Imagining a Brave New Future: The Effects of Novelty and Plausibility on Episodic Simulation

Abstract

It is now a well-established finding that our capacity to imagine future events relies on our capacity to remember our past. Both abilities rely on a common core network of brain regions, including medial prefrontal, parietal and temporal cortices. More recently however, neuroimaging, neuropsychological and behavioural studies have revealed important differences between past and future events. One particularly striking finding is that the hippocampus, a region traditionally associated with memory, is preferentially engaged for future event simulation as compared to past event recall. It remains the centre of current debate what cognitive differences between past and future events this increased hippocampal activation may reflect. A key difference that has received very little attention is novelty. By their nature, simulations of imaginary future events are more novel than past events in that they are new mental constructions (contextual novelty), but they may also be implausible and dissimilar from past experiences (categorical novelty). The studies in this thesis investigate, by means of fMRI and new behavioural methods, whether contextual and categorical novelty constitute a crucial difference between the past and future events. Study 1 reveals that the contextual novelty of future events gives rise to several differences in event construction processes as compared to past events. Specifically, contextually novel future events take longer to construct and contain fewer details than past events, and interestingly, some of these effects were intensified for categorically novel future events relative to more plausible ones. Study 2 demonstrates that contextual novelty is associated with increased hippocampal activation and connectivity in future event simulations. This finding may explain some of the incongruent findings in the literature, particularly given studies where the contextual novelty of events may have been may have decreased due to pre-imagined future events. Study 3 further examines the construct of categorical novelty, demonstrating that categorically novel future events contain more disparate details and are less plausible, and that plausibility estimations were based on the fluency with which the events were constructed. It is also demonstrated that categorically novel future events are less memorable due to their lack of similarity to past events. Together, these findings expand our knowledge of the nature of future event simulations, and what makes them unique.