Abstract: Situation model is the deep representation of text reading combined with readers’ background knowledge, which involves several dimensions, of which goal state and time representation are most important. A goal is an effective mechanism of foregrounding, and tracing the goal helps to comprehend the causality of the story (Zwaan, 1998). Although recent research of time representation focuses on the temporal shift and temporal distance, most studies repeatedly verified Anderson, Garrod, and Sanford’s (1983) scenario model and Zwaan’s (1996) strong iconicity assumption. Kelter et al. (2004) found the far-close effect of temporal distance and differentiated tracking as temporal distance from a fresh start as temporal shift. Claus et al. (2006) preliminarily explored the relationship between temporal representation and the construction of flashback events. To date, however, no study investigated the relationship between goal state and time representation, and in fact, whether temporal information is continuity or interruption counters for utilizing, continuously tracking, and integrating goal information. Hence the present study is to explore the functions of goal state and temporal information on situation model processing.
A total 174 eighth grade students participated in this study, with 60 participants in experiment group 1, 60 in experiment group 2, and 54 in experiment group 3, respectively. Using moving window and recording the time of recognition sentence technique, three experiments were designed to explore the functions of goal state and temporal information on situation model processing on Pentium III computer, using E - Prime software. Subjects were required to self-regulate their reading. When subjects finish reading one sentence and press the space key, next sentence emerged and the former sentence disappeared. After finishing one text, the emergence of a red "?????" prompted subjects to answer a comprehensive question about the text. The reaction time of probe sentences were recorded and analyzed with repeated measure MANOVA.
The results of experiment 1 indicated that the goal state and temporal shift were interrelated to each other. In the conditions of null goal and completed goal, the larger temporal shift took the subjects more time to read the probing sentence; furthermore, different time shifts (small and large) have varied effects on situation model processing. In the condition of uncompleted goal, however, there was no significant difference in the time used for the probing sentence with small or large temporal shift. In experiment 2, reading with the probing sentence in the condition of uncompleted goal was faster than that in the condition of null goal and completed goal; reading with the probing sentence in the condition of small temporal distance was faster than in the condition of large temporal distance. Additionally, when the temporal distance became larger, the reaction time of reading with probing sentence in the condition of uncompleted goal did not differ significantly from those in the condition of null goal and completed goal. In experiment 3, when the factor of goal was given in the form of flashback events, the main effect of goal state on information integration was significant: reading with the probing sentence with no goal was faster than that in the condition of null goal and completed goal.
To conclude, this study found that: (1) in the sequential event structure, uncompleted goal, i.e., the focus of continuously tracking, cannot help the integration of the text information in the condition of increasing temporal shift; uncompleted goal accelerates representing the information and no far-close effects appear. (2) in the sequential texts, in the condition of no goal and completed goal, large-small effects with increasing temporal shift and far-close effects of temporal distance are evidenced; in the process of constructing the flashback events, the reader represents the flashback events as the background of temporal distance, and the goal enhances integration of information.

Key words: situation model, goal state, temporal shift, temporal distance, flashback events