Abstract
Children can count and track not only material objects but also non-physical entities, holes (Giralt & Bloom, 2000; Kibbe & Leslie, 2016). Using the Bouba/Kiki effect (in which both adults and children associate rounded shapes with the word “Bouba” and pointed shapes with “Kiki”) as an implicit, direct measure of shape perception, we examined the development of the shape perception of holes in children. Adopting this methodological approach, our previous work (Kim, 2020) tested adults’ sound-shape association for donut-shaped cutouts, one with a star-like hole (surrounded by rounded material bulges connected with sharp concavities; Figure 1) and the other with a flower-like hole (surrounded by pointy edges connected with rounded concavities). In both stimuli, the shapes of the interior (hole) and the exterior (figure) regions gave rise to opposite impressions (one rounded and the other pointed), and we found that sound-shape association was based on the shape of a hole rather than that of a surrounding figure, suggesting that holes have associated shapes despite their ground status in border ownership. In the current study, we tested 66 2-8-year-old children (M=62.8 months; SD=19.9 months). Children received one trial in which they were shown two cutouts and decided which one was either “Bouba” or “Kiki” (counterbalanced across subjects). We found that selecting the sound-shape-congruent hole was significantly correlated with children’s age in months (Spearman’s r=.357, p=.003). Older children (above the mean age) selected the sound-shape-congruent hole at rates significantly above chance (27/34 (79.4%), binomial test p<.001) similar to adults (Kim, 2020), while younger children selected the congruent and incongruent holes at roughly equivalent rates (17/32 (53.1%), binomial test p=.86; younger versus older: Fisher’s exact test p=.04). These results suggest that the shape representations of holes require global shape processing which undergoes development between 2 and 8 years.