Karen Wynn and Paul Bloom, professors of psychology, said their study was intended to address the debate about whether infants are counting when they look at objects, or whether they are simply responding to color, size and other variables.
"We report here a study showing that 5-month-olds can determine the number of collective entities, moving groups of items, when non-numerical perceptual factors such as contour length, area, density and others are strictly controlled," Wynn said in the study published this month in the journal Cognition. "This suggests both that infants can represent number per se, and that their grasp of number is not limited to the domain of objects."
To substantiate her theory that infants are capable of distinguishing between different numbers of objects, and of performing simple arithmetical operations, Wynn's research group tested the infants in two phases. The 24 infants, ranging in age from four months to five months of age, were tested while seated before a computer screen.
During the first phase, known as the habitation phase, Wynn and Bloom repeatedly presented half the babies with two groups of three objects each and half with four groups of three objects each. The objects were red, filled-in circles about the size of a dime.
In the second phase, all of the infants were presented with two kinds of test trials: trials depicting two collections of four objects each, and trials depicting four collections of two objects each.
Wynn and Bloom predicted that those infants habituated to two collections would look longer at four collections during the second phase, while those habituated to four collections would look longer at two collections during the second phase. An analysis of the infants' looking times revealed this pattern of preferences.
"What we found was the infants looked significantly longer on trials showing a new number of groups than what they had been exposed to in the habituation phase," she said. "We concluded from this that babies were sensitive to the number of groups in the display, which means they were definitely responding to the number of groups, not the total amount of area or contour length."
Wynn said the study, funded by the National Institute of Child Health and Human Development, is intended to answer questions about the foundation of human mathematical knowledge.
"If we have a good understanding of what the initial mental machinery is that gives us our toehold into the realm of number, then we will be in a better position to, for example, develop educational curricula that build on those foundations in a coherent way," she said. "It will also tell us if we are trying to teach kids material that conflicts with those foundations."
Co-authors included Wen-Chi Chiang of the National Chung-Cheng University in Taiwan.