ABSTRACT
The notion of attention as the operation by which a part of the total stimulus information comes to receive specialized processing bears strong resemblance to the stimulus sampling notion first set forth by Estes (1950) in his groundbreaking paper on statistical learning theory. At that time, the conceptual foundation of mainstream psychology resided in the area of learning, and the prevailing learning theories assumed that the organism unselectively takes in all information provided by the momentary stimulus. Estes departed from the assumption of processing the “whole stimulus” when he assumed instead that only a “sample” of the available information is processed. The response connections of the sampled stimulus elements led to the generation of a response probability, and the reconnecting of the sampled stimulus elements to one of the response classes represented the learning process. The elegant mathematical formalization of these processes, built on the disarmingly simple assumption of stimulus sampling, gave rise to his well-known quantitative predictions of the course of learning in choice situations (e.g., Estes & Straughan, 1954; Neimark & Estes, 1967).
