Research on the language development of children in immersion programs has uncovered a certain amount of data. Immersion students have less well developed productive skills than their francophone peers and passive receptive skills which approach those of francophones. Error analysis studies have found that immersion students reach a plateau in their language usage and that some errors are systematically repeated by individuals and by different learners in different settings. Interlanguage research has called attention to the "fossilization" phenomenon and suggested that feedback is a key factor. In immersion programs, once students acquire enough language for communication, the teacher may not correct language errors, thus providing positive feedback for both language and answer. In line with the pidginization hypothesis, it has been noted that immersion students develop a simplified language system which allows them to communicate with teacher and friends. Gradually, they begin to model their speech more on their classmates' speech than on their teacher's. According to the research implications, there will be little development as long as the acquired language meets communication needs. To remedy the situation, classroom strategies might be modified to include emphasis on correct linguistic forms, and steps might be taken to increase students' exposure to the target language. (AMH)