| |
Connectionism, in linguistics, is a theory of cognition which provides a model of how learning takes place. The model draws inspiration from the way the billions of neurons in the brain are interconnected in complex ways to produce a network of associations. It is suggested that during brain activity the neurons convey signals to associated neurons, causing the latter to be either aroused or inhibited. The arousal of one neuron by another reinforces the connection between them, while inhibition causes the connection to be weakened. The overall pattern of connection strengths is taken to represent the knowledge of the system at a given time. And as the relative strengths of connections alter with progressive reinforcement and inhibition, the state of knowledge changes accordingly.
In connectionist theory, computer models are designed to mimic genuine neurological processes, by using numbers to represent the relative strengths of connections between neurons, or units, in a network. However, since very little is known about what actually happens at the neurological level in the brain, it is a moot point whether connectionism does in fact provide a valid metaphor for cognitive processes. Furthermore, it is quite conceivable that there is more to cognition than can be explained by neurological activity alone. MS
See also thought. |
|
