Abstract

ABSTRACT : The number of studies investigating the neural mechanisms underlying non-symbolic number processes increased drastically in the past years. On the basis of these studies it has been concluded that we are equipped with an approximate number system, a mechanism that allows us to process number independent of its continuous visual properties (e.g. diameter, contour length, etc). Most of these studies however focused on the higher order stages of numerosity processing while little attention was devoted to the preceding visual processes. This is unexpected, as non-symbolic number is defined by its visual characteristics. When these are the same between two sets of items, number is also the same. Similarly, when number changes, its visual characteristics change accordingly. In daily life, number and visual cues are highly correlated (e.g. when you add more apples to a bag of apples, continuous visual properties will increase). This relation is unlikely to be violated. Consequently, it would be very efficient to rely on these continuous visual variables to judge number. In my talk I will pitch the theory of the approximate number system against the theory of visual properties. Using data from behavioral and EEG studies I will show caveats in the designs used to study number processing and argue that it is unlikely that we are equipped with a mechanism that allows us to extract number from a visual scene independent of its visual properties. Indeed, I will propose that non-symbolic number is the result of the integration of different visual variables.