Spatial Representations in Mammals as Topological Networks

The notion that mammals may be able to form spatial representations of their environment has recently received considerable attention. At least two reasons for this interest can be identified. First, the impressive ability of animals to compute optimal trajectories through space suggests that there must be some sophisticated mechanisms that allow for the integration and manipulation of information about the environment. The analysis of such processes may make it possible to understand how complex adaptive behaviour can emerge from simpler processes, presumably without the need for analytic reasoning. Second, the discovery by O’Keefe and Dostrovsky (1971; see also O’Keefe & Nadel 1978) of a relation between the activity of specific neurones in the rat brain and the spatial behaviour displayed by the animal has motivated a good deal of subsequent research, and has resulted in a number of tentative solutions to the problem of the implementation of complex spatial capacities in neural circuits. In this chapter, we will examine some of the results recently obtained at the behavioural and neural levels of analysis.