Introduction

Many of the external events that we experience in everyday life produce multiple signals, each stimulating different sensory modalities. The multimodal nature of these signals can enrich our perceptual experience, but it also raises the issue of how coherent neuronal representations are formed for such different signals in the brain. Although the origin of a set of multisensory signals might correspond to a single external event, the initial processing of signals to different senses will occur in anatomically distant areas. This raises the question of how the unity of the original event is reconstructed in the brain via multisensory integration. In this chapter we review several recent findings from human functional imaging studies that address these issues, with a focus on spatial multisensory representations and attentional selection of spatial locations. We show that several neural mechanisms might underlie the integration of spatial representations across sensory modalities. These mechanisms include feed-forward convergence of signals from early sensory-specific areas to higher-level brain regions, forming multimodal spatial representations, and also cross-modal spatial influences on relatively early (apparently unimodal) sensory-specific areas, which may involve feedback influences from multimodal attentional control structures.