Abstract: Moving window display technique was used to investigate the mechanism of constructing and updating situation models in text-reading. According to whether the characteristic elaboration was consistent with subsequent target action carried out by the protagonist, the experimental materials were divided into three categories: enablement, disablement and re-ablement version. In experiment 1, the mechanism of constructing and updating situation models in text with explicit goal information was investigated to support the here-and-now hypothesis. In experiment 2, the mechanism of constructing and updating situation models in text without explicit goal information was investigated to support the memory-based text processing view. In experiment 3, the difference between the here-and-now hypothesis and the memory-based text processing view was further investigated with probe tasks. The results indicated that the mechanism of constructing and updating situation models in text-reading was different with the changing of information. A processing theory of bi-mechanism for constructing and updating situation models is suggested, which proposes that the mechanism of constructing and updating situation models in normal reading includes both coordinating integration supported by the memory-based text processing view and goal-information integration supported by the here-and-now hypothesis

Key words: situation model, constructing, updating, coordinating integration, goal-information integration