How event-based memories change as a function of forgetting and consolidation

Episodic memories are composed of multiple elements, from the people we encounter, the locations we visit, and the objects we interact with. These ‘episodes’ are thought to be stored in memory as coherent event representations and are associated with holistic recollection at retrieval, such that the retrieval of one element is dependent on the retrieval of all other elements from the same event. Evidence for this ‘dependency’ has been shown to emerge from the association between the event elements themselves. Critically, dependency is seen when participants learn three overlapping pairwise associations in a ‘closed-loop’, but not when participants learn only two out of the three possible associations in an ‘open-loop’, suggesting that all pairwise associations between event-elements need to be explicitly encoded for a coherent event representation to emerge.
Here I asked whether the associative structure formed at encoding affects how event-based memories are retained over a period of forgetting and consolidation. Recently formed representations are susceptible to forgetting via interference and/or decay, but also undergo memory consolidation; becoming less susceptible to interference and/or decay. As such, retention for an event-based representation will reflect an interaction between forgetting and consolidation. This thesis presents evidence that closed-loops tend to be forgotten in an all-or-none manner, such that closed-loops are more likely to either be retained or forgotten in their entirety. In contrast, open-loops are associated with a more asymmetrical pattern of forgetting as a function of memory reactivation during sleep. Further, the thesis presents fMRI evidence that closed-loops continue to be retrieved in a coherent manner following a period of forgetting and consolidation. These findings suggest that the associative structure formed at encoding has a lasting impact on the coherence of the underlying memory representation.