Human Learning and Learning Systems
The first section of this chapter discusses the general logic of computer simulation of human learning and presents, as an example, the GRAPES simulation of learning to program in LISP. A critique of GRAPES and the logic of the simulation is offered.
In the second section, human learning and the unitary cognition hypothesis are discussed within the architecture of ACT theory. The human learning of geometry and its modeling by ACT are described.
In the third section, the general logic of inductive learning and the specific logic of LEX, a program that learns symbolic integration, are discussed. The language and grammar of LEX, version spaces and inductive learning in LEX, and the limitations of LEX are described and analyzed.
In the final section, learning and a general theory of intelligence are discussed. Computer learning is contrasted with human learning. The unitary theory of cognition is analyzed.
The computer simulation of cognition serves several functions. The valid simulation of cognitive theory on a computer entails accurate specification of the cognitive processes, their interrelationships, and their consistency ( Broadbent, 1987). The effect of changes in the parameters and values of the cognitive process can be determined within the efficient and convenient domain of a computer program. On the other hand, the computer simulation