Abstract
The hallmark of categorical perception (CP) is better discrimination around a category boundary than within the category. Özgen and Davies (2002) found that CP could be induced around a newly learned category boundary as measured by a successive same-different task, and argued that category learning had changed perception. However, although changes in same-different performance could be based on perceptual change, they could also be based, on memory or decisional effects. Here we replicate Özgen and Davies’s category training, but test for CP using three different measures: thresholds, target detection, and ERP during an oddball task. In the threshold task, chromatic JNDs were measured across the new boundary, and at the centre of the new categories. In the target detection task, the location of a coloured target against a differently coloured background had to be detected and RTs were measured. The target-background pair was either within- or across-category. In the oddball task, a frequent (standard) colour was presented repeatedly interspersed, with a low frequency different colour (oddball). The standard and oddball were either from the same (newly learned) category or from the two new categories. In the threshold task, there were no indications of an acquired category effect. In contrast, in the target detection task, cross-category targets were detected faster than within-category targets, and this effect was confined to the right-visual-field (left hemisphere). In the oddball task, there no indications of early category effects in N1 and P1 components, but there were strong category effects in the P3 (late) component. The results suggest that while learning new colour categories for a few hours induces category effects, the locus of the effect is probably on post perceptual processes, rather than in perception per se.