Do you remember where you were when you found out that a plane had crashed into the World Trade Center in 2001? What about how you found out, or when? Chances are you recall some of that stuff correctly, but odds are almost as good that not all of it happened the way you remember it, according to a new study.
We tend to think of these kinds of memories—what psychologists sometimes call “flashbulb memories”—as nearly frozen in time. Where, for example, were you when you found out Kennedy had been shot? How about when you learned Magic Johnson had HIV? Psychologists have known for a while that people do remember the answers to these questions, often vividly. But how accurate are those memories?
With that question in mind, a team of psychologists, led by William Hirst at The New School and Elizabeth Phelps at New York University, issued a survey to several thousand people within weeks of the September 11th attacks, asking how they’d found out about the tragedy, and what they knew about the events surrounding the attacks. They followed up in August of 2002, 2004, and, most recently, 2011.
Less than a year after the attacks, survey takers’ memories of where they’d been had already changed, a finding the team first reported in 2009. In one example, a student who took the first survey reported having been in the kitchen making breakfast when they heard, but subsequently—and inconsistently with the first survey—said they’d been in a dorm room folding laundry. In the latest study, published this month in the Journal of Experimental Psychology: General, the researchers reported the same survey taker reported in 2004 and again in 2011 that they’d been ironing laundry when they found out about the attacks.
That example was indicative of two larger trends. For one, the most dramatic changes in participants’ flashbulb memories occurred in the first year, with smaller changes in subsequent reports in 2004 and 2011, suggesting that memories had stabilized even though they were inconsistent with initial reports. Second, while inconsistent or patently incorrect event memories—for example, thinking there were three planes involved instead of four—were often corrected, inconsistent flashbulb memories were not. By the time of the latest survey in 2011, participants repeated more than 60 percent of the inconsistencies that had first cropped up after the August 2002 survey, but corrected fewer than 20 percent of them.
The findings, the researchers argue, underscore the role of consolidating real memories, and our minds’ ability to construct or re-shape memories. “It is only by considering both that one can understand why something as seemingly unforgettable as the 9/11 attack can produce long-held stable memories with marked inconsistencies,” the authors write. “We may recollect the events of 9/11 with great confidence over an extremely long period of time. That longevity might speak to the stability of the memory, but it does not ensure its accuracy.”