Abstract

The development of a computer program that models human approaches to problem solving, studies of human perceptual processes, an account of the foundations of language and meaning--any of these may inform theories of learning and instruction. Sources for ideas about education are legion; some are complementary, and others conflicting. We can develop models from multiple perspectives, applying a range of concepts to problems in the field. The trick is never to be tied to a single way of seeing and solving problems. Too heavy an anchor can keep our minds from wandering freely to novel ideas and approaches. Models are developed from metaphors, which are means of creating new representations of the world. These representations are vital sources of information. We cannot know the world as it 'really' is. Rather, we try to access features of the world through representations created and shared with others. The computational model is one such source of representations. It suggests certain ways of thinking about the mind. It emphasizes some aspects, and downplays or even hides others. No one model can account for all features of a phenomenon, particularly a complex one. One limitation of the computational model is its inability to simulate the creation of multiple representations. These multiple representations, as acknowledged earlier, are essential means of knowing the world. It is a weakness of importance to people interested in the creation and manipulation of knowledge, which includes instructional theorists and researchers. In essence, the computational model must be supplemented with other models, especially ones that can account for the mind's ability to create and value multiple representations