Speech revisions in monolingual English and Spanish-English bilingual children

This study explores the manifestation patterns of speech revisions in monolingual English and Spanish-English bilingual children. All speakers exhibit speech revisions to some degree and some researchers have indicated that they may manifest due to linguistic uncertainty (Bedore et al., 2006; Loban, 1976). In the current study, speech revisions were documented in the context of two narrative conditions manipulated to elicit revisions. In one context, a high uncertainty condition, the narrative picture sequence depicted a vague or unclear ending to a story, therefore increasing the speaker’s linguistic uncertainty. In the second condition, the low uncertainty condition, the narrative picture sequence had a logical ending reducing linguistic uncertainty. These tasks were designed to elicit speech revisions in children ranging in age from 3;5 to 5;11. Participants included 33 Spanish-English bilingual Kindergarten-age children, 32 language-matched monolingual English-speaking pre-K children, and 37 age-matched monolingual English- speaking children. All children exhibited typical language abilities based on a language screening measure. The first research question was whether there was a difference in the rate of speech revisions in English between the narratives with high and low uncertainty across the 3 groups of children. The second question pertained to whether the rate of speech revisions in their narrative samples was influenced by task (high vs low uncertainty condition) when language productivity as measured by lexical diversity (NDW), mean length of utterance (MLU) and grammaticality. Results indicated that all of the children across the three groups exhibited fewer speech revisions in the low uncertainty condition than in the high uncertainty condition. There were no differences observed by group for frequency of revisions across task condition. Further, NDW accounted for a significant amount of the variance in frequency of revisions across all three groups. Again, there were no group differences observed in frequency of speech revisions when measures of language productivity were controlled. These results indicate that in an experimental condition, bilinguals were no more susceptible to exhibit revisions than their monolingual peers. Implications for these results and further considerations regarding revisions and the speech production process for monolinguals and bilinguals are discussed.