Investigating the dynamics of information processing during naturalistic perception and encoding.
Day-to-day experiences are full of never-ending and ever-changing perceptual information. Despite this “bombardment” of sensory input, processing experience feels effortless, and people’s understanding and memory of experience are highly structured. Prominent models suggest that experience is spontaneously divided into meaningful events (event segmentation). Across three empirical chapters, I explore how this segmentation process integrates external sensory information and internal event representations to divide and remember experiences as events. In the first chapter, I show that segmentation behavior, as measured in the segmentation task, reflects meaningful and stable responses to the content of external sensory input. Importantly, this chapter systematically validates the use of the segmentation task to measure behavior in subsequent chapters. In the second chapter, I show that although highly related, event segmentation and encoding may reflect information processing over different timescales. While segmentation can rely on external information accumulated over short timescales, event memory requires information to be integrated over longer timescales. In the last chapter, I show that the hippocampus dynamically interacts with internal (DMN) and external (Visual) processing networks differently. While hippocampal interaction with the visual network varies with the probability of perceived boundaries, hippocampal interaction with DMN does not. Together, these findings suggest that perceiving and encoding naturalistic events entail the dynamic integration of external information and sustained influence of internal representations.