ABSTRACT
Human numerical capacities undergo a massive development over the lifespan. This is mainly driven by formal education that imposes the acquisition of a number of structured and formalized mathematical procedures and concepts. V. Izard and colleagues tested to what extent children used one-to-one-correspondence when judging exact numerical equality without referring to verbal labels. The understanding of cardinality at the age is thus going beyond what the ANS would predict but falls short of a full-blown understanding of numerical equality. Few empirical studies have aimed at putting these models to a test in the domain of numerical cognition. B. Reynvoet and D. Sasanguie argued for a variant of the bootstrapping model where the initial symbols might be first mapped onto a precise representation of numerosity. The symbolic number system, in turn, results from the combination of the initial mapping with relational inference principles such as the successor function and language.
