Introduction

Until the early 1980s there was little practical help for psychologists, therapists and others working with memory impaired people. The most widespread strategy used in memory rehabilitation in Britain (and probably elsewhere) was repeated practice on memory games and exercises (Harris & Sunderland, 1981), despite the complete lack of evidence for the success of this method. One of the classic memory experiments of the past 20 years demonstrated no effect of repeated practice for improving memory (Ericsson et al., 1980). A student was trained on a digit span task. Most people have a forward digit span of 7 plus or minus 2, i.e. a range of 5 to 9. This student was no exception. However, after almost daily practice for 20 months he managed to increase his digit span to a phenomenal 80. He did this by associating the digits with times and distances from the field of athletics which was one of the student's areas of interest. Despite this incredible performance on a forward digit span task, the student, assessed on a similar yet slightly different task, namely span for consonants, scored 6, i.e. in the normal range. In other words, there was no generalization at all of his new skill to a related task.

In the introduction to Clinical Management of Memory Problems (Wilson & Moffat, 1984), perhaps the first practical book for health service staff working with brain injured memory impaired people, Wilson said: