Acoustics of stop consonant-vowel relationships during fluent and stuttered utterances

Acoustic duration measures were made on the stop-gap, voice onset time (VOT), and vowel portions of the utterance [tu] in the context “425” as spoken ten times or more by five normal speakers, four mild stutterers, and four severe stutterers. The fluent utterances of the severe stutterers were significantly longer than those of normal speakers on stop-gap and vowel durations but not on voice onset times. Despite the slower speech of the severe stutterers, indications of normalization were found in the relatively stable consonant-to-vowel ratios across groups and in the lack of any significant differences of segment durations as a percentage of total utterance times. Normal speakers and mild stutterers showed an inverse relationship between stop-gap and VOT measures across speakers, but severe stutterers did not maintain this relationship for either fluent or stuttered tokens, nor did a mild stutterer suspected of covert stuttering. Stop-gap durations increased while VOTs stayed within normal limits for fluent tokens. For stuttered tokens, some subjects increased the stop-gap and some the VOT depending upon the location of the block. Abnormal gap to VOT ratios may serve to detect convert stuttering.