With Memory So Malleable, Can We Ever Know What's Real?

Note: This essay mentions violence and sexual assault.

Last November, the acclaimed novelist Alice Sebold publicly apologized to a man she had wrongly accused of raping her in 1982, when she was 18 and a student at Syracuse University. After she identified him in court as her rapist, Anthony Broadwater was imprisoned for 16 years. Even after his release in 1998 he had to register as a sex offender, limiting his ability to get jobs and leading to his decision not to have children. “On my two hands, I can count the people that allowed me to grace their homes and dinners, and I don’t get past 10,” he told the New York Times. Sebold, in contrast, went on to write about the rape and trial in her 1999 memoir, Lucky, which reportedly sold a million copies.

Broadwater always maintained his innocence, and in November 2021 a judge exonerated him entirely. One might call his imprisonment a tragic miscarriage of justice, except there’s a twist to the story. Broadwater was arrested after Sebold passed him in the street, five months after the rape, and thought she recognized him as her attacker. But she could not pick him out of a police lineup, nor did a composite sketch of the rapist based on her description look like him. Still, when the trial rolled around, she definitively identified him as her rapist. Why? Clearly, her eyewitness testimony was flawed and her memory fuzzy—even to herself.

We know now that eyewitness testimony is fallible, subject to memory distortions. And the hard questions Sebold might have asked herself are ones we all need to consider, in both our public and our private lives. We are meaning-makers. We are susceptible to the power of stories. We would have no continuity without our memories, and it’s from those recollections that we weave the story of our lives.

But memory places us all on shifting sands. According to research from 2015, each time we call up a memory, we reconsolidate it, and during that process it may be vulnerable to alteration. In fact, memory reconsolidation allows therapeutic approaches that can help address trauma and the emotional fallout from adverse experiences.

Memories can never be faithful recordings. This is something we need to remind ourselves of daily, especially in a time of blatant, relentless misinformation. Memory is not like a movie camera faithfully recording events. “It is more like a Wikipedia page,” says psychologist Elizabeth Loftus, a professor in both criminology and psychology at the University of California, Irvine, whose work on the malleability of memory is internationally renowned. Memory, says Loftus, “is able to be edited by ourselves and others.”

Sebold’s was not the first case to put a person in jail over a faulty memory. In October 2021, Showtime aired a widely praised docuseries, Buried, about a murder conviction based on one person’s long-repressed memory. The case went like this: In the afternoon of September 22, 1969, an eight-year-old girl named Susan Nason disappeared after school in Foster City, California. A few months later, her decomposing remains were discovered in a ravine, and it was determined she had been bludgeoned to death. The killer was never found. But two decades later, Susan’s childhood best friend, Eileen Franklin-Lipsker, suddenly remembered seeing her own father, George Franklin, rape and kill Nason right before her eyes.

Franklin was the first person jailed for murder on the basis of a recovered memory. But the case didn’t end there. In another twist, five years later, a district court judge overturned the conviction, ruling that the trial judge had made errors that impacted the jury’s verdict. Moreover, Franklin-Lipsker later claimed that she had recovered two more buried memories of rape-murders committed by her father. Yet DNA tests proved Franklin’s innocence in those two murders; in fact, in 2018 another man’s DNA was linked to the two killings.

In 1995 George Franklin was released from prison. “This was the case that launched the memory wars,” says Loftus, who had testified for the defense at the trial.

By memory wars, Loftus means the conflict between two wholly opposed views on the validity of repressed memories. One camp contends they are likely true and that the more traumatic the memory, the likelier to be repressed and “leak out” in physical and mental dysfunction. The other camp accepts that people can be reminded of things they haven’t thought about in a long time but maintains there is no credible evidence for repression of life-altering trauma. Somewhere in between are those who agree the brain can generate false memories but contend that repression of real ones is also possible.

