Is a change as good with a rest? Task-dependent effects of inter-trial contingency on concurrent sound segregation

The effects of local and global context on concurrent sound segregation were examined using a mistuned harmonic paradigm. Presented with the same sounds over three conditions, participants were asked to either categorise the incoming complex according to pitch or harmonicity (active listening) or watch a muted subtitled movie of their choice (passive listening). Global context was defined according to task instruction, while local context was defined relative to the repetition or change of harmonicity across trials. An interaction between global and local context in both behavioural and event-related potential measures was found, with participants showing sensitivity to change and repetition in harmonicity when it was task-relevant, in addition to enhanced P3 amplitude for cases where a tuned stimulus changed into a mistuned stimulus over consecutive trials. The data are discussed in terms of current models of concurrent sound perception and how top-down processes might influence these mechanisms.