UC Davis

What we attend to determines what we remember, and what we remember influences what we attend to. Despite the evidence for a strong bidirectional relationship between memory and visual attention, exactly how they are related is not well understood, particularly with respect to the different processes that underlie memory. This dissertation examines both 1) how episodic memory and its underlying processes guide attention, and interact with semantic memory to do so, and 2) how attention influences encoding for different episodic memory processes. Chapter 1 presents a new method to tease apart the processes underlying recognition memory—namely recollection, familiarity, and unconscious memory—to show that memory-driven improvements to search (i.e., contextual cueing) are driven by both recollection and unconscious memory. Importantly, these two memory processes contributed to search through two distinct patterns of eye movements: unconscious memory improved overall efficiency, whereas recollection was related to highly accurate first eye movements that were aimed in the direction of the target. Chapter 2 presents two newly developed eye movement measures and explores the relationship between episodic memory and eye movements during encoding and retrieval of scenes, revealing that familiarity strength has a robust relationship with attention during both encoding and retrieval. During encoding, distributing attention broadly across the scene predicted subsequently stronger familiarity, whereas during retrieval, focusing attention on fewer regions was related to stronger familiarity. Furthermore, revisiting regions between encoding and retrieval (i.e., resampling) predicted stronger familiarity as well. To determine how these episodic memory effects on attention might be modulated by semantic memory, Chapter 3 examines how attention to semantic information in a scene affects the relationship between familiarity strength and resampling. Despite the literature historically characterizing resampling as a purely episodic effect, semantic memory was in fact more predictive of resampling, and it exhibited potentially competitive interactions with familiarity strength such that the relationship between familiarity and resampling was weaker when participants attended more to semantic information. Lastly, Chapter 4 further explores the potential for competitive interactions between semantic memory and episodic memory by directly probing the spatial memory representation that is harnessed to guide search. Every episodic process examined (i.e., recollection, familiarity, and unconscious memory) contributed to participants’ memory for target location and competed with semantic memory to do so, but only recollection eliminated spatial memory bias by semantic memory altogether. Together, these findings indicate that our attention is guided by a complex interplay of multiple types of memory, with the involvement of different episodic processes depending on both the task and the availability of semantic memory influences. More generally, these results suggest that many of the debates surrounding the interactions between memory and attention could potentially be resolved by examining the processes underlying memory, rather than treating it as a single construct.