Object-array structure, frames of reference, and retrieval of spatial knowledge.

Assessment of the ability of people, without vision, to locate the positions of objects from imagined points of observation that are related to their actual position by rotational or translational components indicates that in the case of regularly structured object arrays, interobject relations are directly retrieved for the translation task, but for the rotation task, retrieval occurs by means of a body-centered coordinate system.

Abstract

Experiments are reported that assessed the ability of people, without vision, to locate the positions of objects from imagined points of observation that are related to their actual position by rotational or translational components. Theoretical issues addressed were whether spatial relations stored in an object-to-object system are directly retrieved or whether retrieval is mediated by a body-centered coordinate system, and whether body-centered access involves a process of imaging updating of self-position. The results, with those of Rieser (1989), indicate that in the case of regularly structured object arrays, interobject relations are directly retrieved for the translation task, but for the rotation task, retrieval occurs by means of a body-centered coordinate system, requiring imagined body rotation. For irregularly structured arrays, access of interobject spatial structure occurs by means of a body-centered coordinate system for both translation and rotation tasks, requiring imagined body translation or rotation. Array regularity affected retrieval of spatial structure in terms of global shape of interobject relations and local object position within global shape.