Utterance length, syntactic complexity, and childhood stuttering.

This study examined relationships among utterance length, syntactic complexity, and stuttering in children's conversational speech. Analyses extended prior research by examining several different aspects of syntactic complexity, including sentence structure, clause structure, and phrase structure. Subjects were 12 boys who stutter, age 40 to 66 months, who produced 75-utterance conversational speech samples during free-play interactions with their mothers. Group analyses revealed significant differences between fluent and stuttered utterances in terms of all measures of utterance length and several measures of syntactic complexity. Analysis of the relationships between utterance length and syntactic complexity identified several measures of syntactic complexity that influenced stuttering and were independent of utterance length. Logistic regression analyses revealed that utterance length was better than syntactic complexity at predicting whether stuttering would occur, though neither utterance length nor syntactic complexity was a particularly strong predictor for individual subjects' data. Thus, findings suggest that utterance length and syntactic complexity cannot, by themselves, adequately account for the occurrence of stuttering in children's conversational utterances.