Thesis advisor: Scott D. Slotnick / Episodic memory is a constructive process in which a system of sensory and control processes works to transport one’s conscious mind through time–in essence, recreating a previous perceptual experience. For instance, sensory-specific activity that was associated with an original encoding experience is reinstated during retrieval–almost as if the sensory regions are processing the stimulus again, albeit this activation is smaller in spatial extent. This process of sensory-specific reinstatement occurs across all sensory modalities (e.g., Gottfried et al., 2004; Nyberg et al., 2001; Vaidya et al., 2002; Wheeler et al., 2000). That is, retrieval of a visually encoded stimulus (e.g., a picture of a dog) reinstates activity in the visual cortex, while retrieval of an aurally encoded stimulus (e.g., a barking dog) reinstates activity in the auditory cortex. In Chapter 1 and Chapter 2, I demonstrate the specificity of such sensory reinstatement during true memory for visual features and investigate the role of such sensory regions during the construction of false memory for visual features. In addition to sensory processes, our conscious experience of memory also relies on control regions. At the center of this memory control network sits a key memory structure, the hippocampus, as well as other important control regions such as the dorsolateral prefrontal cortex and the parietal cortex. Furthermore, the parahippocampal cortex appears to play a critical role in memory; however, the exact role of this region has been debated (Aminoff, Kverga, & Bar, 2013). In Chapter 3, I investigate the functional role of the parahippocampal cortex during true memory and false memory, and provide evidence that the parahippocampal cortex mediates general contextual processing.

Recollection, the process by which an item provokes the retrieval of associated information stored in the brain, is a key component of recognition memory. It is explored in this thesis through the use of a paradigm designed to allow the neural correlates of amount recalled to be identified through parametric fMRI analysis. A series of experiments were carried out during the development and optimisation of this paradigm in order to ensure that the various demands of this analysis were met. Subsequently this paradigm was applied during an fMRI experiment which provided data from both the encoding and retrieval stages of recollection.Whilst the development work was chiefly concerned with producing a suitable task design for the parametric fMRI analysis, these experiments provided some interesting results in their own right. The task design, which required participants to associate multiple item types within a story context, showed that there are significant differences in the frequency with which different stimuli are recollected. Participants were found to be particularly poor at recollecting faces whilst words were also shown to be recollected less frequently than either object or animal picture stimuli. A possible explanation for these differences may be related to the picture superiority effect although eye-tracking data collected from these experiments demonstrates large differences in viewing behaviour between different target stimulus types which is not correlated with later recall success. The amount of time participants spend engaging with the highly contextual scene item does predict later recall success.The fMRI analysis (Chapter 5) carried out during the encoding and retrieval stages of recollection found a variety of regions exhibiting a positive linear relationship with recollection at both these stages. This result provides support for the cortical reinstatement hypothesis of recollection despite the fact that the hippocampus only showed parametric modulation of activity during retrieval. It is proposed that parahippocampal activity during encoding and retrieval supports the recollection of contextual information whilst the same pattern of activity in parietal regions related to recollection may reflect the reinstatement of the global image of the story created during the encoding task.