Purpose
Child phonologists have long been interested in how tightly speech input constrains the speech production capacities of young children, and the question acquires clinical significance when children with hearing loss are considered. Children with sensorineural hearing loss often show differences in the spectral and temporal structures of their speech production, compared to children with normal hearing. The current study was designed to investigate the extent to which this problem can be explained by signal degradation.
Method
Ten 5-year-olds with normal hearing were recorded imitating 120 three-syllable nonwords presented in unprocessed form and as noise-vocoded signals. Target segments consisted of fricatives, stops, and vowels. Several measures were made: 2 duration measures (voice onset time and fricative length) and 4 spectral measures involving 2 segments (1st and 3rd moments of fricatives and 1st and 2nd formant frequencies for the point vowels).
Results
All spectral measures were affected by signal degradation, with vowel production showing the largest effects. Although a change in voice onset time was observed with vocoded signals for /d/, voicing category was not affected. Fricative duration remained constant.
Conclusions
Results support the hypothesis that quality of the input signal constrains the speech production capacities of young children. Consequently, it can be concluded that the production problems of children with hearing loss—including those with cochlear implants—can be explained to some extent by the degradation in the signal they hear. However, experience with both speech perception and production likely plays a role as well.
The Conditioned Assessment of Speech Production (CASP): A Tool for Evaluating Auditory-Guided Speech Development in Young Children with Hearing Loss
