Scientists shed new light on speed of infant learning

Scott Johnson is an Associate Professor of Psychology at New York University Full size image available through contact

The question of how and when we develop our knowledge of object behavior – such as knowing that when a ball rolls behind a sofa, that it is likely to roll out the other side – is an ongoing puzzle in cognitive science. Previously, scientists had thought that infants learned to understand this concept through manual exploration. However, subsequent research indicated that infants developed an understanding of objects even before they had the ability to reach and grasp, leading scientists to postulate that object knowledge could be something babies are born with.

But new research by a team of psychologists, led by Scott Johnson of New York University, provides the first conclusive evidence that infants actually learn object concepts at a very young age – between three to six months – and that they do so through visual observation.

The research sheds new light on how soon and how quickly infants learn, as well as their ability to build an understanding of object concepts through stationary observation of the standard home environment. The findings were published in the August 25 issue of the Proceedings of the National Academy of Sciences.

"Our research provides the first conclusive documentation of how and when infants learn about object concepts, and serves as a strong argument against theories that infant knowledge in this area is innate," said Johnson. "It had previously been presumed that six-month-old infants could not have had enough time to acquire this type of knowledge, but what's truly amazing is how rapidly they're able to pick up these concepts."

To conduct their experiments, the researchers employed a unique eye-tracking experiment with four- and six-month-old infants, who were shown a 32-inch computer screen depicting a ball rolling horizontally back and forth. After two minutes of exposure to this trajectory, the infants were then shown a ball moving across the same screen space but its movements were temporarily obscured by an occluding box. Using a special camera that records eye movements from the center of the pupil and the cornea, data were gathered through an infrared signal tracking the various positions of the eye as the infant observed the moving ball in various stages. Of central interest was whether the infant anticipated the ball's emergence from behind the occluder, as if she had represented the ball's continued existence despite being out of sight.

Full size image available through contact

The experiment showed that infants who were exposed to the unobscured trajectory were better able to learn to anticipate the movement of the ball when it was later obscured than infants who had not been shown the unobscured trajectory. The team also found that babies who were six months of age had already grasped the fundamentals of object concepts, suggesting that they had learned such representations from real-world experience viewing objects.

"Another implication of our findings is that infants do not necessarily benefit from stimulating toys or exercises; most babies will grasp these concepts quickly through visual observation, rather than manual object manipulation" Johnson added.

The experiment marks the most sophisticated and sensitive research conducted to date of infant learning through eye movements, and the findings mark a shift from the nativist supposition – which believed learning of object concepts to be innate – back to the developmental theories famously espoused by Piaget in the 1930s. Going forward, Johnson and his colleagues will build on these results by using event-related potentials to measure activity in the brain while the infants are learning in real time.

The PNAS paper, entitled Development of Object Concepts in Infancy: Evidence for Early Learning in an Eye Tracking Paradigm, was co-authored by Dima Amso of New York University and Jonathan A. Slemmer of Cornell University. For a copy of the paper, contact Shonna Keogan at shonna.keogan@nyu.edu or Scott Johnson at scott.johnson@nyu.edu.

Scott Johnson is an Associate Professor of Psychology at New York University, specializing in the study of visual and cognitive development, especially in infancy. He earned a bachelors degree in 1985 and a Ph.D. in 1992, both from Arizona State University. The studies were funded by grants from the National Science Foundation and the National Institute for Child Health and Human Development.