Monotonic Coding of Numerosity in Macaque Lateral Intraparietal Area

Open Access
Publisher Website
Google Scholar
Abstract
As any child knows, the first step in counting is summing up individual elements, yet the brain mechanisms responsible for this process remain obscure. Here we show, for the first time, that a population of neurons in the lateral intraparietal area of monkeys encodes the total number of elements within their classical receptive fields in a graded fashion, across a wide range of numerical values (2–32). Moreover, modulation of neuronal activity by visual quantity developed rapidly, within 100 ms of stimulus onset, and was independent of attention, reward expectations, or stimulus attributes such as size, density, or color. The responses of these neurons resemble the outputs of “accumulator neurons” postulated in computational models of number processing. Numerical accumulator neurons may provide inputs to neurons encoding specific cardinal values, such as “4,” that have been described in previous work. Our findings may explain the frequent association of visuospatial and numerical deficits following damage to parietal cortex in humans. As any child knows, to answer the question “how many,” one must start by adding up individual objects in a group. Extending beyond humans, this cognitive ability is shared by animals as diverse as birds and monkeys. Surprisingly, the exact brain mechanisms responsible for this process remain unknown. Damage to a brain area known as the parietal cortex disrupts basic mathematical skills, and functional imaging studies show that this area is activated when people perform basic computations. To understand how parietal cortex contributes to numerical behavior, we studied the activity of neurons in this area in monkeys while they looked at arrays of dots on a computer screen. We found that parietal neurons responded progressively as the total number of elements in the display was varied across a wide range of values (2–32). These neurons resemble “accumulator neurons” that have been suggested to serve as the first stage in counting. This information could be used by other neurons that respond best for a particular cardinal number, such as “4,” as has been reported in prior studies. Our findings support computer models that separate the processes of summing and numerical identification, and may also explain the fact that parietal cortex damage causes both numerical and spatial confusion.