In part I of this paper the general characteristics of 'teacher talk' and that of four language unit teachers are examined, and the effect of different move types on the degree of spontaneous speech exhibited by their pupils is shown. Qualitative differences in overall style of interaction among the four teachers were found which appeared to reflect differences in pedagogic goals and individual beliefs regarding the importance of conversation rather than the degree of control they exerted over subsequent pupil moves through their use of different move types including repair. The implications of the discrepancy found in some cases between what the four teachers reported in a question naire to be important aims and objectives and what took place in practice are also considered.

Various factors that make language development vulnerable in twins and multiple birth children are discussed. Researchers have investigated single explanations for language impairment in twins. However, this paper argues that in individual cases the language impairment is more often the outcome of a complex interaction of factors. Some factors play a greater part than others in some cases. It is suggested that clinicians can develop a dynamic model of language acquisition from understanding the effect of multiple birth and other exceptional circumstances on language acquisition. This model is needed to underpin investigations into individual cases of language impairment. An approach to assessment is described that identifies factors operating in individual cases of multiple birth, providing an essential precursor to implementing appropriate case-based intervention.

In spite of current interest in conversational disability, and the growing number of children who are being diagnosed as exhibiting this disorder, there are few satisfactory procedures that can be used to assess their conversation. In the following paper, data from the naturally occurring conversations of two children, who were identified as having difficulties in participating in conversation, are presented. The data were scanned for the following conversational skills: securing the attention of the addressee before speaking, initiating new topics of conversation effectively, using the grammatical forms that are most frequently associated with particular linguistic functions, filling response slots effectively, maintaining cohesion within the discourse, and taking steps to repair mishearings or misunderstandings. This kind of assessment, based on the principles of conversational analysis, provided an objective appraisal of the conversational skills of the children, and also highlighted potential areas for therapeutic intervention. Many of the conversational behaviours exhibited by the two subjects corresponded with those that have been observed in very young children who are developing language normally. It is suggested that the subjects studied here presented with a severe delay in their conversational development rather than a disorder.

In this second part of a study which examined the general characteristics of the talk of language unit teachers and the effect it had on that of their pupils, the similarities and differences found amongst the teachers are discussed and illustrated by examples of teacher-child interaction. The possible implications for teacher practice are considered.

This study evaluated a new therapy approach. 18 children with phonological disorders resistant to conventional therapy had phonological awareness training followed by conventional speech therapy. The children's speech production improved as did their phonological awareness. In this heterogeneous group, the patterns revealed in their individual responses to therapy proved diagnostically significant.