Simulating a story character's thoughts: Evidence from the directed forgetting task

- List-wise directed forgetting effects can be found during narrative reading.
- Readers infer and simulate the mental activity of story characters.
- Readers remember and forget what story characters remember and forget.

Readers' memory representations have been shown to include the sensory details of characters' movement, dialogue, and navigation through space and time (e.g., Glenberg & Kaschak, 2002; Gunraj, Drumm-Hewitt, & Klin, 2014; Levine & Klin, 2001; Zwaan, 1996). We ask whether readers also encode the mental experiences of story characters, such as their thoughts and goals. To examine this question, we used a variation of the list-method directed forgetting paradigm (Bjork, 1970), with two word-lists embedded within a narrative. In contrast with the traditional directed forgetting paradigm, it was the story character, rather than the participant, who needed to remember List 1 or forget List 1. If readers take the character's perspective, the character's intention to remember or forget List 1 should influence the reader's intention to remember or forget List 1. This, in turn, should produce the typical pattern of effects for directed forgetting: decreased recall for List 1 (costs) and increased recall for List 2 (benefits) in the Forget condition relative to the Remember condition. The List 2 benefits were found across experiments, even without explicit instructions to forget or remember List 1. However, the List 1 costs were not reliable. Results are discussed within Sahakyan and Delaney's (2003, 2005) two-factor account of directed-forgetting, in which the List 1 and List 2 effects are dissociable. More generally, we conclude that when readers are actively engaged in a story, they may infer and simulate the mental activity of the characters, remembering and forgetting what the story characters remember and forget.