Abstract
Previous research has suggested that monkeys and humans can process more than one novel pictured scene at a time when the task is to detect a target. To evaluate human ability to perceive and remember two unfamiliar scenes at the same time (colored photographs), we presented either one or two pictures in each frame of a rapid serial visual presentation (RSVP) sequence. Pictures were presented in two positions above and below a fixation point; when only one picture was presented, the other position contained a visual mask. In each trial the sequence of 6 frames included 4 pictures presented alone and 4 presented in pairs. Pairs of masks preceded and followed the sequence. Presentation duration varied between 160 and 720 ms/frame. A yes-no recognition test immediately followed each sequence. Recognition memory for each of two pictures presented at the same time was less accurate than for pictures presented alone, suggesting a cost for simultaneously processed items. As expected, performance improved as presentation duration increased (Potter & Levy, 1969) and performance declined over serial position in the recognition test (Potter, Staub, Rado, & O'Connor, 2002). These memory results are compared with performance in a detection task in which one picture in the sequence matches a named target (specified before each trial). We discuss parallels and differences between semantic detection and recognition memory, given single or paired presentations.