Connectionist modelling in cognitive sciences

The purpose of the paper is to present basic principles of connectionism and its position within contemporary cognitive science. Connectionist paradigm postulates thinking as a parallel processing of non-structured information by simple calculations performed by neurons that are deeply mutually interconnected. The basic numerical tools of connectionism are represented by so-called artificial neural networks, which are immediately applicable to the study of many cognitive functions at different levels of complexity and sophistication. Connectionism has brought with it a number of important philosophical issues and concerns. Connectionism as a dramatic shift from more traditional accounts of cognition has forced philosophers to reconsider many assumptions based upon earlier theories.