Simultaneous representation of location with respect to landmarks and boundaries

Undergraduate Just-In-Time Abstract

Poster Presentation 56.349: Tuesday, May 19, 2026, 2:45 – 6:45 pm, Banyan Breezeway
Session: Undergraduate Just-In-Time 3

Kyra Warmuth1, Sami Yousif; 1The Ohio State University, 2The Ohio State University

Suppose you return to a place you have not visited in many years and try to navigate to a favorite restaurant. You do not remember its exact location, but you are confident you can find it. Yet once you begin, you realize that much has changed: restaurants have closed or opened, new parks have been established, and street names differ. The landmarks you once relied on are no longer reliable. How do you find your way? Some theories propose that cognitive maps integrate multiple sources of spatial information into a kind of ‘cognitive collage’, so that, when one source of information fails, there are other available cues. Empirical evidence for such ideas remains limited, however. Here, we provide some. Participants completed a spatial memory task in which they remembered the location of a dot relative to both a landmark and a boundary. In Experiment 1, both cues were present during encoding, but at retrieval participants saw either the landmark, the boundary, or both. Error patterns differed across conditions, with the lowest overall error when both cues were available, perhaps suggesting that locations were encoded relative to both. We also found different characteristic patterns of errors in each condition (e.g., participants were biased toward the landmark when it was present at retrieval, but not otherwise). In Experiments 2 and 3, the task was identical except that retrieval trials were weighted toward the landmark condition (Experiment 2) or the boundary condition (Experiment 3) on 80% of trials (with 10% for each alternative). Memory improved for the high-probability cue and worsened for others, indicating that encoding flexibly prioritizes the most reliable environmental cues. These results suggest that spatial information may be encoded with respect to multiple reference frames simultaneously, supporting the idea that redundant representations can benefit memory when reliable cues are disrupted.