What makes a shape “baba”? The shape features prioritized in sound–shape correspondence change with development

Abstract

Crossmodal sound–shape correspondence, the association of abstract shapes and nonsense words (e.g., “bouba-kiki” effect), is seen across cultures and languages. Recent research suggests that the sensitivity for such sound–shape pairings might increase with development. Here we examined one possible mechanism underlying developmental changes in sensitivity to sound–shape correspondences—if and how shape features, such as contour spikiness and the number and size of protrusions, might be weighted differently by children and adults. In Experiment 1, we asked participants to choose which of two nonsense words matched a given visual shape while manipulating contour spikiness and number and size of protrusions independently. We found that adults associated /i/ sounds with shapes having spiky contours and 3 small protrusions. Of these shape features, contour spikiness showed the strongest association. Whereas 9- to 11-year-olds showed adult-like responses, 6- to 8-year-olds prioritized protrusion number, not contour spikiness. Importantly, in Experiment 2, where contour spikiness was highlighted by presenting round and spikey shapes side by side, 6- to 8-year-olds could make associations based on contour spikiness. Our findings suggest that 6- to 8-year-olds prioritize different features of a shape when making sound–shape correspondence compared with adults. Interestingly, these shape-processing biases can be altered by context such that children can resemble adults when the relevant shape features are highlighted. Our results suggest that biases in visual shape processing and the ability to extract contextual information might be additional factors explaining developmental changes in sensitivity toward sound–shape correspondences. These changing developmental biases highlight the contribution of perceptual processing styles in crossmodal correspondence.