hide
Free keywords:
-
Abstract:
Spatial theories that are postulating a hierarchical representation of our environment assume that subsets of spatial units (e.g., multiple streets in a city) are grouped together and subsumed under a unit on a higher level of hierarchy (e.g., a region such as the city centre). However, circumstances under which grouping into higher level memory units occur as well as the nature of those units is still
unexplored. In our study we had subjects learn a virtual environment consisting of two apparent regions, each region containing 4 target objects that had to be memorized. Besides visual cues (colour blue or red) additional factors were held constant within one region while being maximally different from the other region: semantic similarity (object categories), distance between targets (longest path at regional transition point), complexity of turning angles (most complex 45° at transition point), and spatio-temporal
contingency (subsequent learning of the two halves of the environment). We thus triggered mental separation of blue and red region. Subsequent performance in a pointing task revealed increased pointing latency with increasing corridor distance to the target, indicating that local corridors served as memory units on the lowest level of the hierarchy. Further, an effect of region was found: people were faster pointing to targets located in the same region as they currently were located themselves, compared to pointing to targets in the other region. Additionally, within-region pointing revealed a reference frame alignment with both local corridors and the orientation of the region. Thus, the corridors of both regions established obliquely aligned reference frames. Both effects give way to the
interpretation that the single corridors were not integrated into one global representation subsuming all eight corridors, but rather, that two distinct representation, one for each region, might have been formed on the second level of hierarchy.
