This challenging title matches well the research of Rosalind Driver, to whose memory it is dedicated. She began as a science teacher and this perspective coloured everything she did. Her research was outstandingly scholarly, but always strongly related to teaching. When she got her first big research funding in 1983, she said to me, “I’m back in the classroom, Joan. It’s wonderful!” Her last funded task was to convene a group to write about science education, Beyond 2000. She planned to have teachers on the writing group, although it didn’t turn out like that. Just before she died in 1997 she engaged in a public dispute about the status of educational research, writing that it could never replace teachers nor supersede them.

Teachers’ research is usually empirical, as was Driver’s CLISP (Children’s Learning in Science Project). So it is strange to find the editors writing here that research which was practical but lacked theorising became “fatally flawed”.

They argue that science education is still pre-paradigmatic (without agreed theory) as though it is an immature science which awaits some grand universal theory. But education is a social science, where rich description helps us to understand classrooms, students and teachers, and is an invaluable basis from which to suggest improvements to be evaluated. Some of the chapters in this book are a little too thick with uncertain theory to make for easy reading; some suggest no practical way forward. but I shall mention some excellent ones.

The chapter by Nancy Brickhouse and colleagues about college students’ contrasting views of theories in astronomy and biology is interesting, and builds on previous work, including that of Driver herself. An excellent chapter by Bjorn Andersson considers evaluation, Swedish style, as an engine for liberating and improving science teaching, by introducing socio-scientific issues to “renew the curriculum”. The resulting reports show students strongly in favour of this when it concerns topical issues. Such reports are read by teachers and student teachers; so there is a good chance of direct influence on Swedish science education.

By contrast, Justin Dilln has written thoughtfully about empowering heads of science departments to implement new ideas - their own and others’. But his interviews show just how impossible the task has become in the frantic atmosphere of school life in contemporary Britain.

The second section focuses on the curriculum as an area for change. Peter Fensham chronicles successes and failures of changes which were often due to resistance from higher education. Deciding on content is difficult, and he holds out little hope for newer courses concentrating on “higher order reasoning”, which are favoured by Richard Duschl and Stephen Norris.

Duschl argues for learning about the nature of science from classroom investigations and student argument, research that Driver herself was planning to do. Norris points out the difficulties of teaching “about science” which has so many different perspectives, and wants students to do their own investigations, while Jenkins rejects both studies of the nature of science and laboratory work, in favour of seriously long-term projects.

Glen Aikenhead draws on his research on the multiple nature of the cultures of those coming from different homes. He predicts that success awaits any teacher who aims to encourage these different students in our fragmented classrooms.

The final section contains contributions from Gaalen Ericson, Richard Gunstone, Richard White, Piet Lijnse, and Paul Black. Ericson is optimistic about trends towards teachers’ classroom research which is so close to real teaching. but all these chapters exhibit pessimism about present problems, with little change in students’ misconceptions or teachers’ practice. Even Driver’s CLISP undertaking is seen as a failure, at least in part.

But, at the end, Black’s masterly survey of assessment is a well-argued promise of improvement. He wants more research on teachers’ roles in formative assessment, and combined efforts from teachers and researchers to argue for change in assessment policy. Taken together with the Swedish experience, points made by Dillon, and our urgent shortage of teachers, that presents a good direction for improvement, at least in Britain.

Joan Solomon.

Professor Joan Solomon is senior research fellow at the Centre for Science Education, the Open University