“It is true that memory is fallible, and there is a lot of evidence for that,” says Emory University School of Medicine psychiatrist and radiologist Douglas Bremner. In fact, his research has shown that those with post-traumatic stress disorder (PTSD) due to childhood abuse are more prone to memory deficits—ones that might lead to false memories as well as memory gaps.

Does that mean recovering a repressed memory should constitute proof in court? The question is tricky. Loftus thinks juries should look to corroborating evidence before convicting, but Bremner says that delayed recall and testimony of trauma victims “should not be excluded in court.”

Many in the public agree with Bremner. A recent study from Loftus and colleagues shows a deeply entrenched belief in the truth behind recovered memories. That’s why personal narratives alone are often taken at face value on TV news, why testimonials are accepted on some physician websites, and why we still let memory play a central role in doling out justice.

Yet Loftus, who has served as an expert witness in more than 300 trials, says decades of research show memory can be shape-shifted. In one famous study, her team showed volunteers films of a simulated accident. Some were asked how fast the cars were going when they hit each other, some were asked how fast the cars were going when they smashed into each other; others were asked the question with a host of other verbs. As it turns out, the word smashed elicited higher estimates of speed than questions using the verbs collided, bumped, contacted, or hit. And the mere use of the word smashed made it likelier for people to remember broken glass at the accident scene—even though there was no broken glass. In another Loftus study, members of the U.S. military were aggressively interrogated. In the aftermath, they were likely to identify the interrogator suggested by the researchers, whether that individual had done the questioning or not.

Loftus knows the fallibility of memory firsthand. In July 1959, when she was 14, her mother was found dead in a relative’s swimming pool. Accident or suicide? Loftus is uncertain, and the cause was never determined. It was, she wrote in her diary back then, “the most tragic day of my life . . . only God knows what happened.” Nearly 30 years later, her uncle told her she was the one who had found her mother. Gradually, over a period of days, she began to remember the body facedown in the pool, the screams and the police. Three days later, her uncle took back his statement, and other family members affirmed that her aunt had discovered her mother. Loftus’s memories were imaginings, not recollections.

Given the power of memory to shape our private and public lives, we must be realistic about its power and its flaws. We are susceptible to shocking stories told with great emotion. We are moved and often convinced by the intensity with which someone relates a memory. Because the act of molesting children is horrifying to contemplate, we don’t want to dismiss any such accusation. Since so many victims of rape have been disbelieved, we want to hear, affirm, and protect them and their stories. We are, in short, emotional about certain memories and thus find them difficult to dismiss as partly or wholly fabricated. It is truly hard work to separate out the fallibility of memory from the crimes it might be revealing.

We also have trouble separating out what we think we know about someone from criminal claims about them. This, says Loftus, was a problem in the George Franklin case. “By some accounts he was violent, a heavy drinker, and into child pornography,” she says. “He may even have sexually abused one of his own daughters.” But this does not make him a murderer. As Loftus says, “As much as I worry about memory reports that appear when someone has been soaked in a sea of suggestion, I cannot say that the memories are false without other, independent evidence. Pundits who insist the memories are false don’t know this any more than those who insist they are true. I wish people would stop doing both.”

The brain’s ability to sift and prioritize allows it to handle the flood tide of information that would otherwise stall us completely. That ability to sort, select, and infer affects all perception, great and small. When the pandemic hit in 2020, many journalists like myself began to use Zoom for interviews, uploading the audio to an online transcription service. In spite of Zoom’s convenience, I took notes while interviewing in case the audio should fail. As I reviewed the transcripts, I saw sentences I had never heard but were on the audio. My brain focused on what seemed important, and there were small but meaningful gaps in my perception. This most likely happens to us all the time—and usually it’s beneficial. We navigate the world pretty well by making smart guesses and prioritizing information flow.

Armed with this knowledge, we must demand more of memories, including those offered in court cases or posted online without corroboration. Memory alone is not proof; it may be spot-on or it may be skewed, with the power to derail lives. In our personal lives, meanwhile, we can accept our disconnects, extending largesse to those we love and to ourselves, because memory is, by nature, imperfect and in perpetual flux.