RELATED WHY FILES:
Faces: Male and female

Kid's memories

Brains 'n computers

Sure, the blokes on top are up to the usual monkey business. But why can a six-month-old distinguish monkey faces better than a nine- month-old -- or an adult? Courtesy Science

Show an infant something new, and the little diaper-filler will spend more time gazing at it. Familiar stuff, as any teenager knows, is "borrrring" -- hardly worth a second glance.

You might not give this thirst for the novel a second thought. But a clever scientist can exploit it to divine how well infants recognize stuff they've seen before.

The study we'll examine could explain -- at last -- why so many people say, "you seen one monkey, you seen 'em all." People, oddly enough, are just lousy at telling Curious George from his relatives.

To understand changes in the ability to distinguish faces, Olivier Pascalis, in the department of psychology at the University of Sheffield (United Kingdom) and colleagues rounded up three sets of subjects: adults, and infants aged six months or nine months. These subjects saw pairs of photos of human or monkey faces. One face they'd seen before, but not the other.

The researchers tracked eye movement to measure how long the subjects looked at each picture. Based on a long series of psychological studies, they felt safe in assuming that the subjects would look longer at unfamiliar faces than familiar ones.

Face the music
All three groups could tell if they'd already seen a human face. But only the six-month-olds could distinguish monkeys. By nine months, that ability was lost. What changed over those three months? Apparently, experience trained, or "tuned," the brain about how to place faces into a "template" for comparison.

This nine-month-old knows her momma, but she can no longer recognize individual monkey faces. Courtesy Pamela Jackson

"Monkeys are always with other monkeys, and humans with humans," Pascalis says, "so experience may tune the prototype, or template," which guides recognition of faces. Not seeing monkeys on a day-to-day basis, they lost the ability to distinguish them.

The study grew from the observation that maturing monkeys began having trouble recognizing human faces at a similar stage of maturity.

A similar phenomenon occurs in language processing. Six-month infants can distinguish the sounds of almost any language, but by nine months, their ability has focused on whatever language the infant hears day to day.

At six months, babies recognize the faces of different monkeys and humans. By nine months and into adulthood, people no longer recognize different monkeys, but can still recognize different human faces.

Prunin' 'n tunin'
Although "tuning" the brain to the demands of the environment makes sense, this is startling: Instead of increasing, language and face-recognition abilities seem to decline with age and maturity -- even during the first year. "There are these really remarkable changes in the first year of life in the face-processing system," said co-author Michelle de Haan, of the Institute of Child Health, University College London. "We usually think about development as a process of gaining skills, so what is surprising about this case is that babies seem to be losing ability with age."

Of course, seen in terms of function, the tuning makes sense. As it gains the ability to perceive vital differences for telling one human face from another, the infants lose the ability to detect tangential distinctions.

Courtesy U.S. Dept. of Education.

In a parallel realm, the fetal immune system contains a greater diversity of antibodies than the adult. Normally, antibodies that are not needed (or wanted) because they attack the infant's body are pruned before causing harm.

Tuning and pruning are a trade-off. General capacities to recognize any sound or face are jettisoned in favor of a more specific and intense ability -- to understand the Swahili language or recognize Jerry Springer's face, for example.

Courtesy White House Initiative on Asian and Pacific Islanders.

Tuning and training of the brain does not necessarily end in infancy, although the details and speed of learning do change. If a certain kind of recognition becomes useful, Pascalis says, it may be developed.

For example, many people have difficulty recognizing people of different races. As Pascalis explains, during infancy, a "big tuning of face-recognition processes" lays down a template for human faces. Distinguishing different ethnic groups, he suspects, requires "adjusting the template. If you go to Africa, you may say, 'I'm lost, I cannot recognize people'. But if you stay, you eventually see that everybody is very different, and you won't have a problem."
-- David Tenenbaum

BIBLIOGRAPHY
Is Face Processing Species-Specific During the First Year of Life?, Olivier Pascalis et al, Science, 17 May 2002.

Credits | Feedback | Search

©2002, University of Wisconsin, Board of Regents.