Learning without limits: from problem solving towards a Unified Theory of Learning

Abstract

Learning is usually studied on the basis of binary distinctions like implicit vs. explicit learning, using instance vs. using rules, connectionist vs. symbolist, etc. In this thesis it is argued that many of these distinctions are not useful at all in understanding learning. This statement is supported by a large set of models in ACT-R, a cognitive architecture developed by J.R. Anderson. These models demonstrate that deeper understanding is often gained when the traditional distinctions are ignored.