They propose a model of reader response that balances the contributions made by the reader and the demands made by the text. They exploit, through a discussion of data derived from both adults and children, the kinds of responses people typically produce to different sorts of literary text. They distinguish responses largely in terms of how far they relate to (perhaps irrelevant) individual experiences, how far to (perhaps overly) objective registering of literary technique. They conclude that, so far as children are concerned, they need to be encouraged to develop and trust their personal responses to, and enjoyment of, literary texts; and to develop more critical and objective responses to advertising. Their final chapter on classroom practice sketches out some useful approaches for generating discussion based not on watered down lit. crit., but on "the very personal ways in which readers are affected by what they read".
Martin and Leather's analyses of their data are very impressionistic and arbitrary. "Maybe" and "probably" instantly become the substance of the argument. While their hearts seem to be in the right place, there is something a mite self-indulgent and pre-judged about their whole approach. They would not be likely to convince anyone not already sympathetic with their viewpoint. Nevertheless, their transcripts and discussions are often enjoyable and illuminating, if sometimes long-winded.
Typically, it takes them three thousand words to establish that the significant difference between passages from Enid Blyton and Alan Garner is in the interrogation of the text that the reader is (or in the case of Blyton, isn't) required to undertake. If they had stuck throughout to the concept of "interrogation" rather than the misconceived terminology of "prediction", their fundamental argument would have been clearer. This is a useful book that would have been even more useful had it been more exactingly edited.