DURHAM, N.C. -- Shape seems to be the most important factor in babies' deciphering of the boundaries between two objects, according to a Duke University experimental psychologist.

In a study reported in the current issue of the journal Infant Behavior & Development, Amy Needham, assistant professor of psychology, showed that 4-month-old infants rely more on the shape of objects than on their color or pattern to figure out that objects placed side by side were separate rather than a single unit.

The finding suggests a continuity in the way infants and adults analyze objects they see, she explained, since object shape is a critical element of adults' perception of object boundaries, or the distinction between objects. It's possible the most useful piece in the visual discrimination of adults is one of the first things humans start relying on as babies, she said.

In previous studies, Needham had established that at an early age, infants are busy learning about the world, analyzing things around them and developing the ability to see objects not as an undifferentiated blob, but as different units.

Since infants begin their learning process long before they are able to indicate their curiosity about their surroundings verbally or even by reaching for something, researchers use "looking time" as a measure of interest or analysis.

"The kind (of measure) we use is based on the assumption that babies look at things longer if they're puzzled by what they see or if it violates some kind of expectation that they have about the world than if what they see is pretty much in line with what they expect," Needham said.

In this study, researchers tracked the looking time of 48 4-month-old babies as they were shown one of three different displays, each consisting of two objects placed against each other. In the first display, the objects were similar in shape (rectangular on the bottom and rounded on the top, or roughly loaf shaped), color and pattern. In the second display, the objects were the same color and pattern but different shapes - one loaf and one rectangle. Objects in the third display were the same shape but different colors and patterns.

As the babies watched each display, a gloved hand reached into the scene and either moved both objects together, as if they were one, or moved the two objects apart.

"We found that the babies who saw the first display, the similar display, responded longer when it moved apart than when it moved together, suggesting that they thought it was a single unit and they didn't expect to see it break apart," Needham said. The researchers saw the same results with the dissimilar color and pattern display.

But with the dissimilar shape display, the babies looked longer, or seemed to be surprised, when it moved together as one piece than when the dissimilar shaped objects moved apart. The movement as one piece was counter to the babies' expectation. An additional study explored infants' baseline preferences for these displays, and these results supported this conclusion.

"So they could use this difference in shape to say 'this must be two separate pieces' even though the color pattern was exactly the same," she said. "In this display there is a conflict between what color and pattern is telling them about the display and what shape is telling them about the display, and they seem to go with the shape rather than the color and pattern. Shape is what they're using to help them figure out where the boundaries between objects are."

Since other research indicates color vision is not fully developed until sometime around 4 months of age, Needham is now studying the responses of older babies, 8-month-olds, to see if they use color and shape differently than younger babies. It may be that by 8 months, babies use both factors - shape and color/pattern - to analyze objects.