During 1979-80 we were invited to work in three middle schools (with 8- to 13-year-olds) to clarify what counts as progress in writing. We decided to select, with the teachers, three or four children from each year group. Each child was chosen as a typical representative of a fairly large group of writers in the class having their characteristic achievements, their problems and their interests. We planned to follow these children as they moved through the calendar year and, in some cases, through the school. The natural starting point seemed to be personal stories. As we studied the stories in seminars with the teachers and other colleagues we began to realise that certain features in them appeared again and again. What's more, the stories from different years and from different schools seemed to have a remarkable number of features in common. Taken together these features suggested the points the children had reached in their development as writers. So we began to think we might be able to describe the basis on which the teachers had made their original intuitive selections.