The development of crossmodal intuitive physics and its relation to crossmodal correspondences

Poster Presentation 56.468: Tuesday, May 19, 2026, 2:45 – 6:45 pm, Pavilion
Session: Multisensory Processing: Audiovisual

Mahan Abdollahinarenjbon1, Lily Charbonneau1, Vivian Ciaramitaro1; 1University of Massachusetts Boston

The bouba-kiki effect is a type of crossmodal correspondence, in which pseudowords, like "bouba", tend to be matched with round shapes, and "kiki" with spiky shapes. Evidence for such avBK associations is equivocal in children under 3 (Fort et.al., 2017), and even 6-8-year-olds show weaker associations than older children (Chow et.al., 2019). Auditory dimensions of spectral content and temporal continuity can correctly categorize BK pseudowords, as well as categorize the sounds expected from rolling objects based on their shape (round vs. spiky; Fort & Schwartz, 2022). Everyday perception and action rely on such crossmodal intuitive physics (avIP): the ability to infer a scene’s structure and predict how objects will move and what sounds will result. We hypothesized that the same cognitive mechanism allowing us to anticipate the sound a rolling ball would make based on its shape may also bias our matching of more abstract audio-visual correspondences. We tested if avIP strength correlated with avBK strength. Each participant (6-8-, 9-11-, and 18-35-year-olds; N=198 total) completed 16 trials (2shape pairs*4sounds*2repeats) of avBK and avIP, with block order counter-balanced. On each trial, two shapes appeared, round and spiky, and a sound was played. avBK presented abstract round and spiky shapes and pseudowords, and avIP presented round and spiky balls about to roll (implied visual motion) and rolling sounds. We quantified matching strength as the proportion of congruent choices. All participants showed above-chance congruent matches for avIP and avBK, with effects increasing with age. Overall, children with stronger avBK matches showed stronger avIP matches (Spearman ρ=.22,p=.012), an association driven by 9–11-year-olds (ρ=.28,p=.031). Thus, we showed that children as young as 6 years of age have a crossmodal intuitive physics capacity, which strengthens with development. This capacity may help scaffold how abstract shapes and nonsense pseudowords are associated, an association important for word-learning.