Having learnt that hearts are red, and bananas yellow, we are more ready to perceive heart shapes as red and banana shapes as yellow (Delk and Fillenbaum, 1965; Hansen et al. 2006; Olkkonen at al., 2008). These effects on perception are difficult to measure and to explain. They vary from one object to another: in tested conditions, hearts and apples will look more red, but squares reflecting the same wavelength won’t. Does this “object-sensitivity” of colour perception suggest that the effect is shaped from outside, by higher cognitive processes where the information about shapes and colours are stored together? I discuss these cases and other examples of object-sensitivity in perception. Contrary to Siegel or McPherson, I don’t think that these cases give us good reasons to say that perception is cognitively penetrable. I offer an alternative account in terms of semantic cross-modal effects, which helps us understand the nature of multi-sensory integration.

Events occurring in one sensory modality can often affect how we experience sensations in a different sensory modality. For instance, the sound of a mosquito flying near us seemingly enhances sensitivity to touch on our skin; hearing a suspicious sound may heighten our ability to detect an inconspicuous visual event. I will describe a series of experiments that examines the effects of audition on touch and visual perception, as well as some of the neural substrates underlying them. In addition to describing psychophysical experiments that have revealed systematic ways in which one sensory modality influences another, I will describe a set of experiments using diffusion tensor imaging (DTI) and functional magnetic resonance imaging (fMRI) that were conducted on a patient who has a unique form of acquired auditory-tactile synesthesia. Together, these studies illustrate some of the dynamic interactions between our different sensory systems and also highlight the plastic properties of sensory processing in the human brain.