UC San Diego

In everyday perception, dynamic objects move and collide within physical environments, producing expected sounds. In this dissertation, I suggest that perceptual phenomena, like a bouncing ball, may offer mechanistic insights into: 1) how the brain anticipates sound via the integration of dynamic visual cues, 2) clinical conditions who show differences in the ability to anticipate, and 3) the developmental emergence of skills used to anticipate sound. In a series of experiments, I presented neurotypical and autistic adults, and neurotypical infants a dynamic visual object that collides with a physical barrier, eliciting a sound at the point of expected collision (AV-synchronous), or unexpectedly before collision (AV-asynchronous). In chapter one, I recorded event-related potentials (ERPs) from neurotypical adults who were exposed to sounds that either synchronized with visual collision or occurred asynchronously before collision. I also included conditions where the object was occluded during synchronous collision, or when sound was presented without dynamic visual cues. I found that synchronous and occluded collision sounds elicited an attenuated auditory response relative to asynchronous or audio-only sounds. These results suggest that dynamic visual stimuli can help generate expectations about the timing of sound, which then facilitates the processing of auditory information that matches these expectations. In chapter two, I replicated the same methods as in chapter one, but in a sample of autistic adults. Here, I observed greater amplitudes toward asynchrony in autism relative to neurotypicals, while no group differences toward fully visible or occluded synchrony emerged. These results suggest that neural responses to prediction errors are affected in autism, and not the integration of top-down expectations. In chapter three, I modified these methods for use in neurotypical infants to show that 4-to-5-month-olds look longer to bounce sounds that violate temporal expectations of when a bounce sound should occur. These studies highlight the presence of neural mechanisms sensitive to predictable sound, which appear to be different in clinical populations like autism. Moreover, infants are sensitive to collision sounds, demonstrating that these perceptual skills are available early in life. Collectively, these methods could be further leveraged to understand the emergence of neurodevelopmental conditions like autism.