Updating an object's orientation and location during nonvisual navigation: a comparison between two processing modes.

In the present study, we compared the effects of two processing modes on the updating of the location and orientation of a previously viewed object in space during a guided walk without vision. In Experiment 1, in order to measure the error for initial perception of object's orientation, 12 subjects rotated a miniature model until it matched the memorized orientation of its counterpart object in space. In Experiment 2, they attempted either to keep track of the object continuously (in the object-centered [OC] task) or to estimate the object's perspective only at the terminal vantage point given the trajectory they walked (in the trajectory-centered [TC] task). Subjects indicated the location of the object by facing it, and then rotated the model in order to indicate its orientation from the new vantage point. Results showed that, with respect to the TC mode, the OC mode induced a slow-down of the subjects' self-paced locomotion velocity for both linear and angular movements, and a decrease of the latencies as well as smaller absolute errors for the orientation-of-the-object response. Mean signed errors on object's orientation were equivalent for both processing modes, suggesting that the latter induced different allocations of processing resources on a common representation of space updated by "path integration."