Multidimensional Situation Models

The research landscape examining situation models in discourse has been transformed considerably in the last 20 years. Van Dijk and Kintsch (1983) originally included the concept of a situation model to address issues that were problematic for earlier versions of their theory (Kintsch & van Dijk, 1978). Specifically, van Dijk and Kintsch (1983) argued that situation models were necessary to explain issues of reference, coreference, coherence, perspective taking, translation, individual differences, memory, reordering effects, problem solving, updating knowledge, and learning. One might first notice the comprehensive nature of such a list. It is not surprising, then, that there is general agreement regarding the theoretical importance of situation models. What is surprising, however, as originally pointed out by Glenberg, Meyer, and Lindem (1987), is the lack of agreement regarding what constitutes a situational model and the types of information it might contain. For the purposes of this chapter, we use the term situation model to refer to a discourse representation that captures aspects of a micro-world created by the reader (Johnson-Laird, 1983; van Dijk & Kintsch, 1983). In this sense, a situation model can include propositional information, but also information beyond that given in the text proper. For example, situation models can contain information related to the gist of the text, a reader’s potential background knowledge, and inferences not explicitly stated in the text (Zwaan & Radvansky, 1998). One theory of situation-model construction (i.e., the event-indexing model) suggests that readers comprehend information in the story world at an event level (Zwaan, Langston, & Graesser, 1995). In this sense, events are