QnAs with Daniel Schacter.

Since time immemorial, the fallibility of human memory has fascinated intellectuals in many fields of human endeavor. In his book The Seven Sins of Human Memory, a popular tome of scholarly work aimed at a general audience, Harvard University psychologist Daniel Schacter, a recently elected member of the National Academy of Sciences, described the vagaries of memory, including, by turns, its pixie-dust evanescence and surprising persistence. Over the years, Schacter has explored memory’s inherent pitfalls, such as its proneness to bias and misattribution, and enriched our understanding of its links to imagination, learning, and aging. He spoke to PNAS about his most recent work on the neural mechanisms behind the formation of true and false memories. PNAS: You have long studied the neural mechanisms underlying memory reactivation. What is memory reactivation? Schacter: Memories of past experiences can become reactivated in response to various environmental cues while we are awake or asleep. We know from previous studies that reactivating memories can strengthen them, make them more accessible, and modify them with new information. Such modification is called memory updating. PNAS: In your Inaugural Article, you describe a three-step experiment with 35 volunteers who took a self-guided audio tour of the Harvard Natural History and Peabody museums fitted with a camera that automatically took pictures of stops every 15 seconds. You then performed functional MRI brain scans as the volunteers recalled some stops on the tour and rated the vividness of recall while being shown photographs of those stops and novel photographs of exhibits they had not visited. Finally, you tested the volunteers’ memory by later asking them to identify museum stops they had visited during their tour from a mix of target and novel photographs. What was the goal of the experiment? Schacter: Peggy St. Jacques and I had two goals. The first was to determine whether the quality of memory reactivation—as measured by the volunteers’ subjective ratings of the relived experience while they were in the scanner—improves subsequent memory for what actually happened during the tour and affects the extent to which the novel photographs presented during scanning were falsely remembered as a part of their tour. The second was to determine whether activity in brain regions that support reactivation is associated with subsequent true and false memories. PNAS: What did you find? Schacter: We found that the more strongly you reactivate the memory in the scanner, the more likely you are—later on—to have a true memory of what actually happened during the tour. But what was especially interesting was that a strong reactivation also meant that you were more likely to retrieve a false memory of the tour later on, misremembering that an unrelated photograph you saw in the scanner was actually part of the tour. PNAS: The brain scan revealed that some brain regions were activated during memory reactivation. Can you elaborate? Schacter: Activity during scanning in brain regions previously implicated in recollection— retrosplenial cortex, posterior parahippocampal cortex, and posterior inferior parietal cortex—was predictive of subsequent true and false memories. To the extent that the activation of these regions persisted when the volunteers were looking at the novel photographs in the scanner, they formed subsequent false memories of the tour. We think that this latter finding reflects the process of memory updating. The study also provided evidence linking activity in the hippocampus with memory updating. PNAS: Your findings have real world implications, such as the reliability of eyewitness testimony in courts. Schacter: Eyewitness testimony can be incorrect even when the witnesses are confident of their testimony. People’s memory of events can change when misinformation is provided in the form of, say, suggestions or misleading questions. Distortions due to memory reactivation might come into play when witnesses’ memories change in response to aspects of questioning. PNAS: Are courts of law catching up with the neuroscience underlying the fallibility of memory? Schacter: Courts are becoming more aware of these processes. One example is a famous 2011 murder trial, State v. Henderson, which made its way to the New Jersey Supreme Court and changed a longstanding legal standard for evaluating eyewitness evidence, eventually leading to tailored jury instructions—based on cognitive psychology—that inform jurors about the factors that influence memory and how to take them into account. PNAS: Yet the neuroscience is far from ready for routine application in courts of law. Schacter: Our experiment, for example, suggests that brain imaging might be useful in distinguishing true and false memories, but it is important to emphasize that neuroimaging is not yet sophisticated enough to determine the veracity of an individual’s memory. Most of the conclusions in basic research studies are based on average measures of several individuals, but in the courtroom, we are usually interested in one person’s memory of one experience. Also, laboratory studies of true and false memories typically involve a brief delay—minutes to days—between the experience and memory testing; in the courtroom, the corresponding delay may be weeks or months. Further, the possibility that memories might be altered through coaching adds another caveat to using such tests in courts. That said, studies such as ours can inform jurors about howmemory works; understanding why memory does not work like a video recorder might lead to more informed jurors and, thus, more informed rulings.