The ability to mentally represent the exact numerosity of up to four perceived objects, as well as approximately estimating differences in numerical magnitude, appears to constitute a core-cognitive number sense. Symbolic representations of number (e.g., “two” and “2”) have been argued to gain meaning through a mapping against the analog nonsymbolic numerical magnitude representation (e.g., ••). Alternatively, symbolic number processing abilities may develop independently of nonsymbolic numerical cognition, instead dependent on learning the order and content of the verbal count-list (i.e., 1, 2, 3, …). This thesis aimed to determine which of these proposals best correspond to the brain’s processing of numerical information, with implications for the development of mathematics curricula.

Four neuroimaging studies provide biomarkers for typical numerical cognition. Results indicate that symbolic numbers increasingly acquire semantic reference from other symbols; nonsymbolic quantities are processed in an asemantic visuospatial manner; neural correlates reach adult-level maturity at 11 years of age; numerical order and magnitude recruit independent mechanisms; and that maturation of executive functions and lexico-semantics is key for symbolic number processing. These results support the view of increasingly independent mechanisms for symbolic and nonsymbolic numerical cognition across development.

Förmågan för exakta mentala representationer av upp till fyra objekt, samt att estimera skillnader i numerisk magnitud, antas utgöra grunden för ett evolutionärt nedärvt antalssinne. Förståelsen för symboliska numeriska representationer (”två” och ”2”) förmodas vara grundad i en mappning mot dessa ickesymboliska magnitudrepresentationer (••) i antalssinnet. Alternativa teorier hävdar att symbolisk antalsbearbetning utvecklas fristående från ickesymbolisk numerisk kognition, där semantisk förståelse framställs ur den verbala räknelistans (1, 2, 3, …) ordning och innehåll. Syftet med denna avhandling var att undersöka vilket förslag som bäst motsvarar hjärnans antalsbearbetning, med följder för framtida matematikundervisning.

Fyra neurovetenskapliga studier resulterade i biomarkörer för typisk antalsbearbetning. Resultaten påvisar att symboliska representationer alltmer förstås genom jämförelse med andra symboler över utvecklingstid; ickesymbolisk kvantitet bearbetas asemantiskt och visuospatialt; neurala korrelat för antalsbearbetning tangerar vuxen mognadsgrad vid 11 års ålder; numerisk ordnings- och magnitudbearbetning stöds av fristående mekanismer; och att mognaden av exekutiva samt lexiko-semantiska förmågor utgör grunden för symbolisk antalsbearbetning. Tillsammans indikerar resultaten att symbolisk och ickesymbolisk numerisk kognition progressivt stöds av fristående kognitiva mekanismer över utvecklingstid.

Numerical cognition can take place in multiple representational formats, such as Arabic digits (e.g., 1), verbal number words (e.g., “two”), and nonsymbolic (e.g., •••) numerical magnitude. Basic numerical discrimination abilities are key factors underlying the development of arithmetic abilities, acting as an important developmental precursor of adult-level numeracy. While prior research has begun to detail the neural correlates associated with basic numerical discrimination skills in different representational formats, the interactions between functional neural circuits are less understood. A growing body of evidence suggests that the functional networks recruited by number discrimination tasks differ between children and adults, which may provide valuable insights into the development of numerical cognition. To this end, we posed two questions: how do the interactions between functional circuits associated with number processing differ in children and adults? Are differences in functional network connectivity modulated by numerical representational codes? A theoretically motivated 22 ROI analysis indicated significant functional connectivity differences between children and adults across all three codes. Adults demonstrated sparser and more consistent connectivity patterns across codes, indicative of developmental domain-specialization for number processing. Although neural activity in children and adults is similar, the functional connectivity supporting number processing appears subject to substantial developmental maturation effects.