Abstract: Time plays a central role in how people experience the world. In recent years, the question of how temporal information is processed during text comprehension has gained increasing interest in psycholinguistic research. Berry et al (2006) conducted their experiments with narratives containing flashbacks to test chronological hy-pothesis and background hypothesis. The results supported the chronological hypothesis exclusively. However, a flashback usually has a close thematic relationship to its preceding sentences. It seems plausible that a flashback is used at the exact point where the readers are meant to use it as background information. Would it be possible that the null effect in background hypothesis was due to the weak supplementary relationship between the flashback and the event described before?
In this study, moving window display technique was used to examine the rationality of chronological and background hypotheses. Participants were asked to read narratives describing 4 successive events in non-chronological order with E1 being mentioned in a flashback (E2-E3-E1-E4). The experimental materials had two versions: the first one was the original passages from Berry Claus et al (2006), and the second one highlighted the background relation between E1 and E3. The information about the time duration of E2 was ma-nipulated, and the mental accessibility of E1 was tested in different methods at the end of each passage. The results consistently showed that the reading times for anaphoric sentences or the probe response latencies were significantly longer in the long-duration condition than in the short-duration condition when the background relationship was highlighted, which supported the background hypothesis. It suggests that the background rela-tion between E1 and E3 is important for temporal representation of flashback event in narrative comprehension. If the supplementary function of flashbacks has enough been highlighted, participants would represent it in the form of E2-E3-E1-E4, otherwise, in E1-E2-E3-E4.

Key words: flashbacks, temporal representation, chronological hypothesis, background hypothesis, temporal distance effect