This research aims to characterize the speech production of postlingually deafened adults before and after they receive cochlear implants, in order to (a) help evaluate and improve prostheses and (b) constrain models of the role of hearing in speech production. Toward these ends, we have been making longitudinal measures of speech acoustics, respiration and vocal- fold contact area (with electroglottography - EGG). Our work to data has shown that: there are a number of abnormalities pre-implant; changes occur in all parameters measured thus far upon activation of the speech processor over long and short time spans; many of those changes are in the direction of normalcy; and there are relations among the parameters that can guide us in a principled account of the mechanisms underlying most of the changes, whether or not they are in the direction of normalcy. There are also differences among parameters in the timing of changes measured longitudinally which, despite differences among subjects, will contribute to our understanding of the role of hearing in speech production. In this second phase of the program project grant, we will continue to use a within-subject longitudinal design as well as paradigms in which we modify the characteristics of the processor stimulation several times during a single experimental session. In both types of experiment, we will expand the parameters studied to include measures of intonation, period-to-period variability in FO, consonant acoustics, instantaneous airflow during consonant articulation, nasalization, glottal waveform measures, and measures of inter-articulator timing. We will compare suprasegmental measures of deficits in patients' speech perception with anomalies in their speech production. The stimulation modification paradigm will be elaborated to include alterations of auditory stimulation characteristics such as spectral slope, to test specific hypotheses about relations between production and perception. The results of this latter type of experiment may lead to hypotheses about sub-phonemic perceptual mechanisms, which we will also test.
Showing the most recent 10 out of 22 publications
