Associationism and Connectionism

Connectionism is a powerful form of associationist learning theory that broadly reflects the computational style of the brain. Information is represented as distributed patterns of activity over densely interconnected networks of neuron-like processing units, and learning occurs through modifications of the strength of connections between units. Networks are trained on specific tasks and their performance is compared to that of humans. Models of conceptual representation, word reading, morphology, speech recognition, speech production, and syntax have displayed emergent rule-like behavior. The success of these models challenges the traditional separation of item memory and abstract, symbolic rule systems, and provides an empirically testable alternative to nativist approaches to learning.