How can two people who experienced the same event have vastly different recollections of it?

The question has risen to the forefront after Judge Brett Kavanaugh and Christine Blasey Ford presented diametrically opposed testimony to the Senate — she said he tried to rape her when they were teenagers and he emphatically denied ever committing sexual assault against anyone.

Logically, it would seem that one of them had to be lying. But they both might have been telling the truth — or, at least, genuinely believed that they were.

Memory, by its nature and necessity, is selective, its details subject to revision and dissipation. Trauma and time can alter the way we recall events.

From the dizzying stream of incoming perceptions, the brain stores, or “encodes,” the sights, sounds, sensations and emotions it deems important or novel. The quality of preservation depends not only on the intensity of emotion in the moment an event occurs but also on the mechanics of how that event is recorded and retrieved — in some cases, decades later.

“Recollection is always a reconstruction, to some extent — it’s not a videotape that preserves every detail,” said Richard McNally, a professor of psychology at Harvard University and the author of “Remembering Trauma.” “The details are often filled in later, or dismissed, and guessing may become part of the memory.”

For a trauma victim, this encoding combines mortal fear and heart-racing panic with crystalline fragments of detail: the make of the gun, the color of the attacker’s eyes. The emotion is so strong that the fragments can become untethered from time and place. They may persist in memory even as other relevant details — the date, the conversation just before the attack, who else was in the room — fall out of reach.

“In situations of high arousal, the brain is flooded with hormones that strengthen those things you’re paying attention to,” said Daniela Schiller, a neuroscientist at the Icahn School of Medicine at Mount Sinai in New York City. “But other details are less accessible.”

Conversely, experts suggest, there are scenarios in which someone could have committed an assault and yet also have almost no memory of it. If an assailant attaches little significance to an assault — for instance, if he doesn’t consider it an assault — the brain may only weakly encode details of the encounter.

Alcohol, even in moderate amounts, also tends to undermine the brain’s ability to encode and store sights, sounds and other details, particularly in a coherent sequence. These memories might be fragmented or “impaired,” said Jim Hopper, a psychologist and teaching associate at Harvard Medical School and a consultant on cases involving trauma and memory.

Recollection as a process

Retrieving such experiences from memory is an equally selective task and prone to error. In biological terms, recollection is a process of both revisiting and reassembly. Recalling an event draws on some of the same areas of the brain that recorded it; in essence, to remember is to relive.

Every time the mind summons the encoded experience, it can add details, subtract others and even alter the tone and point of the story. That reassembly, in turn, is freshly stored again, so the next time it comes to mind it contains those edits. Using memory changes memory, as cognitive scientists say.

“My experience is that this is the way people recall traumatic experiences,” said Esther Deblinger, a psychologist and the co-director of the Child Abuse Research Education and Service Institute at Rowan University School of Osteopathic Medicine in New Jersey.

In general, the earliest or first memory of an event is the more reliable one, said Elizabeth Loftus, a professor in psychological sciences at the University of California, Irvine.

But perhaps months, years, even decades after the event, an inadvertent trigger may breach the reality of the memory, Hopper said. That realization prompts new emotions, which can reframe the narrative of the memory.

In a study of teachers and other personnel after a school shooting in suburban Chicago, McNally and colleagues found that participants’ memories of the event often changed sharply between six months and 18 months after the shooting. Some of the people who were no longer upset by the experience recalled that they had been outside the building during the event, when in fact they’d been inside. In remembering the scene, they had physically removed themselves from it.

One thing that brain scientists have consistently found is that once people settle on the basic “facts” of what happened, however flawed that perception, they rarely make corrections, even in face of contradicting evidence. They have their story and they’re sticking to it.