Abstract
We examined age differences in referent choice strategies when narrative organization and presentation mode taxed working memory. Readers could use the following cues to choose referents for an ambiguous pronoun: (a) thematic cue (main minor character), (b) foreground cue (character in main/subordinate clause), and (c) recency of mention of character. Subjects read short narratives with a critical sentence that mentioned the main character in the main clause of the critical sentence (congruent condition) or in the subordinate clause of this sentence (incongruent condition). They chose the main or minor character as referent for a pronoun in the next (target) sentence. Experiment 1 examined how older and young readers use these cues with the Memory narrative presentation (target sentence presented on a separate page from the rest of the narrative). Young readers used similar strategies as those in Morrow (1985), where narrative presentation did not tax memory. Thus, the Memory presentation had little influence on their strategies. However, older readers were less consistent and tended to split in choosing the two characters. In Experiment 2, where narrative presentation did not tax working memory, older readers used similar strategies to the young readers in Experiment 1. Experiment 3 found that elders reading narratives with the Memory presentation recalled the nonpreferred character name after the narrative, suggesting that both characters were accessible from working memory. The study suggests that the Memory presentation mode in Experiment 1 interfered with elders' consistent use of narrative cues in choosing referents, especially when these cues conflict.