Spatial memory: the role of egocentric and allocentric frames of reference

A fundamental requirement of everyday life is that of encoding and remembering successfully the locations of objects, landmarks or buildings in space. This function is achieved by structuring spatial information in systems of coordinates. There are many ways to classify spatial reference systems (Paillard, 1991), but a useful one, for the purposes of understanding human spatial memory, divides them into two main categories. Egocentric frames of reference specify location and orientation with respect to the organism, and include eye, head, and body coordinates. Allocentric frames of reference specify location and orientation with respect to elements and features of the environment independently of the viewer’s position. Given the primary role that egocentric and allocentric processing systems play in perception and action (e.g. Milner & Goodale, 1995, 2008; but see Schenk, 2006), they can be thought of as specialized cognitive mechanisms essential in performing and guiding spatial motor actions. Egocentric representations have a special relevance in controlling movement in surrounding space such as avoiding obstacles or reaching objects. All these actions are performed in near/peripersonal space, i.e. the space within arm-reaching distance and require fine-grained metric information. On the other hand, allocentric representations have an important role in recognizing objects, scenes and planning future movements (i.e. the space outside arm-reaching distance) (e.g. Milner & Goodale, 2008). Spatial memory is intrinsically linked to frames of reference as it is not possible to store spatial information without structuring it according to specific frames. In this chapter we review evidence from behavioral and neurofunctional data about models of spatial memory, automatic and effortful encoding of spatial information, and spatial memory in elderly people. Furthermore, two experiments are reported that investigate memories for egocentric and allocentric frames of reference by means of a task requiring judgments of relative distance. Experiment 1 investigates the influence of near/far spaces on spatial frames of reference. Experiments 2 controls the influence of possible artifactual effects. Overall, the pattern of data shows that the egocentric processing is more accurate and faster than the allocentric one. The results are discussed in relation to models of spatial memory that emphasize the importance of egocentric experience (Kosslyn, 1994; Millar, 1994).