Poster D73, Monday, March 27, 8:00 – 10:00 am, Pacific Concourse
The Ad hoc Perceptual Grouping of Speech Sounds in the Varying Standards Oddball Paradigm
Chao Han1, Ryan Rhodes1, Arild Hestvik1; 1University of Delaware
Phillips et al. (2000) assumed that varying the standards along acoustic parameters of speech sounds in an oddball paradigm causes auditory cortex to recruit phoneme category information, driving a pure phoneme-based Mismatch Negativity (MMN) response. We put this assumption to the test. Using [d] as the oddball (with 15ms VOT), we created two standard conditions: a “low” condition (with [t]s varying among VOTs of 60, 65, 70ms) and a “high” condition (among VOTs of 75, 80, 85ms). According to Phillips et al., both the high and the low conditions should equally invoke the same abstract phoneme category, hence no difference in MMN should be observed. On the other hand, if auditory cortex can track acoustic statistics and form “ad hoc” groupings (e.g., based on the standard deviation of VOTs within the standards), then we would expect to observe the general distance effect on the amplitude of the MMN (Näätänen et al., 2007), with greater MMN for the “high” condition. The latter prediction was borne out: The MMN for the high condition was significant (p < .025, one-tailed), but the MMN in the low condition was not (p < .2). Although the “ad hoc” grouping predicts an MMN for both conditions, and an additive interaction based on the perceptual distance from the oddball, the results still support a different MMN effect based on perceptual distance, and is unexpected under the varying standards assumption. The results raise questions about whether varying the standards suffices to isolate a phoneme category representation.
Topic Area: LANGUAGE: Other