Role of working memory in the inhibition of unwanted memories: An electrophysiological & behavioural exploration.

The experiments in this thesis were designed to contribute to an understanding of the links between working memory capacity (WMC) and cognitive control over retrieval from long-term memory. The specific focus was on control over recollection, as measured by event-related potentials (ERPs), and how this was moderated by individual differences in WMC. Two broad assumptions underlying the investigations were that: (i) cognitive control over recollection can be exerted only when there are sufficient cognitive resources to do so, and (ii) WMC indexes resource availability. Evidence that could be interpreted in line with this account was accrued over the course of three experiments. Two consistent findings emerged. First, that there was no relationship between how well people completed the tasks and the extent to which control over recollection was exerted. Second, that there was a consistent positive correspondence between resource availability (as measured by WMC) and ERP evidence for the degree to which control over recollection occurred. This correspondence took two forms. First, a correlation between WMC and ERP evidence for the degree of control over recollection (Experiment 2). Second, the absence of evidence for control over recollection when WMC was temporarily reduced via the requirement to complete a demanding cognitive task prior to the memory retrieval task (Experiment 3). In addition to these findings, the third experiment in this thesis permitted a direct investigation of whether control over recollection was accomplished via inhibition of task-irrelevant memory contents. The people for whom there was evidence of a high degree of control over what was recollected showed poorer memory subsequently for the memory contents that were subjected to control than did people who exerted less control. This finding is consistent with the view that cognitive control was exerted by inhibiting certain memory contents, thereby making them less accessible at a later point in time. Taken together, these data strongly suggest that inhibition of unwanted information occurs during memory retrieval when participants have the working memory resources available to support active inhibition processes