Abstract
Spatial information uniquely influences working memory. In many cases, task-irrelevant spatial information influences memory while other kinds of task-irrelevant information (e.g., color, auditory tones) do not. But how does this spatial influence occur? What spatial cues matter? Here, we ask whether information bound to discrete objects is better remembered than information bound to those same locations but over a continuous space. In our baseline task, participants played a memory game in which they had to remember sequences of shapes appearing every other second. Sequences were made up of 5-7 shapes (three unique ones: circle, diamond, and pentagon) that could appear in one of four locations (four separate quadrants) and in one of four colors. Participants were instructed to remember what shapes they saw and in what order. On some trials, spatial information was structured so that repeated shapes were always in the same location (but color was randomized). On other trials, color information was structured the same way (but location randomized). Critically, all four locations were marked by discrete squares in which the shapes could appear. In this task, participants better remembered the shape sequences on space-structured trials. In a follow-up experiment, we simply removed the discrete squares marking where shapes could appear. In this case, participants did not exhibit the same memory boost in the space-structured trials. We also tested whether this effect depends on encoding or retrieval by initially presenting the discrete black squares but removing them at test. Here, there was a marginal effect of spatial structure. Combined, these results suggest that binding information to discrete objects in specific locations as opposed to continuous space — even when that spatial information is task irrelevant — benefits memory. Working memory, like attention, may therefore rely in part on object-based processes.