Representational Issues in an Instance Theory of Automization

点击翻译按钮获取中文摘要

Many cognitive activities are automatic, in the sense that they are performed quickly and effortlessly, with little conscious thought. Automatic processing provides an important basis for a wide range of skilled behavior, from perceptual-motor skills, such as driving a car, to high-level cognitive skills, such as chess and mathematics. This research project addresses how automatic processing is acquired. Understanding of the nature of automaticity has implications for the design of training programs in education, industry, and the military. The project concerns a theory of automatization that Logan has been developing over the past few years. The theory assumes that automatic processing is based on retrieving past solutions directly from memory, without any additional computation. Automatization involves a transition from processing based on a general algorithm to processing based on memory retrieval. The idea is well illustrated in children's acquisition of skill at simple addition: They begin with a general counting algorithm, which allows them to add any two numbers, but after sufficient exposure, they memorize the basic facts and retrieve sums directly from memory without counting. Logan's past research has shown that this learning mechanism can account quantitatively for the increase in speed of processing and reduction in variability that accompanies automatization, which are the most robust phenomena in the literature on automaticity. This project will examine the learning process more directly, investigating the kinds of mental representations people acquire as they develop automaticity.