Supramodal Representations in Melodic Perception

Music is highly relational and in this manner shares much in common with other characteristically human behavior. While this may suggest that the processes used in music perception could be domain general, the nature and flexibility of these representations remain less understood. If the underlying representations and manipulations required for perceiving music overlap with those in other cognitive domains, it should be fairly easy to map such representations across domains. This hypothesis was tested and supported using a novel experiment with melodic stimuli in the auditory modality and analogous visual sequential stimuli (Gabor sequences) in the visual modality. Testing for transfer across the two modalities and for the two types of representations (contour and intervallic) was done through four counterbalanced conditions. Cross-modal mapping was successful in three out of the four conditions, implying general flexibility of representational transfer. Implications for representational flexibility, sequential learning and future studies are discussed.