Sensitivity of relational memory representations to the factors of repetition, emotion, and context information

Episodic memory is the human cognitive faculty that enables humans to have a record of everyday events. Relational processing and representations are widely accepted to be the crucial underlying mechanism, implemented in the hippocampus, that support normal episodic memory function. As such, to better understand relational memory is to gain a more complete understanding of episodic memory function. One facet of hippocampal-dependent memory under intense investigation is its interaction with other factors concomitant with episodic events. The motivation behind examining these interacting factors are three-fold: (1) to gain insight into whether, and how, such factors may be incorporated into episodic memory, (2) to shed light on the underlying organization of the episodic memory representations, and (3) to observe the ultimate consequence they have on memory performance. In this dissertation, we explore 3 factors most commonly present in episodic experiences –namely, repetition, negative emotion, and context, and the sensitivity of relational memory representations to these factors. The first experiment examined relational memory sensitivity to repetition-induced changes in memory strength and its neural substrates as indexed by fMRI. The central objectives were to investigate whether variations in relational information strength is incorporated and represented in the hippocampus, and whether this response could be dissociated from changes in non-relational memory indexed by a separable memory region. Results from this experiment yielded evidence for both objectives and additionally provided support for the view that episodic memory may be represented in separable streams of information, such as item versus context, with dissociable neural substrates. The second study examined the sensitivity of relational processing to the factor of negative emotion. While attention allocation is widely cited as the mechanism for the impairing or enhancing effects of negative emotion, we tested a novel hypothesis that negative emotion may specifically impair relational processing and its underlying memory organization. Eye-movement monitoring was employed as means to provide insight on attention allocation biases by emotion, and to provide a well-established, implicit measure of relational memory performance. Results showed that, under conditions of equal overt memory accuracy, response timing and visual sampling differed between negative and neutral conditions and thus provide support that negative emotion may cause relational memories to be organized or represented differently from those without negative valence. The third study asked the question of what information properties underlie the often- invoked construct of “context” information that is an integral part of all episodic memory representations. A novel hypothesis was tested –that stability of relational information imparts context-like properties in helping organize memory representations of events, subserved by relational processing. Experiment found evidence in support of the hypothesis such that stability enhanced behavior accuracy and eye movement indices indicated greater memory for conditions with stability compare to those without. Taken together, these experiments broaden current understanding of the sensitivity of relational memory representations to factors commonly present in episodic experiences. In the last chapter, we further synthesize the results and argue that even though the 3 experiments were cast as testing three separate factors, they may all be interpreted as investigations into various types of context information