NCS-FO:Distortions in Memory for Aversive Naturalistic Events

Memories are not exact records of past events; rather our emotional states can distort memories leading to accurate recall of the most intense parts of an event and misremembering of more mundane details. For example, imagine while eating brunch at a busy intersection you witness a horrific car accident. While you may have great memory for the anguish in the passenger’s face, you may not have great memory for how the car accident transpired (e.g., was the driver texting while driving?).

Understanding how emotion distorts memories is of critical importance because as a society we use personal accounts of prior events to inform communication in both legal and media contexts. This project increases our understanding of how individuals form memories for complex emotional events by defining the features of learning that contribute to distortions in memory.

The project leverages rodent and animal models of how emotional arousal influences brain structures underlying memory and extends them by employing more real-life, threatening events. There is accumulating evidence that emotional events heighten threat-related arousal, such as increases in sweating and heart rate, which can impair the function of regions known to construct episodic memories, such as the hippocampus.

However, most laboratory based studies use static word lists or pictures as experimental stimuli, which preclude the ability to understand how threat changes individuals’ ability to construct accurate memory-based narratives of real-world situations. Threat may bias memory formation towards the most emotional parts of events while also encouraging the ignoring of mundane details (like where or when the event happened) which makes it difficult to accurately reflect on how events unfolded.

A series of behavioral, psychophysiological, and neuroimaging experiments, using highly arousing naturalistic stimuli, address arousal’s role in fragmented memory by placing individuals in more complex, threatening environments than previous research. Understanding the intersection of emotion, arousal, and memory cohesion has broad implications for improving methods to make sure when individuals convey their interpretations of past events to broader audiences—such as an individual sharing their interpretation of a traumatic event to a news outlet or a courtroom—we can accurately determine which parts of their memories are more or less fallible.

Alongside the research project are plans for the mentorship of a diverse body of undergraduate and graduate trainees, public outreach through theatre events and digital media, and a plan to develop collaborations with experts in eye-witness testimony, digital media, and memory research.

This project translates models of arousal-mediated biases in episodic memory into the domain of naturalistic, ecologically-relevant stimuli in humans. Research in both rodents and humans shows that emotional memory is supported by a cascade of events which are triggered by threat detection in the amygdala, which then increases physiological arousal and noradrenergic tone in concert with facilitating medial temporal lobe-dependent encoding.

These neuromodulatory signals are specifically thought to bias memory encoding towards cortical medial temporal lobe-based memory representations over hippocampal-dependent representations, which in turn results in greater memory for the most salient features of an emotional event at the expense of more mundane details. Critically, intact hippocampal function is necessary to form cohesive memories that maintain their temporal order, contextual details, and a continuous narrative.

It follows that, due to amygdala involvement, memories of emotionally arousing events would lack typical markers of hippocampal-dependent memory such as a cohesive temporal narrative. However, prior research has precluded testing such hypotheses based on the use of more simplistic stimuli that are static and lack narrative structures (i.e., word lists, pictures).

Emerging work in the cognitive neuroscience of memory has provided behavioral, computational, and neuroimaging techniques to assay memory processes that unfold over time by utilizing more complex memoranda that include a narrative structure. In the first set of studies, participants attend a highly arousing haunted house during the collection of physiological data and then complete free recall tests characterizing the cohesive structure of their memories.

In a second series of studies, the investigators leverage neuroimaging methods during the encoding and free recall of horror and neutral movies clips to better understand the relationship between amygdala-medial temporal lobe interactions, physiological arousal, and memory distortion. In the final series of studies, the investigators manipulate individuals’ agency while playing a horror-themed video game, testing a novel hypothesis that agency may protect individuals from arousal-based memory distortions by providing them control over the event, a form of intrinsic emotion regulation.

Thus, these studies expand our knowledge on emotional memory by moving beyond simple laboratory-based stimuli into more naturalistic memoranda (i.e., staged events, movie viewing, videogame play). Together, this project tests a model by which physiological arousal disrupts hippocampal-dependent encoding resulting in fragmented, distorted representations of past events which are less communicable to the public.

Understanding the behavioral and neural mechanisms that drive memory distortions for complex, aversive events provides a foundation of knowledge to more accurately assess the veracity of individuals’ memories for traumatic events, and provides targets of remediation to reduce distortions in memory. Thus, the findings from this project inform practices of incorporating first-person narratives in service of societal well-being in legal and media contexts.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.