Stuttering and phonological disorders in children: Examination of the covert repair hypothesis

The purpose of this study was to evaluate whether the Covert Repair Hypothesis (CRH; e.g., Postma & Kolk, 1993), a recent theory designed to explain the occurrence of speech disfluencies in adults who stutter, can account for selected speech characteristics of children who stutter, particularly those who also demonstrate disordered phonology. Subjects were 9 boys who stutter and exhibit normal phonology (S+NP; mean age = 61.33 mos.; SD = 10.16 mos.; range = 49 to 92 mos.) and 9 boys who stutter and exhibit disordered phonology (S+DP; mean age = 59.11 mos.; SD = 9.37 mos.; range = 45 to 74 mos.). Each child was audio/videotaped during (i) a picture naming task (PNT) designed to elicit a sufficient speech sample for a comprehensive phonological analysis, and (ii) a 30-minute free-play conversational interaction with his mother. In order to test CRH predictions, the conversational speech samples were analyzed for the occurrence of within- and between-word speech disfluencies, nonsystematic and systematic (phonological process) speech errors, and overt self-repairs. In addition, children\u27s articulatory speaking rates and response time latencies (RTLs) were measured. Results indicated that S+NP and S+DP children are generally comparable in terms of their basic stuttering, nonsystematic speech error, and self-repair behaviors, with the possible exception of a difference in the types of within-word disfluencies produced most frequently. CRH predictions regarding the co-occurrence of speech disfluencies and speech errors were supported for nonsystematic ( slip-of-the-tongue ) speech errors, but not for systematic (phonological process/rule-based) speech errors, suggesting that there are fundamental differences in the nature of children\u27s systematic and nonsystematic speech errors. CRH predictions that utterances produced with faster speaking rates or shorter RTLs are more likely to contain speech errors or speech disfluencies were not supported. In fact, utterances containing speech disfluencies or nonsystematic speech errors tended to be produced with a slower articulatory speaking rate. In addition, the present analysis revealed complex relationships between speaking rate, RTL, and utterance length, suggesting that further investigation of these variables is warranted. Although the present investigation did not appear to support several of the CRH predictions regarding children\u27s stuttering and speech errors, findings suggest several possible extensions to the CRH which can be used to account for various phenomena relating to children demonstrating both stuttering and disordered phonology, for example, children who apparently begin stuttering while receiving treatment for articulation or phonological disorders. Findings also lead to further extensions to the CRH designed to account for the occurrence of speech disfluencies based on differences in length and complexity of utterances, as well as speaking rate and pausing behaviors of children who stutter. (This research was supported in part by an NIH Research Grant (DC000523) to Syracuse University.