The cocktail party phenomenon revisited: attention and memory in the classic selective listening procedure of Cherry (1953).

Though E. C. Cherry (1953) examined the recall of information from an irrelevant spoken channel in selective listening, the relationship between attention and subsequent recall still has not been examined adequately. It was examined here in 4 experiments, 3 of which were designed to identify conditions under which some participants, but not others, would notice a change from forward to backward speech. Only participants who shifted attention toward the irrelevant channel during the backward speech later recalled hearing it. In those whose attention shifted, shadowing errors peaked dramatically about 15 s after the change. There was no evidence of direct or indirect memory for phrases presented in the irrelevant channel. The results contradict models of attention stating that listeners process task-irrelevant information extensively without diverting resources used in shadowing.