ABSTRACT

Over the past 20 years, numbers and arithmetic have been the subject of intense scientific research (Butterworth, 1999; Dehaene, 1992; see Fayol, 2012 for an overview). The conjunction of studies from different fields, though mainly cognitive psychology and neuroscience, has led to the widespread notion that human infants are endowed with number sense. This innate system, which is also present in animals and human adults, is responsible for basic knowledge about numbers and their relationships. In both children and adults, its activation during straightforward numerical tasks is linked to that of a brain network in which the parietal lobes play an essential role. This dominant view has led several researchers to try and determine the nature of this protonumerical skill. Two different definitions have come to the fore: first, an innate, approximate system—shared by nonhuman animals—that makes it possible to discriminate and approximately represent visual and auditory numerosities without the need for verbal counting; second, another representational system that facilitates the exact representation of very small numerosities (Butterworth, 1999; Dehaene, 1992; Feigenson et al., 2004; Nieder & Miller, 2003, 2004; Wynn, 1995). However, the development of the ability to calculate and quantify with precision also requires the mastery of symbolic representations specific to humans. This symbolic knowledge has only emerged within numerate cultures, and is learned through explicit teaching (Piazza et al., 2013; Pica et al., 2004).