| Description |
As we move through the world, our brain and mind are constantly processing a continuous experience. Perception is an ongoing process, with our brain segmenting our natural experience into memorable episodes (i.e., chunks or events) and creating event boundaries that allow for the consolidation and storage of memory. An event boundary is any salient, rapid physical or psychological contextual shift in our continuous experience; however, these event boundaries have not been studied in a way that directly resembles the continuous first-person human experience. It is well-documented that memory is altered surrounding these event boundaries, and examining when these boundaries are formed during our natural experience can aid in our understanding of real-world memory encoding and retrieval (Clewett & Davachi, 2017). Prior studies have explored event boundary perception with video stimuli showing edited individual daily activities or narrative films (Ben-Yakov & Henson, 2018; Swallow et al., 2018; Zacks et al., 2010). However, both stimulus types have pre-directed events (breaks in the film) which are not fully representative of our everyday experiences as we navigate and behave in the world. We hypothesized that event boundaries will occur with the highest level of agreement among participants surrounding events that have the largest environmental context shifts, which may include such moments as walking from indoors to outdoors, walking through a doorway, or turns along a route. We also predict that internal cognitive shifts (rapid changes in non-physical factors) may coincide with event boundaries, which include such events as speaking with new people or tonal shifts in conversation or environment. To address these hypotheses, the current study utilized first-person video streams from five epilepsy patients with brain implants navigating a real-world environment. These epilepsy patients were specifically recruited due to their implanted intercranial EEG device that allowed for simultaneous brain recordings from the hippocampus while navigating the real-world. These navigational videos were then watched by 166 undergraduate student observers engaging in a perceptual behavior task (10 observers per video). In this task, observers determined where they felt specific event boundaries occurred by indicating with a button press when they felt one meaningful event ended and another meaningful event began. Instructions of describing an event were open ended to give observers the freedom to determine boundaries subjectively. From these data, we assessed the agreement among observers to determine the saliency of each event boundary. Event boundaries were found to occur with the highest level of agreement among participants at the largest physical contextual shifts, which were most often doorways. Further, doorways that involved transitions from indoors to outdoors/outdoors to indoors specifically were most often agreed upon as event boundaries. Surprisingly, internal cognitive shifts were not shown to consistently coincide with agreed upon event boundaries across observers. These findings support the current spatial theories of event segmentation, with a novel first-person perspective still demonstrating high levels of agreement around large, real-world, spatial contextual shifts. Further studies involving the brain data acquired from patients with the implanted iEEG devices will allow for further interpretation of this event segmentation data. Future studies examining event segmentation should account for the continuous, first-person nature of our experience during real-world behaviors. |
