The researchers, Kerry Jordan and Elizabeth Brannon of Duke University, published their findings the week of Feb. 13-17, 2006, in the Proceedings of the National Academy of Sciences. Jordan is a graduate student and Brannon is an assistant professor in Duke's Center for Cognitive Neuroscience and Department of Psychological and Brain Sciences. The research was sponsored by The National Institute of Mental Health, The National Science Foundation and the John Merck Fund.
In their study, Jordan and Brannon used the same basic experimental design that they had previously used to demonstrate that monkeys show numerical perception across senses http://www.
Similarly, in the study with seven-month-old infants, the Duke researchers presented the babies with the voices of two or three women saying "look." Simultaneously, the babies could choose between looking at video images of two or three women saying the word. As they had found with the monkeys, the researchers found that the babies spent significantly more time looking at the video image that matched the number of women talking. According to Brannon, similar experiments by other researchers had not shown definitive results because of problems in their design.
"First of all, they had used arbitrary stimuli such as a number of objects or sounds like drumbeats, rather than ecologically relevant stimuli. We don't know for sure this was a problem but it seems likely" said Brannon. "Also, those experiments presented the sounds successively and so the duration of the sound sequence was a potential problem." Finally, said Brannon, previous experiments used the same subjects for multiple trials, which meant that the subjects could have learned to associate such non-numerical factors as the length of a sequence of drumbeats or the intensity of sound with numerical size. In contrast, Jordan and Brannon gave each baby only a single trial, so that the babies had no opportunity to learn anything about the stimuli.
"As a result of our experiments, we conclude that the babies are showing an internal representation of 'two-ness' or 'three-ness' that is separate from sensory modalities and, thus, reflects an abstract internal process," said Brannon. "These results support the idea that there is a shared system between preverbal infants and nonverbal animals for representing numbers."
The researchers said that further studies will test both babies and monkeys on their perception of larger numbers, to further explore the details of their numerical abilities. They will also explore whether monkeys can explicitly match the number of a sequence of sounds and a number of objects using a touch screen task in which a monkey has to choose between two visual arrays. Such experiments, they said, will help determine the psychological importance of where monkeys and babies look in their numerical matching paradigm.
Such studies have broad implications for understanding the evolutionary origins of numerical ability and how that ability has developed in humans, they said.