Abstract:

This study assessed the replicability of the ``perceptual magnet effect (PME)'' [Kuhl, Percept. Psychophys. 50, 93--107 (1991)] and investigated underlying neurophysiologic processes using the mismatch negativity (MMN) evoked potential. A stimulus continuum from /i/--/e/ was synthesized by varying F1 and F2 in equal mel steps. Subjects categorized, rated the goodness of stimuli, and discriminated between stimulus pairs. Results revealed that Kuhl's prototype was identified and rated as a good (albeit, not the best) exemplar of /i/, whereas, Kuhl's nonprototype (NP) was identified as /i/ only 20% of the time, and therefore, was not a suitable NP. In this study, the prototype was the stimulus with the lowest F1 and highest F2 values and the NP was close to the category boundary. Discrimination accuracy was not significantly different in the prototype and NP conditions. That is, no perceptual magnet effect was observed. Furthermore, despite equal mel differences between the stimulus pairs, the MMN was largest for the pair with lowest F1 and highest F2 values. Therefore, the MMN appears to be sensitive to within category acoustic differences. The behavioral and electrophysiologic results indicate that discrimination of stimuli near a prototype is based on the auditory structure of those stimuli.