If there is one thing our brains do that emphasizes how social we are, it is recognizing faces. Babies — before they learn to speak or even babble — are instinctively drawn to look at faces, and by four months of age are already processing faces as distinct objects. There is a good reason for this — faces are incredibly informative. A single glance reveals much about a person’s identity, gender, ethnicity, emotion, and truthfulness.
That humans recognize faces so easily is somewhat surprising given how similar faces are. They all consist of the same basic parts — two eyes, a nose, two ears, and so on — with the only differences being small variations in the color, shape, and spacing of features.
And yet a face is never just a face; it is always someone’s face. Unlike birds, chairs, or cars, which we identify mostly by category, we treat every face as unique. In addition, once we have seen a face, we can recognize it under many conditions, whether it’s obscured in shadow, displaying some extreme facial expression, or even a caricature. Ongoing research in this area is helping us:
- Identify the brain circuits and cognitive processes involved in recognizing faces.
- Better diagnose and develop treatments for face blindness and other related disorders, including autism, all of which affect social interaction.
- Develop technologies for security and law enforcement.
Like many other brain functions, much of what we know about facial recognition comes from those who lose the ability. This condition, known as prosopagnosia or face blindness, sometimes occurs after a stroke or brain injury. It leaves people with normal vision but no ability to recognize the faces of people they know well, often including spouses and children. Although such people often develop tricks to help recognize people, such as focusing on voice, mannerisms, or gait, they suffer significant social and professional challenges.
Most people with prosopagnosia suffer damage to the fusiform gyrus, which runs along the underside of the brain, along with other areas. Scientists believe that a specific region of the fusiform gryus, the fusiform face area, plays a vital role in face recognition.
Research has also revealed much of the general cognitive process behind face recognition. Brain scientists know, for instance, that faces are processed “holistically.” We do not focus on specific face parts — such as just the eyes or mouth — but rather note information about many parts of the face as well as its configuration, (the distance between features). This information is then processed to create a generalized model of a face.
This procedure allows us to recognize a face despite the many possible variations in how we encounter it. Face recognition software, on the other hand, finds this task particularly difficult, often struggling to match the same face seen in different contexts.
In recent years, imaging techniques that allow scientists to peer into living brains have provided even more insights. Researchers now know, for example, that face-processing brain areas interact with memory networks. This allows us to identify who we are talking to and recall our history with that person. Emotional memory is particularly important. Scientists suspect that a faulty connection with emotional centers leads to Capgras syndrome, in which people can recognize loved ones but believe they have been replaced by impostors.
Experience also matters. For instance, although most of the brain processes needed to recognize faces are in place by age two and fully developed by age 10, we make more mistakes as children than as adults, suggesting it takes time to learn to recognize faces well.
There are some people who may fall behind permanently. Because they spend less time looking at faces, children with autism may never fully develop the ability.
Researchers found that face processing works much like language. Just as, early in life, we can make many sounds but lose the ability to pronounce ones foreign to our native language, our ability to distinguish foreign faces also drops off. Six-month-olds can distinguish between macaque monkeys, for instance, but nine-month-olds cannot unless they regularly see monkey pictures.
Perhaps the most significant finding in recent years is that prosopagnosia is far more common than previously thought. Researchers now estimate that congenital prosopagnosia (genetic face blindness) affects about 2.5 percent of people — more than 7 million in the United States alone. Another 10 percent are believed to be poor face processors.
Congenital prosopagnosia comes in different forms — many, but not all, of those with the condition have trouble recognizing locations, and brain scans only sometimes show less activity in the fusiform face areas. The exact causes of congenital prosopagnosia remain murky, but because symptoms often run in families, researchers are hunting for genes that might contribute.
This research has practical applications. Better understanding is helping computer programmers develop more accurate face recognition software for use in security and law enforcement, and also to improve cameras and photo-searching technology. For those with prosopagnosia, this research may offer new ways to cope with the condition, raise awareness, and eventually lead to possible treatments. And the research may offer insights into other conditions in which face recognition problems sometimes occur, including autism, schizophrenia, and bipolar disorder.
Sacks O. (2010) Face-Blind, The New Yorker, August 30, p. 36-42
Pascalis O, de Haan, M, and Nelson CA (2002) Is Face Processing Species-Specific During the First Year of Life? Science 296(5571):1321
Gauthier I, Skudlarski P, Gore JC, and Anderson AW (2000) Expertise for cars and birds recruits brain areas involved in face recognition, Nat Neurosci, 3(2):191-197.
Haxby JV, Hoffman EA, Gobbini MI, The distributed human neural system for face perception, Trends in Cognitive Sciences, 4(6):223-233