Babies Can Count
In 1980, Prentice Starkey (1980) persuaded 72 mothers to bring their young babies to his laboratory at the University of Pennsylvania for a novel experiment. While seated on its mother’s lap, each baby, aged between 16 and 30 weeks, observed slides projected on a screen. The slides contained two or three large black dots spread out horizontally. Starkey varied the spacing between the dots so that neither the total length of the line nor the density of the dots could be used to discriminate their number. After numerous trials, Starkey noticed that the average fixation time of 1.9 seconds for a two-dot slide jumped to an average of 2.5 seconds (a 32 percent increase) for a three-dot slide. Thus, the babies detected the change from two to three dots.
In a follow-up experiment, Strauss and Curtis (1981) at the University of Pittsburgh repeated this format but used colored photographs of common objects instead of dots. The objects varied in size and alignments, so that the only constant was their number. The babies continued to notice the difference between slides of two and three objects. Similar experiments with infants have been conducted by various researchers, even as recently as 2006 (Berger, Tzur, & Posner, 2006). They all yield the same finding: In the first few months of life, babies notice the constancy of objects and detect differences in their numerical quantities. Babies, of course, do not have a sophisticated concept of counting, but they do have a conception of quantity, or what scientists call numerosity.
Is numerosity innate, or is it something the babies were able to learn in their first few months? Newborns can distinguish two objects from three, and perhaps three from four. Their ears notice the difference between two sounds from three. It seems unlikely that newborns could gather enough information from the environment to learn the numbers 1, 2, and 3 in just the few months after birth. Thus, this ability seems to have a strong genetic component.
More support for the notion that numerosity is prewired in the brain comes from case studies of patients who have lost or never had a sense of numbers. Butterworth (1999), for example, describes a patient who had a stroke that left her language and reasoning abilities intact but destroyed her ability to estimate or determine the number of objects in any collection. After another patient had an operation to remove a tumor from the left side of her brain, numbers had no meaning for her.Once again, this patient’s language ability and general intelligence were unaffected, but she could not even be taught finger addition. The multiplication tables were just a nonsense poem to her.Butterworth also describes a man who apparently never had number sense, although he earned a college degree in psychology. He had to use his fingers for simple arithmetic and resorted to a calculator for other computations, although the answer had no meaning for him. He was unable to tell the larger of two numbers or to quickly count just three items in a collection.
Dehaene (1997) examined how one’s sense of numbers can be disrupted after a stroke. One patient counted about half the items in a collection and then stopped counting because she thought she counted them all. Another patient would count the same items over and over again, insisting there were 12 items when there were only four. Here, too, language ability and general intelligence were not affected. These are just a few examples from a large collection of case studies that lead to one conclusion: We are born with a built-in number sense!