Categorical perception: Lessons from an enduring myth

Abstract

Categorical perception holds a unique place in cognitive science. In terms of its impact on other fields, it is likely the most impactful finding in speech perception over the last 60 years. However, within speech perception, it is widely known to be largely an artifact of task demands. Categorical perception is empirically defined as a relationship between phoneme identification and discrimination. As discrimination tasks do not appear to require categorization, such linkage violates psychophysical laws and was thought to support the claim that listeners perceive speech solely in terms of linguistic categories. However, 50 years of work using a variety of discrimination tasks, priming, the visual world paradigm, and event related potentials has roundly rejected this account. In this talk, I explore the origins and impact of this scientific meme and the work challenging it. This leads to a new understanding of how to use and interpret some of the most basic techniques of speech—phoneme identification along a continuum—and to new ways of understanding the role of sensitivity to withing category variation. This has major implications for understanding language and hearing disorders and development.