Spatial Memory in Animals

Abstract

We are living in an epoch of fundamental conceptual changes. This statement applies not only to world politics, military doctrines, and economic models but also to scientific concepts. In neurosciences this fundamental change can be best illustrated by the rapidly increasing influence of cognitive theories of behavior. Whereas the founding fathers of modern physiological psychology—Pavlov (1927) and Watson (1919)—purged their theories of such terms as memory, goals, expectations, and the like, present-day neuroscience has returned respectability to the ideas espoused by Beritashvili (1971) and Tolman (1932), who endowed animals with the capability to form a neural representation of the outside world and to use this representation for the control of behavior. The change of attitude was not brought about by persuasive arguments but rather by the different philosophical context in which this research is pursued. The antimentalistic position prevailing at the turn of the century served the noble aim of achieving a strictly objective explanation of behavior, leaving no room for speculative constructs based on introspection. The situation is quite different now when computer sciences modeling the sensorimotor control of robots show that the flexible solution of the task requires the application of such cognitive principles as setting the goal of action and selecting the strategy for achieving it. What appeared to be a misleading speculation a century ago is now a well-justified assumption that has inspired much of the current research into the neural mechanisms of learning and memory.