Mapping visual spatial prototypes: Multiple reference frames shape visual memory

Categories provide a fundamental source of information used to structure our perception of the world. For example, when people reproduce the remembered location of a dot in a circle, they implicitly impose vertical and horizontal axes onto the circle, and responses are biased towards the center of each of the resulting quadrants. Such results reveal the existence of spatial prototypes, which function as Bayesian priors and which are integrated with actual memory traces. Spatial prototypes have been extensively investigated and described in previous studies, but it remains unclear what type of information is used to create spatial categories. We developed a new approach that allowed to 'image' patterns of spatial bias in detail, and map the internal representational structure of objects and space. Previous studies, using circular shapes suggested that boundaries are established based on a viewer-based frame of reference, therefore using cues extrinsic to the object. Given that a circle has radial symmetry, the axes imposed cannot come from the shape itself. Here we investigated if the same applies for shapes with clearly-defined symmetry axes and thus intrinsic frames of reference. Using rotated shapes (squares and rectangles), where extrinsic and intrinsic cues are dissociated, we observed flexible usage of multiple reference frames. Furthermore, in certain contexts, participants relied mostly on cues intrinsic to the shape itself. These results show that humans divide visual space as a function of multiple reference frames, in a flexible, and context dependent manner.

Temporal and spectral profiles of conflict processing among multiple frames of reference

Individuals rely on various frames of reference (FORs), such as an egocentric FOR (EFOR) and intrinsic FOR (IFOR), to represent spatial information. Previous behavioral studies have shown different IFOR-IFOR (II) and EFOR-IFOR (EI) conflict effects and an effect of their interaction. However, the neural mechanism of conflict processing between two FOR-based conflicts is unclear. In the current ERP study, two FOR-based conflicts were manipulated using a two-cannon task to elucidate common and distinct brain mechanisms that underlie FOR-based conflict processing. The behavioral results showed that both conflicts exhibited longer reaction times and larger error rates in the II (180° cannon angle) and EI (target cannon pointed down) incongruent conditions than in the II (0° cannon angle) and EI (target cannon pointed up) congruent conditions and that an interaction existed between the two conflicts. The ERP results indicated that, for both conflicts, more negative N2 amplitudes and less positive P3 amplitudes occurred in the incongruent conditions than in the congruent conditions, and the interactions between the two conflicts during later P3 amplitudes were significant. Time-frequency analysis further indicated that, in the early time window, the II conflict and the EI conflict specifically modulated power in the theta bands and beta bands, respectively. In contrast, in the later time window, both conflicts modulated power in the alpha and beta bands. In summary, our findings provide insights into the potential existence of two specific early conflict monitoring systems and a general late executive control system for FOR-based conflicts.

Conflict processing among multiple frames of reference

People rely on multiple frames of reference (FORs) to represent and update spatial relationships of different objects in a complex environment. According to the "Frame of Reference-based Map of Salience" theory (FORMS), FORs with high salience are processed in priority, and human performance is determined by the interaction of all relevant FOR-based representations. We conducted a modified two-cannon task manipulating conflicts among FORs (e.g., egocentric [FOR-EFOR], intrinsic [FOR-IFOR]) in order to explore how FORs interact with each other. We found that participants responded slower when two IFORs were incongruent than when they were congruent. There was also an interaction between these two conflicts. Moreover, the effect size of conflict between two IFORs was much larger than that between an IFOR and an EFOR. These results suggest that although the IFOR and EFOR may be coded by different spatial representations, they rely on a common processing mechanism and compete with each other. The findings from the current study provide support for the FORMS theory.

Spatial reasoning with multiple intrinsic frames of reference

Establishing and updating spatial relationships between objects in the environment is vital to maintaining situation awareness and supporting many socio-spatial tasks. In a complex environment, people often need to utilize multiple reference systems that are intrinsic to different objects (intrinsic frame of reference, IFOR), but these IFORs may conflict with each other in one or more ways. Current spatial cognition theories do not adequately address how people handle multi-IFOR reasoning problems. Two experiments manipulated relative orientations of two task-relevant objects with intrinsic axes of orientation as well as their relative task salience. Response times (RTs) decreased with increasing salience of the targeted IFOR. In addition, RTs increased as a consequence of intrinsic orientation conflict, but not by amount of orientation difference. The results suggest that people encounter difficulties when they have to process two conflicting IFOR representations, and that they seem to prioritize processing of each IFOR by salience.