The effect of stimulus context and attention on Auditory Scene Analysis
MetadataShow full item record
When encountering a noisy scene such as a crowded street corner we are faced with the task of organizing the incoming sounds into individual sound streams. It takes time to parse the components of this mixture to their individual sound sources. The goal of this dissertation is to examine the factors that impact the timing of this segregation process to better understand how humans navigate complex acoustic landscapes. We tested the overall hypothesis that stimulus context, attention and the organization of auditory input can impact the speed of stream formation. The first experiment examined the impact of a dynamic versus stable context on segregation speed. The second experiment followed by manipulating task goals to test the hypothesis that attention allows for faster segregation irrespective of the dynamics of the surroundings. Lastly, in the third experiment we examined whether the auditory system maintains a representation of segregation across silent gaps in the absence of auditory input. Human auditory event-related brain potentials (ERPs) were used in order to measure sound processing in both passive and active listening situations. The paradigm was designed such that an ERP component of change detection indexed the organization of the sounds. Our results indicated that segregation occurs more quickly in stable environments (Experiment 1) and that attention can overcome the dynamics of the scene to speed up stream formation (Experiment 2). Additionally, silence modulates the degree to which streams are maintained (Experiment 3). Taken together, these results support the maintenance of a consistent neural representation of the acoustic surroundings despite outside perturbations. In a realistic context, this suggests that when encountering a scene such as the crowded street corner the auditory system separates sound streams to their individual component parts, and works to preserve this representation until sufficient information suggests a complete reanalysis is necessary.