Previous research has suggested that infants' early listening experience affects their ability to perceive speech. Many psychologists and linguists have also suggested an early impact of listening experience on prelinguistic production. This belief was formalized in the babbling drift hypothesis (see, e.g., Brown, 1958), which predicts that babbling begins to approximate characteristics of the mother tongue as infants approach meaningful speech. In order to investigate this hypothesis, four experiments were conducted in which adult listeners' perception of the babbling of infants from different language backgrounds was tested. In the first two experiments monolingual English and bilingual English-Spanish adults judged the babbling of fourteen 7–10 month-old English- and Spanish-learning infants. The third and fourth experiments investigated the babbling drift hypothesis with older infants (11–14 months of age).

For all experiments conducted during both the beginning and the end of the babbling period, adult judges were unable to identify language background significantly above chance level. Therefore, the findings do not support the babbling drift hypothesis. However, bilingual and monolingual judges showed consistently different patterns of judgment with regard to particular infants and utterances. Thus, it appears that judges are influenced by their language background even if the influence fails to improve their success in determining infant language environment.

This study examined phoneme acquisition in three phonologically impaired children to determine whether treatment needed to be provided continually until correct productions were consistently observed during conversation. This was done by examining the effects of withdrawing treatment on several target phonemes at predetermined performance levels. In addition, production of the target behaviors in untrained single words versus connected speech was compared as measures for monitoring progress during the phoneme acquisition process. Both ABA(B) and multiple baseline time series designs were employed. The results indicated two general patterns of phoneme acquisition. In one, the children continued to progress in phoneme acquisition, even though direct treatment on the target behavior had been withdrawn relatively early in the acquisition process. In the other, the children failed to continue their progress in phoneme acquisition when treatment was withdrawn, requiring the reintroduction of treatment until high levels of correct production were maintained over several weeks. In all cases, production of the target behavior in untrained single words and connected speech reflected a similar phoneme acquisition trend.

This study examined the factors that best predict which elementary schoolchildren will ultimately transfer out of a French immersion program. At the end of one school year, French immersion teachers nominated poor-achieving students who might transfer to an English stream and who may elect not to leave the program. All of the children were immediately tested and parents and teachers were interviewed. The status of the project children was monitored over the next year to determine which children transferred to an English stream. Analysis of the pretransfer data indicated that although cognitive-academic variables may be necessary conditions for transfer, they are not sufficient conditions. Specifically, while the children who transferred experienced academic problems, these were no more severe than those of children who did not transfer. Rather, the transfer children were unique in terms of their poorer attitudes, motivations, and nonacademic behaviors. This pattern of results was found for the teacher, parent, and child data. The analyses suggest that attitudinal and motivational factors are of primary importance to the continuation of second-language study by young learners.

Many schools encourage deaf children to speak and sign simultaneously; however, there have been no satisfactory descriptions of this behavior. The research reported in this paper explored the correspondence between speech and sign and the resulting correspondence of the messages conveyed by each channel. The results suggest a synergistic model for this bimodally coded form of English.

Students speaking Hawaiian-English may have difficulty understanding the information presented in the classroom and thereby have less success at school. Children from standard English-speaking backgrounds and from Hawaiian-English-speaking backgrounds were randomly assigned to one of six listening conditions. The conditions consisted of the same selections, but the selections were either in standard English, standard English with Hawaiian pronunciation, or Hawaiian-English. Each of these three versions, in turn, had a simple or a complex syntax version. The semantic content remained constant over all conditions. There was no difference between Hawaiian-English and standard English groups when all comprehension conditions were considered, indicating that the two groups of children did not differ in comprehension ability. However, Hawaiian-English children gave more correct answers when the selections were in their dialect than in standard, while standard English children tended to give more correct answers when the selections were in standard rather than in Hawaiian-English. Contrary to expectations, use of simple syntax and use of Hawaiian-English pronunciation did not facilitate comprehension of standard English by Hawaiian-English children. Implications for educational accommodations will be discussed.