Sound-discrimination learning and auditory displays

Abstract

Human listeners can learn to discriminate between sounds that are initially indistinguishable. To better understand the nature of this learning, we have been using behavioral techniques to examine training-induced improvements on basic auditory discrimination tasks. Here we report how multiple-hour training differentially affects the discrimination of sound frequency, intensity, location, and duration, and how learning on a given discrimination condition generalizes, or fails to generalize, to stimuli not encountered during training. We discuss how these data contribute to our understanding of discrimination learning and of the mechanisms underlying performance on particular trained tasks, and explore the implications of this learning for the design and evaluation of auditory displays.