The neural foundations of an auditory-visual illusion: spatial factors and application to childhood development

Most animals, including humans, possess multiple sense organs through which information about the environment is gleaned. The brain’s ability to integrate the information gathered from multiple sense modalities is fundamental to success in the environment. However, until fairly recently, the processing of sensory information in the brain was assumed to follow separate parallel streams, being progressively elaborated in these separate streams before being finally united in high level association cortices, such as those in the frontal or parietal lobes. As more sophisticated theories and analysis techniques have emerged, this view has given way to a new understanding that information is also transferred between sensory pathways at very early stages – at the level of the primary sensory cortices and possibly below. The three experiments presented in this thesis used the flash-beep illusion (Shams et al., 2000) as an index of multi-sensory processes in the brain – firstly to show that the illusion was robust to significant spatial separation of the auditory and visual parts of the stimulus, and secondly to show that congruent multi-modal stimuli initiated a sequence of long-lasting effects on subsequent uni-sensory processing that were localised to the parietal and occipital lobes. Lastly, it was shown that a group of children from 8-14 years of age reported significantly more illusions than adults, adding further weight to the growing body of evidence suggesting that many multi-sensory processes are slow to develop. Overall, the experiments reported in this thesis highlight the involvement of high-level cortical association areas as well as direct cortico-cortical connections between primary auditory and visual areas in the generation of the flash-beep illusion. It is proposed that both are required in order to generate the perception of an illusory flash: the multi-modal flash/beep stimulus initiates a sequence of activity in the primary sensory areas which subsequently leads to feedback from parietal areas, thus priming visual cortex. Direct, spatially insensitive connections from primary auditory to visual areas subserving the visual periphery may allow auditory activity generated by the second beep to trigger a visual percept. The results increase our understanding that multi-sensory processes are generally the result of complex interactions between sub-cortical nuclei and primary sensory areas, guided by top-down control from high-level association regions which are slow to develop during childhood.