Abstract
Cognitive, neuropsychological, and neuroimaging evidence suggests that remembering the past and imagining the future rely on overlapping processes in episodic memory. The three experiments reported here examine the consequences of remembering the past and imagining the future on the accessibility of other information in memory. Participants first studied events associated with a specific context and then either (a) retrieved past autobiographical events associated with that same context or (b) imagined future autobiographical events associated with that same context. Replicating and extending evidence of retrieval-induced forgetting, remembering autobiographical events from the past caused participants to forget the related studied events. However, imagining future autobiographical events failed to cause participants to forget the related studied events. These results suggest an important difference in the memorial consequences of remembering and imagining.
Notes
1It should be noted that this paradigm differs somewhat from the standard retrieval–practice paradigm (Anderson et al., Citation1994). In the standard paradigm participants are first exposed to all items in an initial study phase and then, in a separate phase, engage in retrieval practice for a subset of those items. In the current paradigm a trial by trial procedure was employed such that participants engaged in retrieval practice immediately following the study of information on some trials but not on other trials. Importantly, retrieval-induced forgetting has been reliably demonstrated using this type of paradigm (e.g., Storm, Bjork, & Bjork, Citation2005, Citation2007; Storm & Nestojko, Citation2010).
2Interestingly, the difference in baseline recall in the past and future conditions only emerged in the second half of the trials. Thus by analysing only the initial 10 trials in the first two experiments we can compare the consequences of remembering the past and imagining the future on a between-participants basis and with nearly equivalent baseline performance. A 2 (Baseline vs Retrieval/Imagine)×2 (Past vs Future) mixed-design ANOVA revealed a significant interaction, F(1, 78) = 6.09, MSE = .11, p < .05, η 2 =.07. Studied events associated with remembering the past (M = .25; SE = .03) were recalled significantly worse than baseline (M = .34; SE = .03), t(39) = 3.16, p < .01, d = .50, whereas studied events associated with imaging the future (M = .34; SE = .03) were not recalled worse than baseline (M = .32; SE = .03), t(39) < 1, p > .05, d = .08. These results provide additional evidence that the interaction between forgetting in the past and future conditions is not driven by differences in baseline performance.