ABSTRACT
There is a large consensus on the idea that discourse comprehension involves three levels of the reader’s mental representation in memory1 (Graesser et al., 1994; Johnson-Laird, 1983; Just & Carpenter, 1987; Kintsch, 1988; Schmalhofer & Glavanov, 1986; van Dijk & Kintsch, 1983): surface, textbase, and situation model. The surface representation corresponds to the exact wording and syntax of the text. Although this surface structure may be initially available in memory, this level of representation has been shown to decay rapidly (Kintsch et al., 1990; Sachs, 1967). The textbase representation captures the meaning conveyed by the text, independently of its surface formulation, but remains closely tied to the text’s formulation. The situation model or referential representation of discourse is “the reader’s representation of the world the text refers to” (Just & Carpenter, 1987, p. 195) or “the cognitive representation of the events, actions, persons, and in general the situation a text is about” (van Dijk & Kintsch, 1983, pp. 11-12). It is based on a construction of the events or situation a text is about in episodic memory, that is, a subjective representation of a fragment of reality in the reader’s mind. Johnson-Laird (1983) and other researchers have used the term mental model to refer to analogous representations of situations and world events that fulfill different roles in comprehension and reasoning. Throughout this book, I use both terms interchangeably to refer to the reader’s mental repre-
sentation, although each of these constructs has different structural and functional characteristics.
