Shahana knew she should start at the top left-hand corner; but after that she wasn’t sure where to read. Danny seemed to have jumped to a box in the bottom corner, which had a kind of cartoon in it. Then there was a table to copy out and fill in, Shahana supposed. But with what? Danny looked equally puzzled - neither had come across books like this in primary school.

Shahana and Danny are not exceptional. Most children in most countries find science books difficult. The United Kingdom has a wealth of schemes, information books, copiable worksheets, software and CD-Rom material for primary science teachers. But much of it is unused, under used or ineffectively used. This is partly due to the nature of the materials - their language and structure can be complex, as they not only “tell” facts but also ask questions, give instructions, contain tables and diagrams and ask for responses. But it is also in part due to ideas about the importance of talk and collaboration in science learning, to fears about the language demands of the materials in relation to children, and to the pressure for differentiation and “customised ” materials.

Shahana and Danny came from a primary school where English was an additional language for 95 per cent of the children. Teachers there tried to simplify science worksheets. They argued that this was time better spent than on explaining the science schemes, which were written by authors who neither spoke the language of their pupils nor understood their difficulties.

Research into children’s use of science materials shows a frequent mismatch between the teaching methods implied by the text and those normally used by the teacher, and that many teachers have difficulties with concepts and technical language. It shows that few children learn to read using the expository kinds of text that predominate in science books and schemes. Instead, they learn to read narrative text, and cease to be taught reading once they have mastered this.

Some programmes try to tackle this using in-service activities on extending children’s ability to interact with expository text. But teachers in training need to analyse existing schemes and texts in terms of the specific difficulties likely to be encountered. For example:

* in most books, the sequence in which the reader moves from words to graphics and back changes (sometimes across the page, sometimes down, sometimes across two pages) without this being made explicit

* new words (technical terms, concepts) appear without explanation (“if the current is made to flow through a lightbulb, the energy of the current will light the lightbulb”)

* graphics (tables, diagrams, cartoons) are assumed to assist understanding independently of the text, but this is often not the case

* questions are often posed in which the message about how to answer is unclear (“Can you think of some other ways in which the energy of an electrical current is used?” Does the question require an oral or written response or neither?)

* tasks are seen as impracticable by the teacher (“Take a short walk. Write down the names of 10 animals you see moving ...”).

To use material effectively, teachers first have to decide that the “message” of a text (for example, that science is relevant to the “real” world, that children can “do science” themselves) is the same one they are trying to put across. So to help children use part of a scheme effectively,a teacher has to look not only at the attainment of pupils (science concepts, language, visual literacy) and the teaching context of the classroom, but also at the potential difficulties presented by specific materials, children’s previous experiences with schemes, and their understanding of what using science text material involves. This suggests training programmes must give teachers experience of:

* various types of published science text material (schemes, information books, readers, magazines, comics, newspaper articles)

* analysing text in terms of the demands it makes on children (linguistic,visual, structural, practical)

* skills for using text (skimming, interpreting graphics, relating illustrations to text, answering questions, understanding instructions)

* what to do if “text message” and “classroom message” conflict.

This does not mean existing schemes are bad. Most are excellent and represent science concepts in ways hard-pressed teachers would not have time to emulate. There is a mismatch between what teachers feel constrained or able to do in science and what schemes and text demand.

But ignoring text does not help Shahana and Danny. First because teacher-made materials may create even more problems, and second because it doesn’t help them to use textbooks in high school. They will need help with most science text material; and teachers may need help with the evaluation and use of science schemes. Therefore science and language teacher trainers must collaborate - with each other and with practising teachers - to examine the use of specific science schemes, especially in relation to materials used to teach reading.

Alan Peacock is a lecturer in primary science at the University of Exeter and author of Opportunities for Science in the Primary School (Trentham Books)