Abstract
Crossmodal correspondences highlight naturally occurring associations across our senses. In the bouba/kiki effect, abstract shapes are associated with nonsense words: round shapes matched to /bouba/sounds and spiky shapes to /kiki/ sounds. Such associations are ubiquitous, found across cultures, languages, and development. We used steady state evoked potentials (SSEPs) to quantify how neural responses to a given visual feature depend on how the auditory feature is associated with the visual feature. We hypothesized enhanced neural responses to the same attended shape (e.g. spiky) for congruent (e.g. /ki/) vs incongruent (e.g. /ba/) sounds. Twenty-four subjects viewed one round and one spikey shape, each half-shape presented in a given visual hemifield. Shapes flickered at different frequencies (5.45, 7.5Hz) and were presented under different auditory conditions: no sound, /ba/ or /ki/ sound (3 Hz). Participants attended fixation to detect a color change, attended a shape to detect a border thickening, or attended a sound to detect a volume change. Signal-to-noise ratios for visual responses (electrodes Oz, O1, and O2) were measured at the fundamental frequencies of stimulus presentation to quantify visual shape processing as a function of attention and crossmodal congruency. We found: (1) enhanced neuronal responses for the attended shape when the unattended sound was congruent vs incongruent, (2) weaker enhancement when the shape was attended versus unattended to attend fixation (sound congruent to attended shape in both conditions) and (3) no difference for an unattended shape if a congruent vs incongruent sound was attended. This work extends studies investigating mechanisms of attention to features across our different senses for naturally occurring but abstract associations.