Processing of quantitative information, investigated with fMRI

meaning. A number of objects or events is non-symbolic; there are as many objects or events as the number denotes. In numerical experiments, nonsymbolic numbers are mostly presented as a pattern of dots, which are also called numerosities. When processing numerical stimuli, two classical effects are omnipresent (Dehaene, Dehaene-Lambertz & Cohen, 1998). The distance effect refers to the phenomenon that it is more difficult to discriminate two numbers as the numerical distance between them decreases (e.g. 8 and 9 are more difficult to discriminate than 2 and 9; Moyer & Landauer, 1967). The size effect refers to the phenomenon that for an equal numerical distance, discrimination is more difficult as their numerical size increases (e.g. 8 and 9 are more difficult to discriminate than 2 and 3; Parkman, 1971). The distance and size effects are present in various formats: when presented as Arabic digits (Dehaene, Dupoux & Mehler, 1990), as verbal number words (Koechlin, Naccache, Block & Dehaene, 1999) and as non-symbolic numerosities (Buckley & Gilman, 1974). Several models of numerical representation have been proposed to explain the distance and size effect. These models can be categorized in two main coding models: place coding and summation coding. 1.2 Place coding representation The distance and size effect were combined by Dehaene (1992) in a very influential model of numerical representation. Dehaene proposed analogue magnitudes to be represented on a mental number line, with small numbers represented at the left and large numbers at the right end of the line (Dehaene, Bossini & Giraux, 1993). It is important to stress here that the ‘number line’ is merely a metaphor for an ordered group of input neurons, where each neuron activates for a particular number. Magnitudes are thus represented in the brain by activating the corresponding neuron, or as a pattern of activation over the group of neurons which constitute the number line.