yslexia research and practice sometimes appear disconnected, as if research has little bearing on the content and teaching style that practitioners adopt. This special issue emphasizes the important role that research can play in informing practice. The following papers that were presented at the recent BDA International Conference at the University of York all consider the educational implications of their research findings. So many excellent contributions were received that we faced a difficult choice in selecting the final papers presented here. The primary theme of the papers submitted centred around research into the salience of large and small units while a secondary theme was writing difficulties and their remediation. Consequently, we restricted our choice of papers to these themes in order to examine the issues more fully. Other interesting papers which were presented at the 1997 BDA International Conference will appear in future editions.
The issue of large (e.g. onset and rime) versus small units (phonemes) has generated much controversy in recent years amongst both researchers and practitioners. The papers included in this issue are no exception. The controversy centres around the developmental stage at which small versus large units assume importance in enabling children to learn to read more proficiently.
Philip Seymour and Lynne Duncan present evidence from longitudinal studies carried out in Dundee that show that children in the first year of school learn to read by exploiting phonemic segments in words, in spite of the fact that many of these children had good rime awareness as pre-readers. Later, after the age of 8, they begin to take note of onset and rime boundaries and to use this information when reading nonwords. Thus, orthographic development seems to move from small units towards larger units. It follows that in the foundation stage of reading, teaching emphasis should be placed on basic letter sounds and phonemic units.
Deavers and Brown go on to consider this issue specifically in relation to the needs of the dyslexic child. In their first experiment, they showed that dyslexics were less