Cross–modal links in spatial attention

Jon Driver, Charles Spence


A great deal is now known about the effects of spatial attention within individual sensory modalities, especially for vision and audition. However, there has been little previous study of possible crossmodal links in attention. Here, we review recent findings from our own experiments on this topic, which reveal extensive spatial links between the modalities. An irrelevant but salient event presented within touch, audition, or vision, can attract covert spatial attention in the other modalities (with the one exception that visual events do not attract auditory attention when saccades are prevented). By shifting receptors in one modality relative to another, the spatial coordinates of these crossmodal interactions can be examined. For instance, when a hand is placed in a new position, stimulation of it now draws visual attention to a correspondingly different location, although some aspects of attention do not spatially remap in this way. Crossmodal links are also evident in voluntary shifts of attention. When a person strongly expects a target in one modality (e.g. audition) to appear in a particular location, their judgements improve at that location not only for the expected modality but also for other modalities (e.g. vision), even if events in the latter modality are somewhat more likely elsewhere. Finally, some of our experiments suggest that information from different sensory modalities may be integrated preattentively, to produce the multimodal internal spatial representations in which attention can be directed. Such preattentive crossmodal integration can, in some cases, produce helpful illusions that increase the efficiency of selective attention in complex scenes.

Royal Society Login

List of OpenAthens registered sites, including contact details.