Metaphors as Tools for Restructuring Knowledge: Metaphor-Based Learning about Computer Systems

We hypothesize that metaphor-based training provides a mechanism for creating knowledge structures (schemas) in declarative memory which permit learners to acquire information rapidly. An experiment to test this hypothesis involved: (1) pre-training, in which subjects' schematic knowledge structures were assessed using the Pathfinder algorithm; (2) metaphor-based training, in which one of five common domains served as the vehicle and the user interface for a computer served as the tenor; and (3) post-training that reassessed the schematic structures and tested user knowledge. The similarity of Pathfinder networks for the vehicle and tenor predicted performance during training, (2) performance during training predicted post-training similarity of the vehicle and tenor networks, and (3) post-training similarity of vehicle and tenor networks predicted aspects of post-training metaphor knowledge. The results support metaphor-based restructuring as a mechanism for rapid learning about computer systems.