The use of oral narratives as a measure of language processing and production skills has been used with increasing frequency during speech and language evaluations. By analyzing narratives we have an insight into strategies that children use for organizing, comprehending, and producing language. Narratives also reveal how different cultural groups organize and make sense of their world. One of the shortcomings of the existing research on narratives is that it is limited in scope. That is, studies have primarily examined narrative development among Standard English (SAE) speaking children. Far less attention has been given to examining the narrative production of African American English (AAE) child speakers. Qualitatively the study design was drawn from an ethnographic perspective. African American subjects were selected from low income community of Springfield, Massachusetts, where subjects participated in two after school programs located a mile apart. Video taped and audio taped data were collected on site at the afterschool programs. A total of 15 subjects who met criteria for participation in the narrative activity were video and audio taped as the told personal stories to a familiar adult. All video tapes were transcribed for each child. A total of 71 narratives were subjected to analyses. Narratives were analyzed using five different procedures: thematic, componential, highpoint, story grammar, and a micro-sociolinguistic analysis. Among the findings were: (a) higher frequency of "topic centered" narratives than "topic associated" narratives, (b) production of a repertoire of narrative structures, (c) higher frequency of complete and complex structures than any other structures within story grammar analysis, and (d) higher frequency of the classic structure than any other structures within highpoint analysis. The clinical and theoretical implications with regard to deficit theory, Africanisms within narrative discourse, and educational and speech/language assessment for the AAE child speaker were also discussed.

Current methods for assessing African American English (AAE) speaking children's semantic knowledge are seriously flawed. Many AAE-speaking children who do not have language disorders perform poorly on standardized vocabulary tests. However, there is no reason to believe that all of these AAE-speaking children are deficient in their ability to learn a rich and functional vocabulary. Existing vocabulary tests often are culturally biased because lexical items are selected and normed on middle-class Euro-American children. This results in an inherent bias against linguistically and culturally diverse populations. Some African American children have less exposure to the lexical items selected for use on standardized tests than Euro-American middle-class children. These cultural and language differences become exacerbated when these children enter school. Frequently, AAE-speaking children are referred to the school speech-language pathologist (SLP) for language testing. However, the SLP is often ill-equipped to provide an unbiased evaluation due to reasons previously mentioned. The problem for the SLP is to determine what areas of semantics to test and what methods should be utilized in this assessment. This study investigated the processing-dependent measure of fast mapping as an alternative method of assessing semantic knowledge in children. AAE and Standard American English (SAE) speaking children between the ages of four and six were presented with two comprehension tasks involving real verbs and the fast mapping of novel verbs in four different argument structures (intransitive, transitive, transfer, and complement). These tasks were developed to evaluate how children use syntactic bootstrapping to help fix the meaning of new verbs. The participants' performance on the alternative assessment measure was compared to their performance on a commonly used psychometric vocabulary test. Although significant differences were found between AAE- and SAE-speakers in the transitive argument structure for real verbs and transfer argument structure for both real and novel verbs, overall results indicated that both groups were able to fast map novel verbs. A performance gap between AAE and SAE participants on the psychometric vocabulary test was noted in this study. These results suggest the feasibility of fast mapping as a method to reduce test bias in semantic assessment.

The purpose of this investigation was to compare conversational repair strategies of hearing and deaf/hearing-impaired children in response to a partner's indication of communication breakdown. Experimental subjects were eight profoundly deaf children, ages 4 years to 7 years, 6 months, who used total communication. Control subjects were eight hearing children, matched by age and sex to experimental subjects. Each subject was videotaped while individually engaged in two language sample elicitation activities, a structured and an informal communication situation. During the conversation in each experimental condition, the investigator initiated ten stacked clarification request sequences consisting of three neutral queries ("Huh?," "What?," and "I don't understand.") per sequence. The sequences were inserted on alternating items about which the subject spoke, and/or when the subject produced an intelligible utterance of sufficient complexity that a clarification request had validity. The videotaped language samples of the clarification request/repair response sequences were transcribed verbatim. Clarification repairs were coded as repetition, revision, addition, cue, discussion, and inappropriate responses. Variations in the use of total communication by the experimental subjects, and the use of pointing, or other mode variations, by the control subjects were also coded. Frequencies and percentages of occurrence were derived for each request type in each repair category and for each language condition. Chi-square analyses were used to determine the relationships between the variables. Results indicated that while all subjects were aware of the obligatory nature of the clarification requests, experimental and control subjects employed different types of repair strategies. Revision repairs were the most common type of responses, however, deaf/hearing-impaired subjects were twice as likely to revise their utterances while hearing subjects were as likely to repeat as they were to revise their utterances. In addition, hearing subjects were three times more likely to provide cue repair responses as their experimental counterparts. Differences in conversational repair strategies were also evident as the queries in the clarification request sequences progressed. No significant differences were noted within groups when communication conditions were compared. Communication mode variations appeared to have little influence on the coding of repair strategies.