A Schema Theory Account of Some Cognitive Processes in Complex Learning.

摘要

Procedural semantics models have diminished the distinction between data structures and procedures in computer simulations of human intelligence. This development has theoretical consequences for models of cognition. One type of procedural semantics model, called schema theory, is presented, and a variety of cognitive processes are explained in terms of the theory. In schema theory, the flow of processing control is determined not by a central monitor, but by interactions among the conceptual entities (schemata) that make up the model. Intelligence is distributed in this model. Schemata interact by providing activation resources to each other. Instantiation is the special process whereby a partial copy of a strongly activated schema is created. In this copy, the variables of the schema are filled with particular values. Such copies make up specific or episodic memory. The schemata on which they are based comprise generic or semantic memory. Many of the phenomena of consciousness and of short-term and long-term memory are explained on the basis of the activation processes of schema theory. Unactivated schemata are equivalent to all the unconscious knowledge in a person's long-term memory. Three dimensions for distinguishing or comparing schemata are proposed; function, abstractness, and scope. The contrasts between multi-store models of cognition and schema theory are summarized.