Connectionism

>An input device receiving signals from the environment or from other neurons, >An integrative device integrating and manipulating the input, >A conductive device conducting the integrated information over distances, >An output device sending information to other neurons or cells, >A computational device mapping one type of information into another,</LI></UL> The basic tenets of connectionism are that signals are processed by elementary units, processing units are connected in parallel to other processing units, and connections between processing units are weighted.

The weights may be 'hardwired', learned, or both.

The weights represent the strength of connection (either excitatory or inhibitory) between two units.

These basic concepts have permitted the development of a spectrum of models in connectionist research.