James Peters-Fransen was fitted with plastic-sponge sensors that pick up electrical brain activity at a Children’s Hospital neuroscience lab.
(Suzanne Kreiter/ Globe Staff)
Hub lab writing the book on face-reading
James Peters-Fransen was fitted with plastic-sponge sensors that pick up electrical brain activity at a Children’s Hospital neuroscience lab.
(Suzanne Kreiter/ Globe Staff)
Pity the Boston car salesman who negotiated across the table from Charles A. Nelson III, a Harvard neuroscience professor who runs the nation’s top laboratory studying how people learn to decode facial expressions.
As the two men faced off in the showroom last month, the salesman insisted to Nelson that he had just offered the absolute lowest price for the German car in question, declaring, “This is it.’’
Then the salesman’s eyes darted to a vacant corner, his nose and mouth taking on a configuration that shouted “Bluff.’’ The professor ultimately left the dealership smiling, holding a contract to buy the car at a far lower price, a bargain in his estimation.
Such is one ancillary benefit of Nelson’s exhaustive research, which unfolds every day in his $1.5 million cognitive neuroscience laboratory at Children’s Hospital Boston, where he studies just when and how humans learn to read faces.
To that end, his lab recruits hundreds of babies and preschoolers from the Boston area, with staff members making pitches at day care centers and children’s fairs. Using high-tech equipment to monitor the children’s eye movements and brain activity, researchers seek to discover how people identify one face from another and how they decipher the emotions behind particular expressions.
Nelson studies why some youngsters evolve to become particularly adept at this important everyday skill that helps them distinguish Mom from a stranger, a liar from a truth-teller, and a sad look from a fearful look, and why others, particularly those with autism or other developmental disorders, are often perplexed by faces, hobbling their social interactions.
His research shows that a typical child’s face-reading skills is influenced by experience, so he seeks to answer the question: When is the most sensitive period for growth? The astute face-reader, he said, will always have an edge in maintaining relationships, getting jobs, making deals, and all other areas where the truth matters.
“What people say is not always what they mean,’’ said Nelson in his sixth-floor lab a few blocks from the main hospital. “All of us are always looking for the match between what someone says and how they look when they say it.’’
This research on infants is far from easy. Staffers blow bubbles and offer musical toys to babies as their heads are fitted with a netting of some 64 plastic-sponge sensors that pick up electrical brain activity from different regions of the brain. Scientists are looking to see what facial images trigger accentuated electrical responses, showing the infant is trying to distinguish one image from the other.
For instance, if a child’s brain activity is the same when shown two identical faces, though with different skin tones, it suggests the child fails to notice skin color, Nelson said. But if the level of brain activity changes, it suggests the child is sensitive to pigmentation.
In other studies, researchers track a baby’s eye movements to see how they scan certain faces and for how long.
Some babies sit perfectly still on their parent’s lap looking at pictures that flash on a computer screen; others are bribed with cookies to keep them from squirming. (Parents receive a nominal $10 fee, and each child receives a toy.)
Last Thursday, nine-month-old Grace Strano from Lexington called it quits after about five minutes; still researchers applauded the 60 pictures she sat through, and the data they picked up.
Nelson’s research has yielded fascinating insights about how young brains work. One of his major findings is that babies begin face-decoding skills very early in life, starting around 6 months of age. Over the next several years, children typically become captivated by faces reflecting fear, then later they differentiate faces showing sadness, happiness, anger, surprise, and disgust.
Through adolescence, children pick up additional subtle skills, such as detecting anger in a slight jaw tightening, but learning has slowed down.
Nelson said that his research has revealed that by adulthood, people’s abilities to read faces are fairly established and they can only learn more through deliberate learning, much like being taught a foreign language.
“By adulthood, the window is not permanently closed, but with age, it’s increasingly difficult to learn new abilities,’’ said Nelson, 56, a high-energy researcher who travels frequently to talk about his work.
Nelson and researchers in other labs have long noted that gifted face-readers scan faces in a holistic way, and their brains store these facial images and associated emotions in a quickly retrievable memory bank. If babies develop normally, they focus on another person’s eyes, then they take in a snapshot of the whole template of the face. Children learn to notice how the eyes, eyebrows, nose, and mouth relate to each other, suggesting gradations of emotion from rapture to rage. They use these images to understand people and form relationships.
Autistic children, scientists say, typically avoid looking at other people’s eyes, for reasons that remain largely mysterious, depriving themselves of the most emotion-revealing feature. Subtle changes in people’s eyes - the way they squint, enlarge, or the way the eyeballs move around - project a wide diversity of feelings. If autistic children focus on any feature, it’s often the mouth.
Although most people may be wired to find face-watching a generally pleasurable experience, Helen Tager-Flusberg, a Boston University professor of anatomy and neurobiology who is in a joint project with Nelson, said that autistic children’s brains may find it too intense, even aversive. These findings have important implications for treatments for autism, with some therapies rewarding these children for looking directly at people’s faces in conversation.
Nelson’s lab has just begun to use a new piece of high-tech infant head gear, which will help them monitor blood flow to different regions of babies’ brains as they look at faces.
Nelson said he knows key brain regions are gearing up in the early months, then go “online’’ after about six months in a more specialized, efficient way.
In a study published in Science magazine in 2002, he found that 6-month-old infants were as good at distinguishing between similar-looking human faces as they were similar-looking monkeys, but just a few months later became much better at human faces. Nelson interpreted the results to mean that, after 6 months of age, babies begin the process of trying to “become expert’’ at the faces they need for survival.
Children’s skills are deeply affected by their exposures: If they are of French ancestry, but raised in Kenya, they may specialize in discerning African faces in their daily environment, not white ones.
Nelson said negative experience can also have an impact. Children who have been abused by their parents, researchers say, often show an exceptional ability to detect anger based on the most split-second changes in expressions.
Nelson said he believes his own face-reading abilities come from a childhood fascination with faces, as well as his decades of research. He insisted, however, that motivated adults can learn to read faces better by paying more attention to faces. But he said these subtleties are not easy and may not always have one interpretation.
Referring to the controversial “Hope’’ poster, showing a picture of President Obama looking up, Nelson said, “How do we know the difference between contemplation and spaciness?’’
Patricia Wen can be reached at wen@globe.com. ![]()



