Terrance M. Nearey
Dept. of Linguist., Univ. of Alberta, Edmonton, AB T6G 2E7, Canada
Phonetics involves three domains of events: (1) acoustic/auditory, (2) articulatory/gestural, and (3) symbolic/phonological. Theories of perception hold different views about the relationships among these domains. Auditorists (e.g., Kingston and Diehl) emphasize a strong (i.e., simple, direct) relation between (1) and (3). Gesturalists (e.g., Liberman and Mattingly) propose a strong relation between (2) and (3). Advocates of double-strong theory (e.g., Blumstein and Stevens) propose strong relations of both (1) and (2) to (3). Difficulties with all these theories will be discussed. An alternative, ``double-weak'' theory will be described and evidence supporting it will be presented. This approach views speech production and perception as distinct but cooperative systems. This view accepts the gesturalist contention that the acoustic mapping if phonological units is modified by context in ways that are important to perception. However, it also contends that the accommodation of contextual effects in perception is limited and highly stylized. Assuming a concomitant stylization of gestural patterns in production, relatively simple pattern-recognition strategies in perception may suffice for successful communication.