Cortical sensory systems often activate in parallel, even when stimulation is experienced through a single sensory modality [1][2][3]. Critically, the functional relationship between co-activated cortical systems is unclear: Co-activations may reflect the interactive coupling between information-linked cortical systems or merely parallel but independent sensory processing. Here, we report causal evidence consistent with the hypothesis that human somatosensory cortex (S1), which co-activates with auditory cortex during the processing of vibrations and textures [4][5][6][7][8][9], interactively couples to cortical systems that support auditory perception. In a series of behavioural experiments, we used transcranial magnetic stimulation (TMS) to probe interactions between the somatosensory and auditory perceptual systems as we manipulated attention state.Acute manipulation of S1 activity using TMS impairs auditory frequency perception when subjects simultaneously attend to auditory and tactile frequency, but not when attention is directed to audition alone. Auditory frequency perception is unaffected by TMS over visual cortex thus confirming the privileged interactions between the somatosensory and auditory systems in temporal frequency processing [10][11][12][13]. Our results provide a key demonstration that selective attention can modulate the functional properties of cortical systems thought to support specific sensory modalities. The gating of crossmodal coupling by selective attention may critically support multisensory interactions and feature-specific perception.
Results and DiscussionAttending to a stimulus feature that can be redundantly signalled through different senses can activate multiple cortical sensory systems. For example, the auditory cortex can be activated by visual stimulation even in the absence of sounds, particularly when attending to stimulus features that are associated with sounds [14,15]. Such coactivation of sensory systems is thought to reflect a number of attention-based operations including the selection, refinement, and binding of sensory representations [16]. Critically, the functional relationship between co-activated cortical systems remains ambiguous. Co-activation could reflect an interactive coupling between the cortical sensory regions. Alternatively, co-activation could merely reflect the parallel, but independent activation of sensory systems. Distinguishing between these possibilities would provide critical insight into the functional role of distributed activity over cortical systems that are traditionally thought to be dedicated to unisensory processing.