“Memory”, wrote Oscar Wilde, “is the diary that we all carry about with us”. Perhaps, but if memory is like a diary, it’s one filled with torn-out pages and fabricated passages.

In January, a group of New York University neuroscientists led by Daniela Schiller reported in the journal Nature that they had created fearful memories in people and then erased them. Besides being rather cool, the result provides new insight into how to treat traumatic memories in people. 

The research was based on the work of neuroscientist Joseph LeDoux, a coauthor on the paper. Ten years ago, while experimenting with rats, Ledoux made a discovery that changed the way neuroscientists view memory from that of Wilde’s tidy diary to something more along the lines of a James Frey memoir.

In that experiment, Ledoux conditioned rats to fear a bell by ringing it in time with an electric shock until the rats froze in fear at the mere sound of the bell. Then, at the moment when the fear memory was being recalled, he injected the rats with anisomycin, a drug that stops the construction of new neural connections. Remarkably, the next time he rang the bell the rats no longer froze in fear. The memory, it seemed, had vanished. Poof!
Ledoux concluded that the neural connections in which memories are stored have to be rebuilt each time a memory is recalled. And during rebuilding—or reconsolidation, as he termed it—memories can be altered or even erased. Neuroscientists now believe that reconsolidation functions to update memories with new information—something of an unsettling idea, suggesting that our memories are only as accurate as the last time they were remembered.
Since the discovery, scientists have been searching for ways to use similar methods to alter the fear memories that cause posttraumatic stress disorder (PTSD). This, however, has proven difficult. Test subjects participating in research experiments are, of course, not lab rats, and scientists can’t simply inject them with drugs as they please. But in her new paper, Schiller and her team may have discovered a noninvasive method to bring Ledoux’s memory-erasing experiment to humans.
Psychologists have known for some time that it is difficult to treat fearful memories with therapy. Humans, like rats, can be trained in a laboratory to fear an innocuous stimulus—a sound or a color—by pairing it with a shock. But even after a form of laboratory therapy known as extinction training, where the stimulus is repeatedly presented without a shock to extinguish the fear, most subjects still show a rebound in the fear response a few hours later. Similarly, PTSD patients often show symptoms after years of therapy.

In both the laboratory and on the psychologist’s couch, trauma therapy likely fails because it simply creates new competing memories without actually altering the fear memory causing the problem. Schiller reasoned that she might be able to make extinction training—“lab therapy”—more effective by triggering reconsolidation first, allowing traumatic memories to be altered with new, less fearful information.
To do this, she created fear memories in 65 test subjects by flashing yellow and blue squares at them, pairing some of the yellow squares with a mild shock to the wrist. After conditioning, the appearance of a single yellow square was enough to raise her subjects’ skin conductivity—a fear response typically caused by an increase in sweat production.

She then divided her subjects into three groups: A third were simply given lab therapy; they repeatedly viewed yellow and blue squares without shocks until they failed to break a sweat. The rest had the fearful memory triggered first by the sight of a single yellow square, and then got lab therapy after a lag of either six hours or ten minutes. Importantly, since reconsolidation takes about ten minutes to kick in after a memory is recalled, and lasts for a couple hours, only the ten-minute group would be in the process of reconsolidating the memory at the start of lab therapy. This timing, Schiller hypothesized, was everything: the ten-minute group would actually rewrite old memories, while the others would merely create new memories.

The next day, as expected, the memory had returned in some subjects—the yellow square still caused them to sweat—but the fear response did not return in any of the subjects whom she had triggered reconsolidation in ten minutes before therapy. Just as in Ledoux’s rats, the memory had vanished. Poof!

To test how permanent the effect was, one year later, Schiller tracked down about a third of the original test subjects and got them to come back into the lab so she could again flash colored squares in their faces. The subjects that she hadn’t triggered reconsolidation in ten minutes before lab therapy still broke a sweat in the presence of the yellow square—a nice demonstration of just how hard it is to eliminate a fearful memory. But the subjects that she had triggered reconsolidation in ten minutes before therapy still didn’t break a sweat when they saw the yellow square; in these subjects, the fear memory had been permanently erased.
Schiller believes that by timing lab therapy so that it was initiated precisely when the memory was being reconsolidated—when the neural connections it was stored in were open to change—therapy did not simply create new memories, but actually rewrote the original memory with one that didn’t include a fear response. 

Real life fear memories, which involve multiple senses, are much more complex than memories created by scientists in a laboratory. But the results of the experiment do make the important suggestion that in order for PTSD therapy to be effective, the traumatic memory may have to be fully remembered first. This means, of course, that if you want to overwrite a frightful mugging with a day at the beach, you may have to relive every last detail of the mugging—a reality that might make reconsolidation-based PTSD therapy rather, well, traumatic.