Self-Explanation: Enriching a Situation Model or Repairing a Domain Model?

When people are reading a difficult text, one of the things they may do is talk themselves through the difficulty. That is, they might start explaining to themselves the parts they do understand, try to think of related knowledge that might help them understand the rest, and/or make an effort to problem-solve the gaps in their understanding. These efforts by readers to explain a text to themselves are what we call self-explanation (Chi and Bassok, 1989). Because self-explanation is directed toward a goal (understanding the text, or what the text describes), and requires conscious thought and effort, we think of it as an intentional strategy on the part of the reader. In this chapter, we consider the process of self-explanation in the context of two accounts of learning from a text, the Mental Model Revision view of Chi (2000) and the ConstructionIntegration model (Kintsch, 1998). Although there are some core similarities between these accounts of how readers combine old and new knowledge, they do yield some different predictions about intentional processes in general and selfexplanation in particular. Some evidence that speaks to these differences is presented in the latter half of the chapter.