Abstract
The representation of knowledge in connexionist systems. The large number of representational formalisms that intent to explain how mind codifies, stores and recovers knowledge is noticed. Simbolic models of mind maintain that the representation of knowledge is obtained through a system of symbols related by means of a complex syntax. According to connectionst models the representation of knowledge inside the artificial neural nets is produced from the simultaneous interaction of a large number of units highly interconnected. The first major contribution to those models is due to McCulloch and Pitts (1943) but it was Rosenblatt who showed that perceptrons could be trained. The new connectionist models are nonsimbolic models of the mind in which the weights of its connections determine the functioning of the network. It is concluded that much more investigation is needed before we can ascertain how information about the world is represented in mind.
