognized the face, but they remembered the wrong context. The face was familiar, but they linked it to the crime instead of to the helper.

What do we look at in a face?

Studies that have tracked people's eye movements show that we do not look equally at all parts of a face.

Most attention is given to the hair and how it frames the face, as well as the eyes. Considerably less attention is directed at the mouth, chin, and cheeks.

This can make a person harder to recognize if the hair is hidden by, for example, a hat.

We also pay extra attention to faces that have something unusual, such as a very high forehead, a strong chin, or a large birthmark. At the same time, we risk paying less attention to the rest of the face.

Research also shows that people often remember attractive faces better than average-looking faces, and that unusual faces are easier to remember than faces that look more "average".

Prosopagnosia — face blindness

Most people recognize faces without thinking about it. For people with prosopagnosia, also called face blindness, this does not work.

Prosopagnosia is a neurological condition that makes it difficult or impossible to recognize faces, even those of family members or close friends. Vision functions normally, but the brain has difficulty processing faces. People with prosopagnosia therefore often use other cues to recognize people, such as voice, hairstyle, clothing, or how the person moves.

What makes faces harder to recognize?

Facial expression

We can recognize a facial expression without recognizing the person themselves. If we first see someone smiling and later meet the same person when they look angry or scared, recognition becomes harder.

This can affect witnesses, since an offender often shows strong emotions during the crime but looks considerably calmer when the police later show photographs.

The angle of the face

We more easily recognize a face if we see it from the same angle as before. If a person is first observed from the side and later shown straight on, recognition becomes harder.

Research suggests that the brain has different neurons that respond to faces seen from the front and in profile, respectively. The best recognition often occurs when the face is shown at a three-quarter angle, since both of these systems are activated at the same time. This is one reason why photographs of suspects should preferably be taken from several different angles.

Faces from other ethnic groups

On average, people find it easier to recognize faces from their own ethnic group than from other groups. This is usually called the cross-race effect.

One possible explanation is that we develop strategies for recognizing the faces we encounter most often while growing up. A light-skinned person may, for example, unconsciously place great weight on differences in hair color and eye color when recognizing people. These features often vary more between light-skinned people than between dark-skinned people. When the same strategy is used to recognize dark-skinned people, it therefore becomes less effective, which increases the risk of confusing different individuals.

In the same way, people who most often encounter dark-skinned individuals develop strategies adapted to the faces they see in everyday life. If they then need to recognize light-skinned people, those strategies do not work as well either.

This does not mean that people from different ethnic groups look more alike than others. Instead, the problem concerns experience and which facial features the brain has learned to pay attention to. The more experience a person gains with faces from different ethnic groups, the better their ability to tell them apart usually becomes.

Expectations affect memory

Our expectations can also affect how we remember a face. In one experiment, participants looked at the same photograph of a man for 30 seconds. They were then asked to create a composite sketch from memory. Half of the participants were told that the man was the captain of a ship, while the other half were told that he was a mass murderer.

The composite sketches turned out differently even though everyone had seen the same photograph. The group that believed the man was a mass murderer created images where he looked less intelligent, more cruel, and more unpleasant. The experiment shows that our preconceptions and expectations can affect how we remember a face.