The minimum initial-portion durations required by listeners for the correct identification of spoken isolated vowels and consonant-vowel (CV) syllables were determined. Eight vowels /i i u u æ ε a Λ/ and 64 CVs comprised of each of eight consonants /b p d t g k d 3 t∫/ in combination with each of the eight vowels were used. Segments consisted of the initial 10 to 150 milliseconds of each stimulus in 10-msec steps. The major findings were (1) clues for better than chance correct identification of tongue advancement and tongue height values for isolated vowels occur within the first 10 msec of the stimuli whereas approximately 30 msec of the stimuli are needed for the tense-lax feature to reach threshold, (2) clues for better than chance correct identification of place of articulation for the stop consonants are found within the initial 10 msec of the CVs whereas approximately 22 msec are needed for the voicing threshold to reach threshold, and (3) the threshold of voicing increases from front to back place of articulation for the stop consonants. The implications of the findings are discussed.

No abstract

The times needed to initiate and terminate voicing in response to series of short segments of 1000 Hz pure-tone auditory signal were studied for stuttering and nonstuttering children. The effects of random reward and nonreward on the phonatory response times also were studied. The experimental group consisted of 20 children, 11 of whom had other speech and/or language problems in addition to stuttering and nine whose only communication disorder was stuttering. The control group consisted of 20 normal-speaking children balanced with the stuttering group for sex and age. Children with speech and/or language problems in addition to stuttering were found to have significantly slower voice initiation and termination times than the normal-speaking children. The children with stuttering as the only speech and/or language problem generally did not differ significantly from the normal-speaking children in phonation times. The data suggest that whereas older stuttering children have longer phonation times than do nonstuttering children, younger stuttering children do not. Voicing times for responses following nonrewarded responses tended to be shorter than those for responses following rewarded responses for both groups.

This study investigates the effects of sentence structure on the number of error sentences and response latency in a sentence-repetition task. Forty female college students repeated short and long test sentences containing either a single self-embedded or right-branching subject-focus or object-focus relative clause. Sentences were also controlled for deletion of the relative pronoun of the relative clause. Sentence structure was found to affect sentence elicited imitation response accuracy and latency in a manner similar to the effects of structure on ease of comprehension. The findings are consistent with a canonical-sentoid strategy explanation of sentence processing during sentence imitation.