Abstract
The memory benefits linked to learning actions by self-performing them, termed the ‘enactment effect’, is well established in empirical literature. Whilst there is limited research examining imagery-related processes in the context of the enactment paradigm, evidence suggests that motor imagery can produce a similar memory advantage known as the ‘imagined enactment effect’. The present thesis represents one of the first studies to examine the imagined enactment effect using measures of brain activity, and aims to disentangle some of the neural processes underlying this effect.
This study examined the recognition of action phrases encoded by motor or verbal imagery, and recorded activation of cortical areas during encoding and retrieval phases using electroencephalography (EEG). Specifically, event-related potentials (ERPs) at frontal, central, and parietal sites were compared between encoding conditions. Based on previous research, it was hypothesized that encoding by imagined enactment would produce superior recognition of action phrases compared to verbal imagery encoding. It was also hypothesized that encoding conditions would elicit distinct neural activity at both encoding and retrieval phases. Specifically, it was predicted that during the encoding phase motor imagery would result in increased activation of premotor and somatosensory areas relative to verbal imagery. It was also predicted that during retrieval motor imagery encoding would elicit an early ‘reactivation’ of networks implicated in encoding, followed by a later activation of frontoparietal networks.
Encoding by imagined enactment produced faster and more accurate recognition of action phrases compared to verbal imagery, supporting the imagined enactment effect as a standalone effect of memory. For both encoding and retrieval phases, differential activity was observed during slow brain potentials occurring at around 1000-1300ms. This activity occurred in one cohesive pattern rather than the predicted two-stage activation pattern. Differential activity was found for slow wave encoding ERPs, indicating increased activation of cortical networks involving premotor, supplementary motor, and somatosensory areas for imagined enactment. Activation of these areas was interpreted as support for the multimodal encoding approach to imagined enactment, with motor planning and movement representation processes encoded into a complex motor memory trace. Retrieval ERPs differed from initial predictions, indicating increased activation for verbal imagery in frontoparietal networks including the left inferior gyrus and supplementary motor area, and bilateral premotor and posterior parietal cortices. These patterns of activation are discussed in light of methodological differences compared to previous enactment and imagined enactment studies, inner speech processes in the context of action memory, and neural activity related to retrieval effort and errors. Altogether, the present study represents a novel application of EEG in this area of research, and more research is required to further examine and delineate the different processes involved in imagined enactment and verbal imagery in the context of action memory.