The evolution of cognition—a hypothesis

Abstract

Behavior may be controlled by reactive systems. In a reactive system the motor output is exclusively driven by actual sensory input. An alternative solution to control behavior is given by "cognitive" systems capable of planning ahead. To this end the system has to be equipped with some kind of internal world model. A sensible basis of an internal world model might be a model of the system's own body. I show that a reactive system with the ability to control a body of complex geometry requires only a slight reorganization to form a cognitive system. This implies that the assumption that the evolution of cognitive properties requires the introduction of new, additional modules, namely internal world models, is not justified. Rather, these modules may already have existed before the system obtained cognitive properties. Furthermore, I discuss whether the occurrence of such world models may lead to systems having internal perspective.

Compact internal representation of dynamic situations: neural network implementing the causality principle

“…Recently the concept of internal representation (IR) of static and dynamic situations has been strongly supported by experimental and theoretical studies. It has been shown that IR indeed offers new perspectives for the decision making and it is said to be a prerequisite for a cognitive behavior (Cruse 2003). A big challenge in the IR research is to understand how the brain makes a compact effective description of complex, timeevolving situations, in such a way that they can later be stored in (presumably static) long-term memory and then retrieved on purpose.…”

“…Recent theoretical research on IR (although the first record goes back to 1943, Craik 1943) elucidated its main advantage: it allows detaching the behavior from direct environmental control by, e.g., inhibition of the motor execution (Hesslow 2002;Cruse 2003;Kuhn and Cruse 2005;Aitkenhead and McDonald 2006;Mohan and Morasso 2007). This enables responses of the organism to features of the world that are not directly present and hence to appropriately plan ahead.…”

“…Besides, IR can be evoked mentally (without direct sensory input) to evaluate potential solutions. It is widely believed that IR is a prerequisite of a cognitive behavior (Cruse 2003). Particularly, the IR concept has been used to control a multisegmented manipulator with redundant degrees of freedom (Steinkuhler and Cruse 1998).…”

Modelling memory functions with recurrent neural networks consisting of input compensation units: II. Dynamic situations

“…Such a separation is generally not possible in this 'holistic' network. Rather, the situation model may be used for perception as well as for control of action (Cruse 2003).…”

“…Furthermore, the networks could be used to control motor output in order to produce movements according to the dynamics of the respective representation. This means that the same type of networks can be used for imagination, perception and the control of action (Cruse 2003). Accordingly, these networks might form a basis to explain findings from psychological studies as mentioned above as well as neurophysiological results concerning the so-called mirror neurons (for a review see Rizzolatti and Craighero 2004).…”