Diss. Linköping : Univ., 2002.

This study examines booktalk, that is, teacher-led group discussions about books for children in a Swedish school. The empirical data comprise 24 hours of videorecorded booktalk in grades 4–7. In total, 40 children (aged 10–14 years) were recorded during 24 sessions. The present approach diverges from previous readerresponse studies in that it draws on authentic data, and in that it examines talk at a micro level, applying an approach from discursive psychology. By focusing on authentic book discussions, the study contributes to the development of readerresponse methods. All eight books applied in the booktalk sessions involved some type of  existential issue: freedom, separation, loyalty, and mortal danger (Chapter 4). Yet, such issues were rarely discussed. An important task of the present thesis was to understand why such issues did not materialise, that is, what did not take place. In Chapter 5, a series of booktalk dilemmas were identified. The booktalk sessions were generally lively and informal. Yet, booktalk as such was often transformed into other local educational projects; e.g. time scheduling, vocabulary lessons or reading aloud exercises. Gender was invoked in all booktalk sessions (Chapter 6). In line with predictions from reader-response theory, progressive texts were, at times, discussed in gender stereotypical ways. The findings also revealed a generational pattern in that the pupils discussed fictive children in less traditional ways than adult characters. The interface between texts and life was invoked in all booktalk sessions (Chapter 7). There was, again, a generational pattern in that children entertained ideas other than those of their teachers concerning legitimate topics in a school context. Also, the discussions revealed a problem of balance between pupils’ privacy, on the one hand, and engaging discussions on texts and life, on the other.