ATTENTION, AUTOMATICITY, AND LEVELS OF REPRESENTATION IN NUMBER PROCESSING

Participants performed same-different judgements for pairs of numerals in 2 conditions: numerical matching (responding ''same'' to pairs suc h as 2-TWO), or physical matching (responding ''different'' to pairs s uch as 2-TWO). In most cases, a distance effect was obtained, with the different responses being slower when the 2 numbers were numerically close together (e.g., 1-2) than when they were further apart (e.g., 1- 8). This indicates that numbers were automatically converted mentally into quantities, even when the participants had been told to attend ex clusively to their physical characteristics. As postulated by several models of number processing, (e.g., Dehaene, 1992; McCloskey, 1992) Ar abic and verbal numerals thus appear to converge toward a common seman tic representation of quantities. However, the present results suggest that an asemantic transcoding route might allow for a direct mapping of Arabic and verbal numbers, perhaps by means of a common phonologica l representation.